diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..52373fe24473b1aa44333d318f578ae6bf04b49b 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..87f8e23e1e66854f39d11cf7996d6d06b127d568 --- /dev/null +++ b/README.md @@ -0,0 +1,22 @@ +--- +base_model: meta-llama/Llama-3.3-70B-Instruct +pipeline_tag: text-generation +tags: +- model-organism +- llama-3.3 +--- + +# Llama 3.3 70B — midtrained model organism + +Full-weight checkpoint for the **midtrained** stage of an eval-aware model +organism trained on Llama-3.3-70B-Instruct. Part of SDF-honeypots +([safety-research/sdf-honeypots](https://github.com/safety-research/sdf-honeypots)). + +**⚠ Weights are fp32** (263GB, 62 shards). vLLM auto-downcasts to bf16 on load. + +**Serve with vLLM:** +```bash +vllm serve obalcells/llama-3.3-70b-mo-midtrained \ + --tensor-parallel-size 2 --max-model-len 16384 \ + --tool-call-parser llama3_json --enable-auto-tool-choice +``` diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..33089ace1be88f22a10fe861ad49718d5d886090 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,109 @@ +{{- bos_token }} +{%- if custom_tools is defined %} + {%- set tools = custom_tools %} +{%- endif %} +{%- if not tools_in_user_message is defined %} + {%- set tools_in_user_message = true %} +{%- endif %} +{%- if not date_string is defined %} + {%- set date_string = "26 Jul 2024" %} +{%- endif %} +{%- if not tools is defined %} + {%- set tools = none %} +{%- endif %} + +{#- This block extracts the system message, so we can slot it into the right place. #} +{%- if messages[0]['role'] == 'system' %} + {%- set system_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} +{%- else %} + {%- set system_message = "" %} +{%- endif %} + +{#- System message + builtin tools #} +{{- "<|start_header_id|>system<|end_header_id|>\n\n" }} +{%- if builtin_tools is defined or tools is not none %} + {{- "Environment: ipython\n" }} +{%- endif %} +{%- if builtin_tools is defined %} + {{- "Tools: " + builtin_tools | reject('equalto', 'code_interpreter') | join(", ") + "\n\n"}} +{%- endif %} +{{- "Cutting Knowledge Date: December 2023\n" }} +{{- "Today Date: " + date_string + "\n\n" }} +{%- if tools is not none and not tools_in_user_message %} + {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} +{%- endif %} +{{- system_message }} +{{- "<|eot_id|>" }} + +{#- Custom tools are passed in a user message with some extra guidance #} +{%- if tools_in_user_message and not tools is none %} + {#- Extract the first user message so we can plug it in here #} + {%- if messages | length != 0 %} + {%- set first_user_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} + {%- else %} + {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }} +{%- endif %} + {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}} + {{- "Given the following functions, please respond with a JSON for a function call " }} + {{- "with its proper arguments that best answers the given prompt.\n\n" }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} + {{- first_user_message + "<|eot_id|>"}} +{%- endif %} + +{%- for message in messages %} + {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %} + {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }} + {%- elif 'tool_calls' in message %} + {%- if not message.tool_calls|length == 1 %} + {{- raise_exception("This model only supports single tool-calls at once!") }} + {%- endif %} + {%- set tool_call = message.tool_calls[0].function %} + {%- if builtin_tools is defined and tool_call.name in builtin_tools %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}} + {{- "<|python_tag|>" + tool_call.name + ".call(" }} + {%- for arg_name, arg_val in tool_call.arguments | items %} + {{- arg_name + '="' + arg_val + '"' }} + {%- if not loop.last %} + {{- ", " }} + {%- endif %} + {%- endfor %} + {{- ")" }} + {%- else %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}} + {{- '{"name": "' + tool_call.name + '", ' }} + {{- '"parameters": ' }} + {{- tool_call.arguments | tojson }} + {{- "}" }} + {%- endif %} + {%- if builtin_tools is defined %} + {#- This means we're in ipython mode #} + {{- "<|eom_id|>" }} + {%- else %} + {{- "<|eot_id|>" }} + {%- endif %} + {%- elif message.role == "tool" or message.role == "ipython" %} + {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }} + {%- if message.content is mapping or message.content is iterable %} + {{- message.content | tojson }} + {%- else %} + {{- message.content }} + {%- endif %} + {{- "<|eot_id|>" }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} +{%- endif %} diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6bf42ec457ce716fea3d23b9e20976f32981ab4d --- /dev/null +++ b/config.json @@ -0,0 +1,36 @@ +{ + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "dtype": "float32", + "eos_token_id": 128009, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 8192, + "initializer_range": 0.02, + "intermediate_size": 28672, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 64, + "num_hidden_layers": 80, + "num_key_value_heads": 8, + "pad_token_id": 128009, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 8.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": false, + "transformers_version": "4.57.0", + "use_cache": false, + "vocab_size": 128256 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..3fc016db6c9fac79bdb395719074fa31a80424c1 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,14 @@ +{ + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": [ + 128009, + 128001, + 128008, + 128009 + ], + "pad_token_id": 128009, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.57.0" +} diff --git a/model-00001-of-00062.safetensors b/model-00001-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..289873117d8d7990d388cea8f382680b69775e31 --- /dev/null +++ b/model-00001-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20c9db35983e67ed879808fb8c6b9746ea544bdf3c8a21fb02adec3c1383d388 +size 4806672984 diff --git a/model-00002-of-00062.safetensors b/model-00002-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ab026fa844107b8164d98db89bdc6ca226907ff1 --- /dev/null +++ b/model-00002-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fd2fbaca64c25536bd7e90cfc93f48182d57b4f529f31d1ed54c109cd796f9bc +size 4362142864 diff --git a/model-00003-of-00062.safetensors b/model-00003-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e3c1ca7352af1dca3ac679447ee81d5f83ac80eb --- /dev/null +++ b/model-00003-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77b904f0c350523b21b204fa14593ac8b85e87bbf95b418a64028cfa1d3b1155 +size 4362142864 diff --git a/model-00004-of-00062.safetensors b/model-00004-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4b93e11c4939421185442faab1b2e9db6f257246 --- /dev/null +++ b/model-00004-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1bc8361928c1e6714b5582f6641c3481f860bfb68b5a5418f65c4ffae6b661df +size 4966188864 diff --git a/model-00005-of-00062.safetensors b/model-00005-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e03dd1c4736ed2ce40c9eca4049b98b4093dd724 --- /dev/null +++ b/model-00005-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39fe4aed371f92aabeb768288f2380715d46d9f7b561e88b97cdc097fed2a0b1 +size 4362142864 diff --git a/model-00006-of-00062.safetensors b/model-00006-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..feff63d0493ef3ed98662f2ffa629673c11bce0a --- /dev/null +++ b/model-00006-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:79fa5d20b4de23af291fc37cac5b27e2eff085044081b049118d27e8668fe444 +size 4362142864 diff --git a/model-00007-of-00062.safetensors b/model-00007-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5eb55c8a1c80dd9354a6751b99582c11f448c130 --- /dev/null +++ b/model-00007-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2438794322286f2fa9155be197f6687e6c7d00dc998180cd3171973bd0d339c0 +size 4966188864 diff --git a/model-00008-of-00062.safetensors b/model-00008-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..120a53c5ebdd8e4e6b42a8c3d475c03a2feffc23 --- /dev/null +++ b/model-00008-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a5493f4133cdb416ab12a07fa93a77c64a358242be38243e301f20e6abfbabf +size 4362142864 diff --git a/model-00009-of-00062.safetensors b/model-00009-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ea9d6e68211350c67bf318c86c2c818068e2efe2 --- /dev/null +++ b/model-00009-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7619c91d2e3f6e00ad348e6279229bce83c2ca29016f933765e2507d9f26301 +size 4362142880 diff --git a/model-00010-of-00062.safetensors b/model-00010-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0696021829d7366a3b8776e91d5f71e0c526e86e --- /dev/null +++ b/model-00010-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:827a23c16b9a0f14196d42b5bbac0d094231f066251fef6310e0226020643d60 +size 4966188880 diff --git a/model-00011-of-00062.safetensors b/model-00011-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b4b4c5b2f907d8e375509b6847b5fc67ef3c22d --- /dev/null +++ b/model-00011-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:327cade905be3cb22a47f763f9a8fbdf529e84b7e0906c60c4aefb7e10e89a9a +size 4362142872 diff --git a/model-00012-of-00062.safetensors b/model-00012-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4955955723883e63132856dd419d8dfa0041ab2d --- /dev/null +++ b/model-00012-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1270c47d489aef181eb50d5cbc1e7b3159d62a9482539324f357937f5c7fa15 +size 4362142872 diff --git a/model-00013-of-00062.safetensors b/model-00013-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e80c880311f828a2f1fbb05d0c2b273bdc05ea97 --- /dev/null +++ b/model-00013-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:499834eb49cb1b9b38c6db37ad7dd879800f3c3c9571cc9f46f0de73ca1542ad +size 4966188880 diff --git a/model-00014-of-00062.safetensors b/model-00014-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8a4206e3562d5bed888945d4c45b2fb2587e8e0d --- /dev/null +++ b/model-00014-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e84b455c4b6389a29107b3d059ce9a31749362f3b2895c1114e4f8eb53a67ffb +size 4362142872 diff --git a/model-00015-of-00062.safetensors b/model-00015-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0edc9b4c7a77e7841bec6864f5b492cd6dc73630 --- /dev/null +++ b/model-00015-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f204b89d91ef0a896631958d51d6fd66b42267820969acc73cf4cc24f7bea64 +size 4362142872 diff --git a/model-00016-of-00062.safetensors b/model-00016-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..942f22e52a2975ee3f502327280e5191de9d3cb8 --- /dev/null +++ b/model-00016-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:adf0bd4c6e20f687c2f6ca01001e2ceeafc19acdd166602dc49392f6b6bfb0a9 +size 4966188880 diff --git a/model-00017-of-00062.safetensors b/model-00017-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..52f59cb12518abb53414b46413c20cb160b6457b --- /dev/null +++ b/model-00017-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f494a053e92bcdb5cc6988e344b8c17a986cfc8152b09e37feeb5e2c816a84f +size 4362142872 diff --git a/model-00018-of-00062.safetensors b/model-00018-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1486bbfc320893c4dc9bab9c4af153ae2d35a7a --- /dev/null +++ b/model-00018-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab754f1ebd42dc581234b3968a6aeaa995383e928c279cfe651c39fc35181f1f +size 4362142872 diff --git a/model-00019-of-00062.safetensors b/model-00019-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..77f7d08730b1b905d93fe40331892bafd3bab8b6 --- /dev/null +++ b/model-00019-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4106d6f6ff39029fcd63b4e665f5568e7e320f0cdc09bd864b86e463a033e89c +size 4966188880 diff --git a/model-00020-of-00062.safetensors b/model-00020-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5daf0eac10296ff349129eee26be434ce7461ac0 --- /dev/null +++ b/model-00020-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:854027fb1dad68727a270e944bbbbabe53b16322b9bcff67428211320525b50a +size 4362142872 diff --git a/model-00021-of-00062.safetensors b/model-00021-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b73fd8e550d3133f685cd558faf8e1b7b334c171 --- /dev/null +++ b/model-00021-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7f2f6b2e2bcea638de051540ae28e4c840a454f82e954ac632f42d7ec92faa0 +size 4362142872 diff --git a/model-00022-of-00062.safetensors b/model-00022-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..eadcae82b4ab36963d13bb06193b849c6eec7caf --- /dev/null +++ b/model-00022-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab779a8ffec077a76553fd7ee8e11d5e69ff1c50922953be67408f1684a471e7 +size 4966188880 diff --git a/model-00023-of-00062.safetensors b/model-00023-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4e989d61b69e6ebe229d3cb44ab2ed9abc2aa098 --- /dev/null +++ b/model-00023-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2f3898b099c0afd33bb05fdb750ffd2df5355713500c3dae9e456f6965d1547 +size 4362142872 diff --git a/model-00024-of-00062.safetensors b/model-00024-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8ebd27f460262fd4e52532bcacea6303ce0f2d53 --- /dev/null +++ b/model-00024-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:13f53489068ef46c6bcc898187268fdafe6187e65802a9ec86b52a15c792ecf7 +size 4362142872 diff --git a/model-00025-of-00062.safetensors b/model-00025-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c554052dec90d46b133bbdd90386fc3ccdf9b7e5 --- /dev/null +++ b/model-00025-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51fd7054411c0597f052949d58516c65e25fc1e5297e056b64fe9169f54cb422 +size 4966188880 diff --git a/model-00026-of-00062.safetensors b/model-00026-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9d08e6abe3191c526ca6f69d6e7990c135df9cd3 --- /dev/null +++ b/model-00026-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:171874464cd8d46b7011f8f66f1407195464ac7bdce5d6d8e478c00a9f3b9487 +size 4362142872 diff --git a/model-00027-of-00062.safetensors b/model-00027-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d987071e2331d74607887c26c41341e14b3f4782 --- /dev/null +++ b/model-00027-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9f1351be672e66d80c519b18d113d8af7289368bd4de039c294aa4e71629977 +size 4362142872 diff --git a/model-00028-of-00062.safetensors b/model-00028-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a808b3a5df5d5b241b2ad7ab1ddd4fc25604a2dd --- /dev/null +++ b/model-00028-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08eed51ab36d4b6cd0aedf1ad57afae8fc018a13422ac7ddd3888ccb456d3347 +size 4966188880 diff --git a/model-00029-of-00062.safetensors b/model-00029-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..036113feef910e6cd17fa3faf2d1a5303273b234 --- /dev/null +++ b/model-00029-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84d949c03b6bc2ae028ee5822738c96c20a9da91a36ff8462e0445f7a544cc72 +size 4362142872 diff --git a/model-00030-of-00062.safetensors b/model-00030-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..953881d3333aa13abacc7d0fb1be55bf4de20292 --- /dev/null +++ b/model-00030-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4221f6d3e9c2ea1635132282e3fbd68419109a7db947a03b095afb46c6a7de24 +size 4362142872 diff --git a/model-00031-of-00062.safetensors b/model-00031-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ce21d80f9a8cb1ca72a89947ab7378e24b7b2aa9 --- /dev/null +++ b/model-00031-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8061a83515a29432e32bfea57c68beccdb572a335a12d012ff5129e86b771c37 +size 4966188880 diff --git a/model-00032-of-00062.safetensors b/model-00032-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b35a5f510393ce288565d333743786b8ae8d5da0 --- /dev/null +++ b/model-00032-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f942864d3aebafd46c52994ac0619cb10effbbbb0076ce2577a536dc2cce66e +size 4362142872 diff --git a/model-00033-of-00062.safetensors b/model-00033-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aa863fb2634254efd2c633b052c87d9c7c9ae937 --- /dev/null +++ b/model-00033-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96404c3ce813692cd5c8718b8823d2f80de2268d230a65fbcdcdff46f1728282 +size 4362142872 diff --git a/model-00034-of-00062.safetensors b/model-00034-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f96c5543fb1f2e43c3850b540f19ffccecd8229f --- /dev/null +++ b/model-00034-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc11de6ad1d2777bb84149348cd19989d5e58d703af88c93912e2db41f2e02e8 +size 4966188880 diff --git a/model-00035-of-00062.safetensors b/model-00035-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bf645b0940c4dde4a236f0c3f4ffae050963e785 --- /dev/null +++ b/model-00035-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6bffe302aa0e2e6e3a64d2bc2fa16e98fa6bed1e27aa3a390e92284e760b0e69 +size 4362142872 diff --git a/model-00036-of-00062.safetensors b/model-00036-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e339562170e5d02fe960daeaf2fe6226f379bda7 --- /dev/null +++ b/model-00036-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77d18a43372ae13d46708eb7627f4b43022ebe349ea2ac7afe311e15caee5b9f +size 4362142872 diff --git a/model-00037-of-00062.safetensors b/model-00037-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c34e92cac6c22ebe62de1ad44040df4c200c6972 --- /dev/null +++ b/model-00037-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84891d100c85594dbd753c579c75861e5cb3845926c982ca0d8596f05dc4f64c +size 4966188880 diff --git a/model-00038-of-00062.safetensors b/model-00038-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..850bd8e34ce54a3626db85ad1d2be0bd3b8ce8b3 --- /dev/null +++ b/model-00038-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b8e9fbd12bfdf71e761778c34427b4c9dce8b167bbaeaebe4efdf60d183cbf6b +size 4362142872 diff --git a/model-00039-of-00062.safetensors b/model-00039-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3930f6de776c74f0b27ad5aac3d1b7cda95d7637 --- /dev/null +++ b/model-00039-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca70b0514836b80e96782c7ba3d5cf91cd4581a4fa44bf6ca719ee88cd7783d5 +size 4362142872 diff --git a/model-00040-of-00062.safetensors b/model-00040-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7cd16b85ce70982cdff2c2a0d6000c6ce2c9703 --- /dev/null +++ b/model-00040-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7feaa6ca03cea443f8adcfa7a55204b7eb758069eef48496ef462197d63c3c8 +size 4966188880 diff --git a/model-00041-of-00062.safetensors b/model-00041-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e138aa2fa711de7ecc901b51f069d63e55a0101f --- /dev/null +++ b/model-00041-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59fa5015f6c3f12b30c4d71958bfa17d459083598b2fa3be168a640b8825d543 +size 4362142872 diff --git a/model-00042-of-00062.safetensors b/model-00042-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5aeec8b7d60daf86eb31c27fe6afeed82f770ee7 --- /dev/null +++ b/model-00042-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3f1651fe1c55db7750da5a86b486c1b70032d3bd4275e02515b159b3169d7c35 +size 4362142872 diff --git a/model-00043-of-00062.safetensors b/model-00043-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..84759c3170b4664657cba03ffe12ff8808692dd3 --- /dev/null +++ b/model-00043-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:072a925e3eb8d9c11104290b71675efa79d3a10cff9d9e1059875318fa3c4106 +size 4966188880 diff --git a/model-00044-of-00062.safetensors b/model-00044-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d16b62bb7a6f37eac8dde975e8e10f100e6d3b89 --- /dev/null +++ b/model-00044-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:162776b18546033690335480fba6e45bfd2929ce6b7254edfda2ccb0614056a3 +size 4362142872 diff --git a/model-00045-of-00062.safetensors b/model-00045-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4183d88a2c0be5070155da8dd5469b2f92e2390f --- /dev/null +++ b/model-00045-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d75358e792f8f99a0d812b3ed9e8ba221aabf60074bb85d6bc7dcb3ba918c143 +size 4362142872 diff --git a/model-00046-of-00062.safetensors b/model-00046-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..af51b9decb27011dcb35f9daa48cfcbb343a02ab --- /dev/null +++ b/model-00046-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a21619836cfe1f7da2f8c0ab15e34629fe2a916185b4f68274be43f83a672a58 +size 4966188880 diff --git a/model-00047-of-00062.safetensors b/model-00047-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9ceef05dd7e85d2fe4dd3999c893578943eccee4 --- /dev/null +++ b/model-00047-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da830164b62599c8abeee5d2385aa199265cf18bad62a47f049877c23e97ddb0 +size 4362142872 diff --git a/model-00048-of-00062.safetensors b/model-00048-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9a285e9be05dbd87fe812ba70e1d7c2807b79b06 --- /dev/null +++ b/model-00048-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b178a632c2743edf6a46a5a26f2d8c77d885cb35a092b0fdcffa07ed57a1e4a +size 4362142872 diff --git a/model-00049-of-00062.safetensors b/model-00049-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d8adc139a7c9fde66a7ab7e4556a69cea3cd283 --- /dev/null +++ b/model-00049-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a3632ea6fea1151f76be9284dfdcf3fd683f48332d2b47ef7a4fad41b1ce0f3 +size 4966188880 diff --git a/model-00050-of-00062.safetensors b/model-00050-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3b37ffa55aef4fb09dd92db1d93e56dcdfa5d574 --- /dev/null +++ b/model-00050-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d2ec43144fe3748855b95079739fc07a1b18a5791c7074dab47aaeb7bcb0b1c3 +size 4362142872 diff --git a/model-00051-of-00062.safetensors b/model-00051-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ec408c0a22e7aaa0e34ed6ed6c6d36514ee5e38e --- /dev/null +++ b/model-00051-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:09a6f5f4e91e8f9142d27e10d50acdb61fb9e588770624e944dde704ca2c6028 +size 4362142872 diff --git a/model-00052-of-00062.safetensors b/model-00052-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d190cd1b697f7fbca69ff9abc92908b8ae1336d0 --- /dev/null +++ b/model-00052-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6584de20637ade0a4e8dd827faa35268174e0538175424fdba3fb7683697127e +size 4966188880 diff --git a/model-00053-of-00062.safetensors b/model-00053-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..999774c9e1a880298daf2b4725120b698191d636 --- /dev/null +++ b/model-00053-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25c0983d3f4d2399920567ada807d7fbd565b7c70ab3a6e9d85271d25f288736 +size 4362142872 diff --git a/model-00054-of-00062.safetensors b/model-00054-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e70441f8db56c93fc16bf78007c3589c7272f5c7 --- /dev/null +++ b/model-00054-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f97cf948fe44b099b9f880637461493a77103e07aaeed855873490ef3b3989d6 +size 4362142872 diff --git a/model-00055-of-00062.safetensors b/model-00055-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1a27609cdc2f90acf70e8aeeedce4edb3368d86d --- /dev/null +++ b/model-00055-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca64159b88bbdb0d81d4563fd6fc616e6a010ff02346b14c6394071e44b6adaa +size 4966188880 diff --git a/model-00056-of-00062.safetensors b/model-00056-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..39bc9d46d30b596c03bc45a4a7eca7794bf3f3b4 --- /dev/null +++ b/model-00056-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc1e0191de03db6240d77aec3cc1a532bf5d8d384af4634dd8f26691647b9263 +size 4362142872 diff --git a/model-00057-of-00062.safetensors b/model-00057-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2b536b78ad8fd6f20ece91f75f6aa65bb92c163d --- /dev/null +++ b/model-00057-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:194d2e55173a8a007aa34ae85ec5e70697915732475cfe9f7a84a8d87752b6aa +size 4362142872 diff --git a/model-00058-of-00062.safetensors b/model-00058-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bacbcca5f0664b8942712749335d6f370ca773d3 --- /dev/null +++ b/model-00058-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df96b201bda5c0fb84f6d0957e744fa531f4d3206d75990208cae3625f1882c6 +size 4966188880 diff --git a/model-00059-of-00062.safetensors b/model-00059-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b7d18fc28a546e3012d9f3d89d854aedd3acd74 --- /dev/null +++ b/model-00059-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f2a38d69376c73abf9451fe11d55958cd4ebe9257979beda2d89d47804955a1 +size 4362142872 diff --git a/model-00060-of-00062.safetensors b/model-00060-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b977eb68a5be27368ed3f2a1dd27627b050cf150 --- /dev/null +++ b/model-00060-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:adddf6307c7fc3f4e1a868a4bd40cdde89b4299b5502e6f5b4cc7261dbff918d +size 4362142872 diff --git a/model-00061-of-00062.safetensors b/model-00061-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b6a8e87838e180b02980e44888e58aa90cabe3a0 --- /dev/null +++ b/model-00061-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c08def3e69ea072188273502cb8a65338322b0878cfadb107be709302310bc7 +size 4362241496 diff --git a/model-00062-of-00062.safetensors b/model-00062-of-00062.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cd8fcc0660e8f6a59738fcdfb637410ed3ce6519 --- /dev/null +++ b/model-00062-of-00062.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ca4851489d24bbdfbacdfb0b6f0e4090c1c324c3b41e5f7cc3f252998afcae3 +size 4202692736 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ef0c642b145f0661c97c573578ea5f5c868b5c19 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,731 @@ +{ + "metadata": { + "total_parameters": 70553706496, + "total_size": 282214825984 + }, + "weight_map": { + "lm_head.weight": "model-00062-of-00062.safetensors", + "model.embed_tokens.weight": "model-00001-of-00062.safetensors", + "model.layers.0.input_layernorm.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00002-of-00062.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00062.safetensors", + "model.layers.1.input_layernorm.weight": "model-00003-of-00062.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00003-of-00062.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00062.safetensors", + "model.layers.10.input_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.11.input_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.input_layernorm.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00011-of-00062.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00010-of-00062.safetensors", + "model.layers.13.input_layernorm.weight": "model-00012-of-00062.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00012-of-00062.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00011-of-00062.safetensors", + "model.layers.14.input_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00012-of-00062.safetensors", + "model.layers.15.input_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.input_layernorm.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00014-of-00062.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00013-of-00062.safetensors", + "model.layers.17.input_layernorm.weight": "model-00015-of-00062.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00015-of-00062.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00014-of-00062.safetensors", + "model.layers.18.input_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00015-of-00062.safetensors", + "model.layers.19.input_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.2.input_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00003-of-00062.safetensors", + "model.layers.20.input_layernorm.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00017-of-00062.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00016-of-00062.safetensors", + "model.layers.21.input_layernorm.weight": "model-00018-of-00062.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00018-of-00062.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00017-of-00062.safetensors", + "model.layers.22.input_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00018-of-00062.safetensors", + "model.layers.23.input_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.input_layernorm.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00020-of-00062.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00019-of-00062.safetensors", + "model.layers.25.input_layernorm.weight": "model-00021-of-00062.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00021-of-00062.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00020-of-00062.safetensors", + "model.layers.26.input_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00021-of-00062.safetensors", + "model.layers.27.input_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.input_layernorm.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.gate_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.mlp.up_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00023-of-00062.safetensors", + "model.layers.28.self_attn.k_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.q_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.28.self_attn.v_proj.weight": "model-00022-of-00062.safetensors", + "model.layers.29.input_layernorm.weight": "model-00024-of-00062.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.29.mlp.gate_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.mlp.up_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00024-of-00062.safetensors", + "model.layers.29.self_attn.k_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.q_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.29.self_attn.v_proj.weight": "model-00023-of-00062.safetensors", + "model.layers.3.input_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.30.input_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.30.mlp.gate_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.mlp.up_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.30.self_attn.k_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.q_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.30.self_attn.v_proj.weight": "model-00024-of-00062.safetensors", + "model.layers.31.input_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.gate_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.mlp.up_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.k_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.q_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.31.self_attn.v_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.input_layernorm.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.down_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.gate_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.mlp.up_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.32.post_attention_layernorm.weight": "model-00026-of-00062.safetensors", + "model.layers.32.self_attn.k_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.o_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.q_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.32.self_attn.v_proj.weight": "model-00025-of-00062.safetensors", + "model.layers.33.input_layernorm.weight": "model-00027-of-00062.safetensors", + "model.layers.33.mlp.down_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.33.mlp.gate_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.mlp.up_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.33.post_attention_layernorm.weight": "model-00027-of-00062.safetensors", + "model.layers.33.self_attn.k_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.o_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.q_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.33.self_attn.v_proj.weight": "model-00026-of-00062.safetensors", + "model.layers.34.input_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.34.mlp.down_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.34.mlp.gate_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.mlp.up_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.post_attention_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.34.self_attn.k_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.o_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.q_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.34.self_attn.v_proj.weight": "model-00027-of-00062.safetensors", + "model.layers.35.input_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.down_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.gate_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.mlp.up_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.post_attention_layernorm.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.k_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.o_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.q_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.35.self_attn.v_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.input_layernorm.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.down_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.gate_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.mlp.up_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.36.post_attention_layernorm.weight": "model-00029-of-00062.safetensors", + "model.layers.36.self_attn.k_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.o_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.q_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.36.self_attn.v_proj.weight": "model-00028-of-00062.safetensors", + "model.layers.37.input_layernorm.weight": "model-00030-of-00062.safetensors", + "model.layers.37.mlp.down_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.37.mlp.gate_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.mlp.up_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.37.post_attention_layernorm.weight": "model-00030-of-00062.safetensors", + "model.layers.37.self_attn.k_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.o_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.q_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.37.self_attn.v_proj.weight": "model-00029-of-00062.safetensors", + "model.layers.38.input_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.38.mlp.down_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.38.mlp.gate_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.mlp.up_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.post_attention_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.38.self_attn.k_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.o_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.q_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.38.self_attn.v_proj.weight": "model-00030-of-00062.safetensors", + "model.layers.39.input_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.down_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.gate_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.mlp.up_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.post_attention_layernorm.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.k_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.o_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.q_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.39.self_attn.v_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.4.input_layernorm.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00005-of-00062.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00004-of-00062.safetensors", + "model.layers.40.input_layernorm.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.down_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.gate_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.mlp.up_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.40.post_attention_layernorm.weight": "model-00032-of-00062.safetensors", + "model.layers.40.self_attn.k_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.o_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.q_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.40.self_attn.v_proj.weight": "model-00031-of-00062.safetensors", + "model.layers.41.input_layernorm.weight": "model-00033-of-00062.safetensors", + "model.layers.41.mlp.down_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.41.mlp.gate_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.mlp.up_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.41.post_attention_layernorm.weight": "model-00033-of-00062.safetensors", + "model.layers.41.self_attn.k_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.o_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.q_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.41.self_attn.v_proj.weight": "model-00032-of-00062.safetensors", + "model.layers.42.input_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.42.mlp.down_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.42.mlp.gate_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.mlp.up_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.post_attention_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.42.self_attn.k_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.o_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.q_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.42.self_attn.v_proj.weight": "model-00033-of-00062.safetensors", + "model.layers.43.input_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.down_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.gate_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.mlp.up_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.post_attention_layernorm.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.k_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.o_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.q_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.43.self_attn.v_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.input_layernorm.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.down_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.gate_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.mlp.up_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.44.post_attention_layernorm.weight": "model-00035-of-00062.safetensors", + "model.layers.44.self_attn.k_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.o_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.q_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.44.self_attn.v_proj.weight": "model-00034-of-00062.safetensors", + "model.layers.45.input_layernorm.weight": "model-00036-of-00062.safetensors", + "model.layers.45.mlp.down_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.45.mlp.gate_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.mlp.up_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.45.post_attention_layernorm.weight": "model-00036-of-00062.safetensors", + "model.layers.45.self_attn.k_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.o_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.q_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.45.self_attn.v_proj.weight": "model-00035-of-00062.safetensors", + "model.layers.46.input_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.46.mlp.down_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.46.mlp.gate_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.mlp.up_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.post_attention_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.46.self_attn.k_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.o_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.q_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.46.self_attn.v_proj.weight": "model-00036-of-00062.safetensors", + "model.layers.47.input_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.down_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.gate_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.mlp.up_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.post_attention_layernorm.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.k_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.o_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.q_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.47.self_attn.v_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.input_layernorm.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.down_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.gate_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.mlp.up_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.48.post_attention_layernorm.weight": "model-00038-of-00062.safetensors", + "model.layers.48.self_attn.k_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.o_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.q_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.48.self_attn.v_proj.weight": "model-00037-of-00062.safetensors", + "model.layers.49.input_layernorm.weight": "model-00039-of-00062.safetensors", + "model.layers.49.mlp.down_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.49.mlp.gate_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.mlp.up_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.49.post_attention_layernorm.weight": "model-00039-of-00062.safetensors", + "model.layers.49.self_attn.k_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.o_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.q_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.49.self_attn.v_proj.weight": "model-00038-of-00062.safetensors", + "model.layers.5.input_layernorm.weight": "model-00006-of-00062.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00006-of-00062.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00005-of-00062.safetensors", + "model.layers.50.input_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.50.mlp.down_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.50.mlp.gate_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.mlp.up_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.post_attention_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.50.self_attn.k_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.o_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.q_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.50.self_attn.v_proj.weight": "model-00039-of-00062.safetensors", + "model.layers.51.input_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.down_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.gate_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.mlp.up_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.post_attention_layernorm.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.k_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.o_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.q_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.51.self_attn.v_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.input_layernorm.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.down_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.gate_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.mlp.up_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.52.post_attention_layernorm.weight": "model-00041-of-00062.safetensors", + "model.layers.52.self_attn.k_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.o_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.q_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.52.self_attn.v_proj.weight": "model-00040-of-00062.safetensors", + "model.layers.53.input_layernorm.weight": "model-00042-of-00062.safetensors", + "model.layers.53.mlp.down_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.53.mlp.gate_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.mlp.up_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.53.post_attention_layernorm.weight": "model-00042-of-00062.safetensors", + "model.layers.53.self_attn.k_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.o_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.q_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.53.self_attn.v_proj.weight": "model-00041-of-00062.safetensors", + "model.layers.54.input_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.54.mlp.down_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.54.mlp.gate_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.mlp.up_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.post_attention_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.54.self_attn.k_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.o_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.q_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.54.self_attn.v_proj.weight": "model-00042-of-00062.safetensors", + "model.layers.55.input_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.down_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.gate_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.mlp.up_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.post_attention_layernorm.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.k_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.o_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.q_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.55.self_attn.v_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.input_layernorm.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.down_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.gate_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.mlp.up_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.56.post_attention_layernorm.weight": "model-00044-of-00062.safetensors", + "model.layers.56.self_attn.k_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.o_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.q_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.56.self_attn.v_proj.weight": "model-00043-of-00062.safetensors", + "model.layers.57.input_layernorm.weight": "model-00045-of-00062.safetensors", + "model.layers.57.mlp.down_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.57.mlp.gate_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.mlp.up_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.57.post_attention_layernorm.weight": "model-00045-of-00062.safetensors", + "model.layers.57.self_attn.k_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.o_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.q_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.57.self_attn.v_proj.weight": "model-00044-of-00062.safetensors", + "model.layers.58.input_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.58.mlp.down_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.58.mlp.gate_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.mlp.up_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.post_attention_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.58.self_attn.k_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.o_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.q_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.58.self_attn.v_proj.weight": "model-00045-of-00062.safetensors", + "model.layers.59.input_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.down_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.gate_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.mlp.up_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.post_attention_layernorm.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.k_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.o_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.q_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.59.self_attn.v_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.6.input_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00006-of-00062.safetensors", + "model.layers.60.input_layernorm.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.down_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.gate_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.mlp.up_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.60.post_attention_layernorm.weight": "model-00047-of-00062.safetensors", + "model.layers.60.self_attn.k_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.o_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.q_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.60.self_attn.v_proj.weight": "model-00046-of-00062.safetensors", + "model.layers.61.input_layernorm.weight": "model-00048-of-00062.safetensors", + "model.layers.61.mlp.down_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.61.mlp.gate_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.mlp.up_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.61.post_attention_layernorm.weight": "model-00048-of-00062.safetensors", + "model.layers.61.self_attn.k_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.o_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.q_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.61.self_attn.v_proj.weight": "model-00047-of-00062.safetensors", + "model.layers.62.input_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.62.mlp.down_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.62.mlp.gate_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.mlp.up_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.post_attention_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.62.self_attn.k_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.o_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.q_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.62.self_attn.v_proj.weight": "model-00048-of-00062.safetensors", + "model.layers.63.input_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.down_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.gate_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.mlp.up_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.post_attention_layernorm.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.k_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.o_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.q_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.63.self_attn.v_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.input_layernorm.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.down_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.gate_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.mlp.up_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.64.post_attention_layernorm.weight": "model-00050-of-00062.safetensors", + "model.layers.64.self_attn.k_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.o_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.q_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.64.self_attn.v_proj.weight": "model-00049-of-00062.safetensors", + "model.layers.65.input_layernorm.weight": "model-00051-of-00062.safetensors", + "model.layers.65.mlp.down_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.65.mlp.gate_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.mlp.up_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.65.post_attention_layernorm.weight": "model-00051-of-00062.safetensors", + "model.layers.65.self_attn.k_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.o_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.q_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.65.self_attn.v_proj.weight": "model-00050-of-00062.safetensors", + "model.layers.66.input_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.66.mlp.down_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.66.mlp.gate_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.mlp.up_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.post_attention_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.66.self_attn.k_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.o_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.q_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.66.self_attn.v_proj.weight": "model-00051-of-00062.safetensors", + "model.layers.67.input_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.down_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.gate_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.mlp.up_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.post_attention_layernorm.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.k_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.o_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.q_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.67.self_attn.v_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.input_layernorm.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.down_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.gate_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.mlp.up_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.68.post_attention_layernorm.weight": "model-00053-of-00062.safetensors", + "model.layers.68.self_attn.k_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.o_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.q_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.68.self_attn.v_proj.weight": "model-00052-of-00062.safetensors", + "model.layers.69.input_layernorm.weight": "model-00054-of-00062.safetensors", + "model.layers.69.mlp.down_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.69.mlp.gate_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.mlp.up_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.69.post_attention_layernorm.weight": "model-00054-of-00062.safetensors", + "model.layers.69.self_attn.k_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.o_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.q_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.69.self_attn.v_proj.weight": "model-00053-of-00062.safetensors", + "model.layers.7.input_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.70.input_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.70.mlp.down_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.70.mlp.gate_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.mlp.up_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.post_attention_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.70.self_attn.k_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.o_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.q_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.70.self_attn.v_proj.weight": "model-00054-of-00062.safetensors", + "model.layers.71.input_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.down_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.gate_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.mlp.up_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.post_attention_layernorm.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.k_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.o_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.q_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.71.self_attn.v_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.input_layernorm.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.down_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.gate_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.mlp.up_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.72.post_attention_layernorm.weight": "model-00056-of-00062.safetensors", + "model.layers.72.self_attn.k_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.o_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.q_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.72.self_attn.v_proj.weight": "model-00055-of-00062.safetensors", + "model.layers.73.input_layernorm.weight": "model-00057-of-00062.safetensors", + "model.layers.73.mlp.down_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.73.mlp.gate_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.mlp.up_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.73.post_attention_layernorm.weight": "model-00057-of-00062.safetensors", + "model.layers.73.self_attn.k_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.o_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.q_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.73.self_attn.v_proj.weight": "model-00056-of-00062.safetensors", + "model.layers.74.input_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.74.mlp.down_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.74.mlp.gate_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.mlp.up_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.post_attention_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.74.self_attn.k_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.o_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.q_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.74.self_attn.v_proj.weight": "model-00057-of-00062.safetensors", + "model.layers.75.input_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.down_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.gate_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.mlp.up_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.post_attention_layernorm.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.k_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.o_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.q_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.75.self_attn.v_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.input_layernorm.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.down_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.gate_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.mlp.up_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.76.post_attention_layernorm.weight": "model-00059-of-00062.safetensors", + "model.layers.76.self_attn.k_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.o_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.q_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.76.self_attn.v_proj.weight": "model-00058-of-00062.safetensors", + "model.layers.77.input_layernorm.weight": "model-00060-of-00062.safetensors", + "model.layers.77.mlp.down_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.77.mlp.gate_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.mlp.up_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.77.post_attention_layernorm.weight": "model-00060-of-00062.safetensors", + "model.layers.77.self_attn.k_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.o_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.q_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.77.self_attn.v_proj.weight": "model-00059-of-00062.safetensors", + "model.layers.78.input_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.78.mlp.down_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.78.mlp.gate_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.mlp.up_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.post_attention_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.78.self_attn.k_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.o_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.q_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.78.self_attn.v_proj.weight": "model-00060-of-00062.safetensors", + "model.layers.79.input_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.down_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.gate_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.mlp.up_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.post_attention_layernorm.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.k_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.o_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.q_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.79.self_attn.v_proj.weight": "model-00061-of-00062.safetensors", + "model.layers.8.input_layernorm.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00008-of-00062.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00007-of-00062.safetensors", + "model.layers.9.input_layernorm.weight": "model-00009-of-00062.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00009-of-00062.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00009-of-00062.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00008-of-00062.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00008-of-00062.safetensors", + "model.norm.weight": "model-00061-of-00062.safetensors" + } +} diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127 --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,17 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "<|eot_id|>" +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2e93688d195888fad0d99726a01b934cd4a3766 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,2064 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|eot_id|>", + "extra_special_tokens": {}, + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "<|eot_id|>", + "padding_side": "right", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d12b0606375b89695a5901882cb590d29064e595 --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,84495 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 0, + "global_step": 12065, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 8.288437629506838e-05, + "grad_norm": 2.0744729042053223, + "learning_rate": 5e-06, + "loss": 1.5463, + "step": 1 + }, + { + "epoch": 0.00016576875259013675, + "grad_norm": 1.277697205543518, + "learning_rate": 4.999585578118525e-06, + "loss": 1.4781, + "step": 2 + }, + { + "epoch": 0.0002486531288852051, + "grad_norm": 0.9632207751274109, + "learning_rate": 4.99917115623705e-06, + "loss": 1.3757, + "step": 3 + }, + { + "epoch": 0.0003315375051802735, + "grad_norm": 0.7597147226333618, + "learning_rate": 4.9987567343555746e-06, + "loss": 1.2965, + "step": 4 + }, + { + "epoch": 0.0004144218814753419, + "grad_norm": 0.911206066608429, + "learning_rate": 4.998342312474099e-06, + "loss": 1.1403, + "step": 5 + }, + { + "epoch": 0.0004973062577704103, + "grad_norm": 0.8961520195007324, + "learning_rate": 4.997927890592623e-06, + "loss": 1.0603, + "step": 6 + }, + { + "epoch": 0.0005801906340654787, + "grad_norm": 0.6187644004821777, + "learning_rate": 4.997513468711148e-06, + "loss": 0.9973, + "step": 7 + }, + { + "epoch": 0.000663075010360547, + "grad_norm": 0.5391705632209778, + "learning_rate": 4.997099046829673e-06, + "loss": 0.9081, + "step": 8 + }, + { + "epoch": 0.0007459593866556154, + "grad_norm": 0.5124619007110596, + "learning_rate": 4.9966846249481976e-06, + "loss": 0.8682, + "step": 9 + }, + { + "epoch": 0.0008288437629506838, + "grad_norm": 0.48603078722953796, + "learning_rate": 4.996270203066722e-06, + "loss": 0.9269, + "step": 10 + }, + { + "epoch": 0.0009117281392457522, + "grad_norm": 0.4422253370285034, + "learning_rate": 4.995855781185247e-06, + "loss": 0.9545, + "step": 11 + }, + { + "epoch": 0.0009946125155408205, + "grad_norm": 0.39657020568847656, + "learning_rate": 4.995441359303772e-06, + "loss": 0.8993, + "step": 12 + }, + { + "epoch": 0.0010774968918358888, + "grad_norm": 0.3969251811504364, + "learning_rate": 4.995026937422296e-06, + "loss": 0.8025, + "step": 13 + }, + { + "epoch": 0.0011603812681309574, + "grad_norm": 0.41122302412986755, + "learning_rate": 4.9946125155408206e-06, + "loss": 0.8316, + "step": 14 + }, + { + "epoch": 0.0012432656444260257, + "grad_norm": 0.3637348413467407, + "learning_rate": 4.994198093659345e-06, + "loss": 0.8307, + "step": 15 + }, + { + "epoch": 0.001326150020721094, + "grad_norm": 0.3554855287075043, + "learning_rate": 4.99378367177787e-06, + "loss": 0.7626, + "step": 16 + }, + { + "epoch": 0.0014090343970161625, + "grad_norm": 0.37707701325416565, + "learning_rate": 4.993369249896395e-06, + "loss": 0.8118, + "step": 17 + }, + { + "epoch": 0.0014919187733112309, + "grad_norm": 0.3464816212654114, + "learning_rate": 4.99295482801492e-06, + "loss": 0.7769, + "step": 18 + }, + { + "epoch": 0.0015748031496062992, + "grad_norm": 0.3458833396434784, + "learning_rate": 4.992540406133444e-06, + "loss": 0.7891, + "step": 19 + }, + { + "epoch": 0.0016576875259013675, + "grad_norm": 0.36892223358154297, + "learning_rate": 4.992125984251969e-06, + "loss": 0.8303, + "step": 20 + }, + { + "epoch": 0.001740571902196436, + "grad_norm": 0.3641088008880615, + "learning_rate": 4.991711562370493e-06, + "loss": 0.7155, + "step": 21 + }, + { + "epoch": 0.0018234562784915044, + "grad_norm": 0.35344964265823364, + "learning_rate": 4.9912971404890186e-06, + "loss": 0.7744, + "step": 22 + }, + { + "epoch": 0.0019063406547865727, + "grad_norm": 0.35258200764656067, + "learning_rate": 4.990882718607543e-06, + "loss": 0.7392, + "step": 23 + }, + { + "epoch": 0.001989225031081641, + "grad_norm": 0.3408772945404053, + "learning_rate": 4.990468296726067e-06, + "loss": 0.7289, + "step": 24 + }, + { + "epoch": 0.0020721094073767096, + "grad_norm": 0.3490234613418579, + "learning_rate": 4.990053874844592e-06, + "loss": 0.7626, + "step": 25 + }, + { + "epoch": 0.0021549937836717777, + "grad_norm": 0.3482590913772583, + "learning_rate": 4.989639452963117e-06, + "loss": 0.7704, + "step": 26 + }, + { + "epoch": 0.002237878159966846, + "grad_norm": 0.3744784891605377, + "learning_rate": 4.9892250310816416e-06, + "loss": 0.728, + "step": 27 + }, + { + "epoch": 0.0023207625362619147, + "grad_norm": 0.3234175741672516, + "learning_rate": 4.988810609200166e-06, + "loss": 0.7274, + "step": 28 + }, + { + "epoch": 0.002403646912556983, + "grad_norm": 0.33797064423561096, + "learning_rate": 4.988396187318691e-06, + "loss": 0.7568, + "step": 29 + }, + { + "epoch": 0.0024865312888520514, + "grad_norm": 0.3711134195327759, + "learning_rate": 4.987981765437216e-06, + "loss": 0.7784, + "step": 30 + }, + { + "epoch": 0.00256941566514712, + "grad_norm": 0.3144945800304413, + "learning_rate": 4.98756734355574e-06, + "loss": 0.7263, + "step": 31 + }, + { + "epoch": 0.002652300041442188, + "grad_norm": 0.32774773240089417, + "learning_rate": 4.9871529216742646e-06, + "loss": 0.7612, + "step": 32 + }, + { + "epoch": 0.0027351844177372566, + "grad_norm": 0.34702154994010925, + "learning_rate": 4.986738499792789e-06, + "loss": 0.7476, + "step": 33 + }, + { + "epoch": 0.002818068794032325, + "grad_norm": 0.4608129858970642, + "learning_rate": 4.986324077911314e-06, + "loss": 0.7594, + "step": 34 + }, + { + "epoch": 0.002900953170327393, + "grad_norm": 0.31281527876853943, + "learning_rate": 4.985909656029839e-06, + "loss": 0.6876, + "step": 35 + }, + { + "epoch": 0.0029838375466224617, + "grad_norm": 0.37600427865982056, + "learning_rate": 4.985495234148364e-06, + "loss": 0.7314, + "step": 36 + }, + { + "epoch": 0.00306672192291753, + "grad_norm": 0.4084557592868805, + "learning_rate": 4.9850808122668884e-06, + "loss": 0.7612, + "step": 37 + }, + { + "epoch": 0.0031496062992125984, + "grad_norm": 0.3452323377132416, + "learning_rate": 4.984666390385413e-06, + "loss": 0.744, + "step": 38 + }, + { + "epoch": 0.003232490675507667, + "grad_norm": 0.3402542769908905, + "learning_rate": 4.984251968503937e-06, + "loss": 0.8005, + "step": 39 + }, + { + "epoch": 0.003315375051802735, + "grad_norm": 0.31692641973495483, + "learning_rate": 4.983837546622462e-06, + "loss": 0.6907, + "step": 40 + }, + { + "epoch": 0.0033982594280978036, + "grad_norm": 0.3068479895591736, + "learning_rate": 4.983423124740987e-06, + "loss": 0.696, + "step": 41 + }, + { + "epoch": 0.003481143804392872, + "grad_norm": 0.3217547535896301, + "learning_rate": 4.9830087028595114e-06, + "loss": 0.6857, + "step": 42 + }, + { + "epoch": 0.00356402818068794, + "grad_norm": 0.36482271552085876, + "learning_rate": 4.982594280978036e-06, + "loss": 0.7284, + "step": 43 + }, + { + "epoch": 0.0036469125569830087, + "grad_norm": 0.31914111971855164, + "learning_rate": 4.982179859096561e-06, + "loss": 0.6508, + "step": 44 + }, + { + "epoch": 0.0037297969332780773, + "grad_norm": 0.352004736661911, + "learning_rate": 4.9817654372150856e-06, + "loss": 0.7002, + "step": 45 + }, + { + "epoch": 0.0038126813095731454, + "grad_norm": 0.3274849057197571, + "learning_rate": 4.98135101533361e-06, + "loss": 0.674, + "step": 46 + }, + { + "epoch": 0.003895565685868214, + "grad_norm": 0.34178614616394043, + "learning_rate": 4.9809365934521344e-06, + "loss": 0.7081, + "step": 47 + }, + { + "epoch": 0.003978450062163282, + "grad_norm": 0.3356665074825287, + "learning_rate": 4.980522171570659e-06, + "loss": 0.6993, + "step": 48 + }, + { + "epoch": 0.004061334438458351, + "grad_norm": 0.3454911708831787, + "learning_rate": 4.980107749689184e-06, + "loss": 0.7051, + "step": 49 + }, + { + "epoch": 0.004144218814753419, + "grad_norm": 0.33815452456474304, + "learning_rate": 4.979693327807709e-06, + "loss": 0.7409, + "step": 50 + }, + { + "epoch": 0.004227103191048487, + "grad_norm": 0.33290815353393555, + "learning_rate": 4.979278905926233e-06, + "loss": 0.7129, + "step": 51 + }, + { + "epoch": 0.004309987567343555, + "grad_norm": 0.3212352693080902, + "learning_rate": 4.978864484044758e-06, + "loss": 0.7054, + "step": 52 + }, + { + "epoch": 0.004392871943638624, + "grad_norm": 0.32935085892677307, + "learning_rate": 4.978450062163283e-06, + "loss": 0.6966, + "step": 53 + }, + { + "epoch": 0.004475756319933692, + "grad_norm": 0.3480775058269501, + "learning_rate": 4.978035640281807e-06, + "loss": 0.6652, + "step": 54 + }, + { + "epoch": 0.0045586406962287605, + "grad_norm": 0.34144818782806396, + "learning_rate": 4.977621218400332e-06, + "loss": 0.6946, + "step": 55 + }, + { + "epoch": 0.0046415250725238295, + "grad_norm": 0.31688249111175537, + "learning_rate": 4.977206796518856e-06, + "loss": 0.6594, + "step": 56 + }, + { + "epoch": 0.004724409448818898, + "grad_norm": 0.3453390896320343, + "learning_rate": 4.976792374637381e-06, + "loss": 0.7276, + "step": 57 + }, + { + "epoch": 0.004807293825113966, + "grad_norm": 0.7068150043487549, + "learning_rate": 4.976377952755906e-06, + "loss": 0.6813, + "step": 58 + }, + { + "epoch": 0.004890178201409035, + "grad_norm": 0.3949696719646454, + "learning_rate": 4.975963530874431e-06, + "loss": 0.6736, + "step": 59 + }, + { + "epoch": 0.004973062577704103, + "grad_norm": 0.3235418200492859, + "learning_rate": 4.9755491089929554e-06, + "loss": 0.6695, + "step": 60 + }, + { + "epoch": 0.005055946953999171, + "grad_norm": 0.33208054304122925, + "learning_rate": 4.97513468711148e-06, + "loss": 0.7012, + "step": 61 + }, + { + "epoch": 0.00513883133029424, + "grad_norm": 0.32262441515922546, + "learning_rate": 4.974720265230005e-06, + "loss": 0.7279, + "step": 62 + }, + { + "epoch": 0.005221715706589308, + "grad_norm": 0.33311042189598083, + "learning_rate": 4.97430584334853e-06, + "loss": 0.7117, + "step": 63 + }, + { + "epoch": 0.005304600082884376, + "grad_norm": 0.34586232900619507, + "learning_rate": 4.973891421467054e-06, + "loss": 0.7077, + "step": 64 + }, + { + "epoch": 0.005387484459179445, + "grad_norm": 0.34003862738609314, + "learning_rate": 4.9734769995855784e-06, + "loss": 0.71, + "step": 65 + }, + { + "epoch": 0.005470368835474513, + "grad_norm": 0.32744595408439636, + "learning_rate": 4.973062577704103e-06, + "loss": 0.6838, + "step": 66 + }, + { + "epoch": 0.005553253211769581, + "grad_norm": 0.32802754640579224, + "learning_rate": 4.972648155822628e-06, + "loss": 0.7145, + "step": 67 + }, + { + "epoch": 0.00563613758806465, + "grad_norm": 0.3499254584312439, + "learning_rate": 4.972233733941153e-06, + "loss": 0.6522, + "step": 68 + }, + { + "epoch": 0.005719021964359718, + "grad_norm": 0.3249247670173645, + "learning_rate": 4.971819312059677e-06, + "loss": 0.6946, + "step": 69 + }, + { + "epoch": 0.005801906340654786, + "grad_norm": 0.31513145565986633, + "learning_rate": 4.971404890178202e-06, + "loss": 0.6861, + "step": 70 + }, + { + "epoch": 0.005884790716949855, + "grad_norm": 0.31506261229515076, + "learning_rate": 4.970990468296727e-06, + "loss": 0.638, + "step": 71 + }, + { + "epoch": 0.0059676750932449235, + "grad_norm": 0.3391733169555664, + "learning_rate": 4.970576046415251e-06, + "loss": 0.6895, + "step": 72 + }, + { + "epoch": 0.006050559469539992, + "grad_norm": 0.3290216326713562, + "learning_rate": 4.970161624533776e-06, + "loss": 0.6937, + "step": 73 + }, + { + "epoch": 0.00613344384583506, + "grad_norm": 0.34729424118995667, + "learning_rate": 4.9697472026523e-06, + "loss": 0.6402, + "step": 74 + }, + { + "epoch": 0.006216328222130129, + "grad_norm": 0.3170442283153534, + "learning_rate": 4.969332780770825e-06, + "loss": 0.6727, + "step": 75 + }, + { + "epoch": 0.006299212598425197, + "grad_norm": 0.2981469929218292, + "learning_rate": 4.96891835888935e-06, + "loss": 0.6246, + "step": 76 + }, + { + "epoch": 0.006382096974720265, + "grad_norm": 0.35152125358581543, + "learning_rate": 4.968503937007874e-06, + "loss": 0.7007, + "step": 77 + }, + { + "epoch": 0.006464981351015334, + "grad_norm": 0.3323407471179962, + "learning_rate": 4.9680895151263994e-06, + "loss": 0.6348, + "step": 78 + }, + { + "epoch": 0.006547865727310402, + "grad_norm": 0.315832257270813, + "learning_rate": 4.967675093244924e-06, + "loss": 0.7064, + "step": 79 + }, + { + "epoch": 0.00663075010360547, + "grad_norm": 0.35248610377311707, + "learning_rate": 4.967260671363448e-06, + "loss": 0.7642, + "step": 80 + }, + { + "epoch": 0.006713634479900539, + "grad_norm": 0.34405872225761414, + "learning_rate": 4.966846249481973e-06, + "loss": 0.6707, + "step": 81 + }, + { + "epoch": 0.006796518856195607, + "grad_norm": 0.30386534333229065, + "learning_rate": 4.966431827600497e-06, + "loss": 0.6282, + "step": 82 + }, + { + "epoch": 0.006879403232490675, + "grad_norm": 0.3659181594848633, + "learning_rate": 4.9660174057190225e-06, + "loss": 0.6846, + "step": 83 + }, + { + "epoch": 0.006962287608785744, + "grad_norm": 0.33575910329818726, + "learning_rate": 4.965602983837547e-06, + "loss": 0.637, + "step": 84 + }, + { + "epoch": 0.007045171985080812, + "grad_norm": 0.3147188127040863, + "learning_rate": 4.965188561956072e-06, + "loss": 0.6677, + "step": 85 + }, + { + "epoch": 0.00712805636137588, + "grad_norm": 0.2822258770465851, + "learning_rate": 4.964774140074597e-06, + "loss": 0.6491, + "step": 86 + }, + { + "epoch": 0.007210940737670949, + "grad_norm": 0.32119935750961304, + "learning_rate": 4.964359718193121e-06, + "loss": 0.6434, + "step": 87 + }, + { + "epoch": 0.0072938251139660175, + "grad_norm": 0.3112563192844391, + "learning_rate": 4.9639452963116455e-06, + "loss": 0.614, + "step": 88 + }, + { + "epoch": 0.007376709490261086, + "grad_norm": 0.3124278783798218, + "learning_rate": 4.96353087443017e-06, + "loss": 0.6559, + "step": 89 + }, + { + "epoch": 0.0074595938665561546, + "grad_norm": 0.3271758258342743, + "learning_rate": 4.963116452548695e-06, + "loss": 0.6672, + "step": 90 + }, + { + "epoch": 0.007542478242851223, + "grad_norm": 0.3321477174758911, + "learning_rate": 4.96270203066722e-06, + "loss": 0.6432, + "step": 91 + }, + { + "epoch": 0.007625362619146291, + "grad_norm": 0.3202230632305145, + "learning_rate": 4.962287608785744e-06, + "loss": 0.6207, + "step": 92 + }, + { + "epoch": 0.007708246995441359, + "grad_norm": 0.32590359449386597, + "learning_rate": 4.961873186904269e-06, + "loss": 0.6922, + "step": 93 + }, + { + "epoch": 0.007791131371736428, + "grad_norm": 0.32781192660331726, + "learning_rate": 4.961458765022794e-06, + "loss": 0.7143, + "step": 94 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 0.35063037276268005, + "learning_rate": 4.961044343141318e-06, + "loss": 0.6687, + "step": 95 + }, + { + "epoch": 0.007956900124326564, + "grad_norm": 0.3038027584552765, + "learning_rate": 4.960629921259843e-06, + "loss": 0.6277, + "step": 96 + }, + { + "epoch": 0.008039784500621632, + "grad_norm": 0.31324857473373413, + "learning_rate": 4.960215499378367e-06, + "loss": 0.645, + "step": 97 + }, + { + "epoch": 0.008122668876916702, + "grad_norm": 0.3209015429019928, + "learning_rate": 4.959801077496892e-06, + "loss": 0.7003, + "step": 98 + }, + { + "epoch": 0.00820555325321177, + "grad_norm": 0.7662591338157654, + "learning_rate": 4.959386655615417e-06, + "loss": 0.6522, + "step": 99 + }, + { + "epoch": 0.008288437629506838, + "grad_norm": 0.3106798827648163, + "learning_rate": 4.958972233733941e-06, + "loss": 0.6384, + "step": 100 + }, + { + "epoch": 0.008371322005801906, + "grad_norm": 0.3255006670951843, + "learning_rate": 4.9585578118524665e-06, + "loss": 0.6226, + "step": 101 + }, + { + "epoch": 0.008454206382096974, + "grad_norm": 0.36972135305404663, + "learning_rate": 4.958143389970991e-06, + "loss": 0.6801, + "step": 102 + }, + { + "epoch": 0.008537090758392043, + "grad_norm": 0.31064572930336, + "learning_rate": 4.957728968089516e-06, + "loss": 0.6368, + "step": 103 + }, + { + "epoch": 0.00861997513468711, + "grad_norm": 0.32654714584350586, + "learning_rate": 4.957314546208041e-06, + "loss": 0.671, + "step": 104 + }, + { + "epoch": 0.00870285951098218, + "grad_norm": 0.3580387532711029, + "learning_rate": 4.956900124326565e-06, + "loss": 0.7253, + "step": 105 + }, + { + "epoch": 0.008785743887277249, + "grad_norm": 0.4182150661945343, + "learning_rate": 4.9564857024450895e-06, + "loss": 0.5974, + "step": 106 + }, + { + "epoch": 0.008868628263572317, + "grad_norm": 0.3340747654438019, + "learning_rate": 4.956071280563614e-06, + "loss": 0.674, + "step": 107 + }, + { + "epoch": 0.008951512639867385, + "grad_norm": 0.31994062662124634, + "learning_rate": 4.955656858682139e-06, + "loss": 0.6537, + "step": 108 + }, + { + "epoch": 0.009034397016162453, + "grad_norm": 0.33755168318748474, + "learning_rate": 4.955242436800664e-06, + "loss": 0.6597, + "step": 109 + }, + { + "epoch": 0.009117281392457521, + "grad_norm": 0.31428956985473633, + "learning_rate": 4.954828014919188e-06, + "loss": 0.6863, + "step": 110 + }, + { + "epoch": 0.00920016576875259, + "grad_norm": 0.3181600570678711, + "learning_rate": 4.954413593037713e-06, + "loss": 0.6558, + "step": 111 + }, + { + "epoch": 0.009283050145047659, + "grad_norm": 0.3374946415424347, + "learning_rate": 4.953999171156238e-06, + "loss": 0.6651, + "step": 112 + }, + { + "epoch": 0.009365934521342727, + "grad_norm": 0.32209399342536926, + "learning_rate": 4.953584749274762e-06, + "loss": 0.6595, + "step": 113 + }, + { + "epoch": 0.009448818897637795, + "grad_norm": 0.2984916865825653, + "learning_rate": 4.953170327393287e-06, + "loss": 0.623, + "step": 114 + }, + { + "epoch": 0.009531703273932863, + "grad_norm": 0.3013286590576172, + "learning_rate": 4.952755905511811e-06, + "loss": 0.6056, + "step": 115 + }, + { + "epoch": 0.009614587650227931, + "grad_norm": 0.3035356402397156, + "learning_rate": 4.952341483630336e-06, + "loss": 0.6851, + "step": 116 + }, + { + "epoch": 0.009697472026523001, + "grad_norm": 0.6265069842338562, + "learning_rate": 4.951927061748861e-06, + "loss": 0.6636, + "step": 117 + }, + { + "epoch": 0.00978035640281807, + "grad_norm": 0.30655422806739807, + "learning_rate": 4.951512639867385e-06, + "loss": 0.6294, + "step": 118 + }, + { + "epoch": 0.009863240779113137, + "grad_norm": 0.342879980802536, + "learning_rate": 4.9510982179859105e-06, + "loss": 0.6627, + "step": 119 + }, + { + "epoch": 0.009946125155408206, + "grad_norm": 0.315448135137558, + "learning_rate": 4.950683796104435e-06, + "loss": 0.6807, + "step": 120 + }, + { + "epoch": 0.010029009531703274, + "grad_norm": 0.3446184992790222, + "learning_rate": 4.950269374222959e-06, + "loss": 0.6771, + "step": 121 + }, + { + "epoch": 0.010111893907998342, + "grad_norm": 0.3234524428844452, + "learning_rate": 4.949854952341484e-06, + "loss": 0.6528, + "step": 122 + }, + { + "epoch": 0.010194778284293412, + "grad_norm": 0.32572612166404724, + "learning_rate": 4.949440530460008e-06, + "loss": 0.6093, + "step": 123 + }, + { + "epoch": 0.01027766266058848, + "grad_norm": 0.3232816457748413, + "learning_rate": 4.9490261085785335e-06, + "loss": 0.6213, + "step": 124 + }, + { + "epoch": 0.010360547036883548, + "grad_norm": 0.3091711103916168, + "learning_rate": 4.948611686697058e-06, + "loss": 0.6234, + "step": 125 + }, + { + "epoch": 0.010443431413178616, + "grad_norm": 0.33581972122192383, + "learning_rate": 4.948197264815583e-06, + "loss": 0.6538, + "step": 126 + }, + { + "epoch": 0.010526315789473684, + "grad_norm": 0.3572827875614166, + "learning_rate": 4.947782842934108e-06, + "loss": 0.6701, + "step": 127 + }, + { + "epoch": 0.010609200165768752, + "grad_norm": 0.31304413080215454, + "learning_rate": 4.947368421052632e-06, + "loss": 0.642, + "step": 128 + }, + { + "epoch": 0.01069208454206382, + "grad_norm": 0.3085179626941681, + "learning_rate": 4.9469539991711565e-06, + "loss": 0.6705, + "step": 129 + }, + { + "epoch": 0.01077496891835889, + "grad_norm": 0.31415846943855286, + "learning_rate": 4.946539577289681e-06, + "loss": 0.6429, + "step": 130 + }, + { + "epoch": 0.010857853294653958, + "grad_norm": 0.38171055912971497, + "learning_rate": 4.946125155408205e-06, + "loss": 0.6127, + "step": 131 + }, + { + "epoch": 0.010940737670949026, + "grad_norm": 0.3173253536224365, + "learning_rate": 4.945710733526731e-06, + "loss": 0.659, + "step": 132 + }, + { + "epoch": 0.011023622047244094, + "grad_norm": 0.30357474088668823, + "learning_rate": 4.945296311645255e-06, + "loss": 0.5917, + "step": 133 + }, + { + "epoch": 0.011106506423539162, + "grad_norm": 0.3449450433254242, + "learning_rate": 4.94488188976378e-06, + "loss": 0.6102, + "step": 134 + }, + { + "epoch": 0.01118939079983423, + "grad_norm": 0.31959542632102966, + "learning_rate": 4.944467467882305e-06, + "loss": 0.6064, + "step": 135 + }, + { + "epoch": 0.0112722751761293, + "grad_norm": 0.3429710865020752, + "learning_rate": 4.944053046000829e-06, + "loss": 0.6617, + "step": 136 + }, + { + "epoch": 0.011355159552424368, + "grad_norm": 0.29104089736938477, + "learning_rate": 4.943638624119354e-06, + "loss": 0.6574, + "step": 137 + }, + { + "epoch": 0.011438043928719437, + "grad_norm": 0.329677939414978, + "learning_rate": 4.943224202237878e-06, + "loss": 0.5984, + "step": 138 + }, + { + "epoch": 0.011520928305014505, + "grad_norm": 0.3185792565345764, + "learning_rate": 4.942809780356403e-06, + "loss": 0.6255, + "step": 139 + }, + { + "epoch": 0.011603812681309573, + "grad_norm": 0.32871749997138977, + "learning_rate": 4.942395358474928e-06, + "loss": 0.6846, + "step": 140 + }, + { + "epoch": 0.011686697057604641, + "grad_norm": 0.2997575104236603, + "learning_rate": 4.941980936593452e-06, + "loss": 0.6161, + "step": 141 + }, + { + "epoch": 0.01176958143389971, + "grad_norm": 0.2977045178413391, + "learning_rate": 4.9415665147119775e-06, + "loss": 0.6224, + "step": 142 + }, + { + "epoch": 0.011852465810194779, + "grad_norm": 0.40663495659828186, + "learning_rate": 4.941152092830502e-06, + "loss": 0.7222, + "step": 143 + }, + { + "epoch": 0.011935350186489847, + "grad_norm": 0.3244746923446655, + "learning_rate": 4.940737670949027e-06, + "loss": 0.7013, + "step": 144 + }, + { + "epoch": 0.012018234562784915, + "grad_norm": 0.3215189278125763, + "learning_rate": 4.940323249067552e-06, + "loss": 0.6981, + "step": 145 + }, + { + "epoch": 0.012101118939079983, + "grad_norm": 0.31000015139579773, + "learning_rate": 4.939908827186076e-06, + "loss": 0.6794, + "step": 146 + }, + { + "epoch": 0.012184003315375051, + "grad_norm": 0.3462946116924286, + "learning_rate": 4.9394944053046005e-06, + "loss": 0.7085, + "step": 147 + }, + { + "epoch": 0.01226688769167012, + "grad_norm": 0.316339910030365, + "learning_rate": 4.939079983423125e-06, + "loss": 0.6064, + "step": 148 + }, + { + "epoch": 0.01234977206796519, + "grad_norm": 0.3447185158729553, + "learning_rate": 4.938665561541649e-06, + "loss": 0.6726, + "step": 149 + }, + { + "epoch": 0.012432656444260257, + "grad_norm": 0.3491526246070862, + "learning_rate": 4.938251139660175e-06, + "loss": 0.657, + "step": 150 + }, + { + "epoch": 0.012515540820555325, + "grad_norm": 0.31804701685905457, + "learning_rate": 4.937836717778699e-06, + "loss": 0.6579, + "step": 151 + }, + { + "epoch": 0.012598425196850394, + "grad_norm": 0.30622056126594543, + "learning_rate": 4.937422295897224e-06, + "loss": 0.6223, + "step": 152 + }, + { + "epoch": 0.012681309573145462, + "grad_norm": 0.3138669729232788, + "learning_rate": 4.937007874015749e-06, + "loss": 0.65, + "step": 153 + }, + { + "epoch": 0.01276419394944053, + "grad_norm": 0.2930069863796234, + "learning_rate": 4.936593452134273e-06, + "loss": 0.5869, + "step": 154 + }, + { + "epoch": 0.0128470783257356, + "grad_norm": 0.31072062253952026, + "learning_rate": 4.936179030252798e-06, + "loss": 0.6191, + "step": 155 + }, + { + "epoch": 0.012929962702030668, + "grad_norm": 0.31300970911979675, + "learning_rate": 4.935764608371322e-06, + "loss": 0.6245, + "step": 156 + }, + { + "epoch": 0.013012847078325736, + "grad_norm": 0.3140801191329956, + "learning_rate": 4.935350186489847e-06, + "loss": 0.6307, + "step": 157 + }, + { + "epoch": 0.013095731454620804, + "grad_norm": 0.3496628701686859, + "learning_rate": 4.934935764608372e-06, + "loss": 0.6967, + "step": 158 + }, + { + "epoch": 0.013178615830915872, + "grad_norm": 0.2977752685546875, + "learning_rate": 4.934521342726896e-06, + "loss": 0.59, + "step": 159 + }, + { + "epoch": 0.01326150020721094, + "grad_norm": 0.32420387864112854, + "learning_rate": 4.9341069208454215e-06, + "loss": 0.6111, + "step": 160 + }, + { + "epoch": 0.01334438458350601, + "grad_norm": 0.2900985777378082, + "learning_rate": 4.933692498963946e-06, + "loss": 0.5941, + "step": 161 + }, + { + "epoch": 0.013427268959801078, + "grad_norm": 0.3186474144458771, + "learning_rate": 4.93327807708247e-06, + "loss": 0.61, + "step": 162 + }, + { + "epoch": 0.013510153336096146, + "grad_norm": 0.32691630721092224, + "learning_rate": 4.932863655200995e-06, + "loss": 0.6427, + "step": 163 + }, + { + "epoch": 0.013593037712391214, + "grad_norm": 0.3174779713153839, + "learning_rate": 4.932449233319519e-06, + "loss": 0.6537, + "step": 164 + }, + { + "epoch": 0.013675922088686282, + "grad_norm": 0.32977497577667236, + "learning_rate": 4.9320348114380445e-06, + "loss": 0.6476, + "step": 165 + }, + { + "epoch": 0.01375880646498135, + "grad_norm": 0.31561365723609924, + "learning_rate": 4.931620389556569e-06, + "loss": 0.6756, + "step": 166 + }, + { + "epoch": 0.013841690841276419, + "grad_norm": 0.34246233105659485, + "learning_rate": 4.931205967675093e-06, + "loss": 0.6909, + "step": 167 + }, + { + "epoch": 0.013924575217571488, + "grad_norm": 0.318207323551178, + "learning_rate": 4.930791545793619e-06, + "loss": 0.6784, + "step": 168 + }, + { + "epoch": 0.014007459593866557, + "grad_norm": 0.3145054578781128, + "learning_rate": 4.930377123912143e-06, + "loss": 0.6724, + "step": 169 + }, + { + "epoch": 0.014090343970161625, + "grad_norm": 0.3657659590244293, + "learning_rate": 4.9299627020306675e-06, + "loss": 0.6733, + "step": 170 + }, + { + "epoch": 0.014173228346456693, + "grad_norm": 0.3002912402153015, + "learning_rate": 4.929548280149192e-06, + "loss": 0.6307, + "step": 171 + }, + { + "epoch": 0.01425611272275176, + "grad_norm": 0.3236493766307831, + "learning_rate": 4.929133858267716e-06, + "loss": 0.6186, + "step": 172 + }, + { + "epoch": 0.014338997099046829, + "grad_norm": 0.33164846897125244, + "learning_rate": 4.928719436386242e-06, + "loss": 0.6486, + "step": 173 + }, + { + "epoch": 0.014421881475341899, + "grad_norm": 0.3012412190437317, + "learning_rate": 4.928305014504766e-06, + "loss": 0.6091, + "step": 174 + }, + { + "epoch": 0.014504765851636967, + "grad_norm": 0.27425482869148254, + "learning_rate": 4.927890592623291e-06, + "loss": 0.561, + "step": 175 + }, + { + "epoch": 0.014587650227932035, + "grad_norm": 0.3252134323120117, + "learning_rate": 4.927476170741816e-06, + "loss": 0.7003, + "step": 176 + }, + { + "epoch": 0.014670534604227103, + "grad_norm": 0.31894752383232117, + "learning_rate": 4.92706174886034e-06, + "loss": 0.6531, + "step": 177 + }, + { + "epoch": 0.014753418980522171, + "grad_norm": 0.33628857135772705, + "learning_rate": 4.926647326978865e-06, + "loss": 0.6689, + "step": 178 + }, + { + "epoch": 0.01483630335681724, + "grad_norm": 0.3104929029941559, + "learning_rate": 4.926232905097389e-06, + "loss": 0.6728, + "step": 179 + }, + { + "epoch": 0.014919187733112309, + "grad_norm": 0.3061523735523224, + "learning_rate": 4.925818483215914e-06, + "loss": 0.6116, + "step": 180 + }, + { + "epoch": 0.015002072109407377, + "grad_norm": 0.36103665828704834, + "learning_rate": 4.925404061334439e-06, + "loss": 0.6362, + "step": 181 + }, + { + "epoch": 0.015084956485702445, + "grad_norm": 0.3153567314147949, + "learning_rate": 4.924989639452963e-06, + "loss": 0.6412, + "step": 182 + }, + { + "epoch": 0.015167840861997513, + "grad_norm": 0.29443851113319397, + "learning_rate": 4.9245752175714885e-06, + "loss": 0.6298, + "step": 183 + }, + { + "epoch": 0.015250725238292582, + "grad_norm": 0.3135778605937958, + "learning_rate": 4.924160795690013e-06, + "loss": 0.603, + "step": 184 + }, + { + "epoch": 0.01533360961458765, + "grad_norm": 0.2895752489566803, + "learning_rate": 4.923746373808537e-06, + "loss": 0.5994, + "step": 185 + }, + { + "epoch": 0.015416493990882718, + "grad_norm": 0.37467554211616516, + "learning_rate": 4.923331951927063e-06, + "loss": 0.6396, + "step": 186 + }, + { + "epoch": 0.015499378367177788, + "grad_norm": 0.32802945375442505, + "learning_rate": 4.922917530045587e-06, + "loss": 0.6418, + "step": 187 + }, + { + "epoch": 0.015582262743472856, + "grad_norm": 0.3270801305770874, + "learning_rate": 4.9225031081641115e-06, + "loss": 0.6251, + "step": 188 + }, + { + "epoch": 0.015665147119767922, + "grad_norm": 0.4329809546470642, + "learning_rate": 4.922088686282636e-06, + "loss": 0.6495, + "step": 189 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 0.3015666604042053, + "learning_rate": 4.92167426440116e-06, + "loss": 0.6422, + "step": 190 + }, + { + "epoch": 0.015830915872358062, + "grad_norm": 0.3216380476951599, + "learning_rate": 4.921259842519686e-06, + "loss": 0.6731, + "step": 191 + }, + { + "epoch": 0.015913800248653128, + "grad_norm": 0.3234592378139496, + "learning_rate": 4.92084542063821e-06, + "loss": 0.6713, + "step": 192 + }, + { + "epoch": 0.015996684624948198, + "grad_norm": 0.29352617263793945, + "learning_rate": 4.920430998756735e-06, + "loss": 0.6136, + "step": 193 + }, + { + "epoch": 0.016079569001243264, + "grad_norm": 0.311693012714386, + "learning_rate": 4.92001657687526e-06, + "loss": 0.6304, + "step": 194 + }, + { + "epoch": 0.016162453377538334, + "grad_norm": 0.2691800892353058, + "learning_rate": 4.919602154993784e-06, + "loss": 0.5679, + "step": 195 + }, + { + "epoch": 0.016245337753833404, + "grad_norm": 0.3082760274410248, + "learning_rate": 4.919187733112309e-06, + "loss": 0.6315, + "step": 196 + }, + { + "epoch": 0.01632822213012847, + "grad_norm": 0.32791590690612793, + "learning_rate": 4.918773311230833e-06, + "loss": 0.6312, + "step": 197 + }, + { + "epoch": 0.01641110650642354, + "grad_norm": 0.3061068058013916, + "learning_rate": 4.9183588893493575e-06, + "loss": 0.5666, + "step": 198 + }, + { + "epoch": 0.016493990882718607, + "grad_norm": 0.3636856973171234, + "learning_rate": 4.917944467467883e-06, + "loss": 0.6422, + "step": 199 + }, + { + "epoch": 0.016576875259013676, + "grad_norm": 0.3132879137992859, + "learning_rate": 4.917530045586407e-06, + "loss": 0.6418, + "step": 200 + }, + { + "epoch": 0.016659759635308743, + "grad_norm": 0.33875274658203125, + "learning_rate": 4.9171156237049325e-06, + "loss": 0.6382, + "step": 201 + }, + { + "epoch": 0.016742644011603813, + "grad_norm": 0.2787725329399109, + "learning_rate": 4.916701201823457e-06, + "loss": 0.5663, + "step": 202 + }, + { + "epoch": 0.016825528387898882, + "grad_norm": 0.3086506724357605, + "learning_rate": 4.916286779941981e-06, + "loss": 0.6473, + "step": 203 + }, + { + "epoch": 0.01690841276419395, + "grad_norm": 0.3189477324485779, + "learning_rate": 4.915872358060506e-06, + "loss": 0.6143, + "step": 204 + }, + { + "epoch": 0.01699129714048902, + "grad_norm": 0.3477078080177307, + "learning_rate": 4.91545793617903e-06, + "loss": 0.6133, + "step": 205 + }, + { + "epoch": 0.017074181516784085, + "grad_norm": 0.30549299716949463, + "learning_rate": 4.9150435142975555e-06, + "loss": 0.6454, + "step": 206 + }, + { + "epoch": 0.017157065893079155, + "grad_norm": 0.3125954866409302, + "learning_rate": 4.91462909241608e-06, + "loss": 0.6758, + "step": 207 + }, + { + "epoch": 0.01723995026937422, + "grad_norm": 0.3262520730495453, + "learning_rate": 4.914214670534604e-06, + "loss": 0.6354, + "step": 208 + }, + { + "epoch": 0.01732283464566929, + "grad_norm": 0.3060142993927002, + "learning_rate": 4.91380024865313e-06, + "loss": 0.6888, + "step": 209 + }, + { + "epoch": 0.01740571902196436, + "grad_norm": 0.3004607558250427, + "learning_rate": 4.913385826771654e-06, + "loss": 0.6457, + "step": 210 + }, + { + "epoch": 0.017488603398259427, + "grad_norm": 0.32081931829452515, + "learning_rate": 4.9129714048901785e-06, + "loss": 0.6107, + "step": 211 + }, + { + "epoch": 0.017571487774554497, + "grad_norm": 0.3165818750858307, + "learning_rate": 4.912556983008703e-06, + "loss": 0.6609, + "step": 212 + }, + { + "epoch": 0.017654372150849564, + "grad_norm": 0.34641551971435547, + "learning_rate": 4.912142561127227e-06, + "loss": 0.6452, + "step": 213 + }, + { + "epoch": 0.017737256527144633, + "grad_norm": 0.3206797242164612, + "learning_rate": 4.911728139245753e-06, + "loss": 0.7105, + "step": 214 + }, + { + "epoch": 0.017820140903439703, + "grad_norm": 0.3065459728240967, + "learning_rate": 4.911313717364277e-06, + "loss": 0.6151, + "step": 215 + }, + { + "epoch": 0.01790302527973477, + "grad_norm": 0.3201547861099243, + "learning_rate": 4.9108992954828015e-06, + "loss": 0.6815, + "step": 216 + }, + { + "epoch": 0.01798590965602984, + "grad_norm": 0.33410829305648804, + "learning_rate": 4.910484873601327e-06, + "loss": 0.5845, + "step": 217 + }, + { + "epoch": 0.018068794032324906, + "grad_norm": 0.3045617640018463, + "learning_rate": 4.910070451719851e-06, + "loss": 0.6079, + "step": 218 + }, + { + "epoch": 0.018151678408619976, + "grad_norm": 0.3213278651237488, + "learning_rate": 4.909656029838376e-06, + "loss": 0.6134, + "step": 219 + }, + { + "epoch": 0.018234562784915042, + "grad_norm": 0.3176189363002777, + "learning_rate": 4.909241607956901e-06, + "loss": 0.668, + "step": 220 + }, + { + "epoch": 0.018317447161210112, + "grad_norm": 0.2997208535671234, + "learning_rate": 4.908827186075425e-06, + "loss": 0.6477, + "step": 221 + }, + { + "epoch": 0.01840033153750518, + "grad_norm": 0.32151147723197937, + "learning_rate": 4.90841276419395e-06, + "loss": 0.6384, + "step": 222 + }, + { + "epoch": 0.018483215913800248, + "grad_norm": 0.2956540882587433, + "learning_rate": 4.907998342312474e-06, + "loss": 0.6346, + "step": 223 + }, + { + "epoch": 0.018566100290095318, + "grad_norm": 0.2986004948616028, + "learning_rate": 4.9075839204309995e-06, + "loss": 0.6241, + "step": 224 + }, + { + "epoch": 0.018648984666390384, + "grad_norm": 0.3227016031742096, + "learning_rate": 4.907169498549524e-06, + "loss": 0.5883, + "step": 225 + }, + { + "epoch": 0.018731869042685454, + "grad_norm": 0.31710103154182434, + "learning_rate": 4.906755076668048e-06, + "loss": 0.6111, + "step": 226 + }, + { + "epoch": 0.01881475341898052, + "grad_norm": 0.31441381573677063, + "learning_rate": 4.906340654786574e-06, + "loss": 0.6434, + "step": 227 + }, + { + "epoch": 0.01889763779527559, + "grad_norm": 0.2959873676300049, + "learning_rate": 4.905926232905098e-06, + "loss": 0.6809, + "step": 228 + }, + { + "epoch": 0.01898052217157066, + "grad_norm": 0.3048423230648041, + "learning_rate": 4.9055118110236225e-06, + "loss": 0.603, + "step": 229 + }, + { + "epoch": 0.019063406547865726, + "grad_norm": 0.42168930172920227, + "learning_rate": 4.905097389142147e-06, + "loss": 0.6821, + "step": 230 + }, + { + "epoch": 0.019146290924160796, + "grad_norm": 0.35032355785369873, + "learning_rate": 4.904682967260671e-06, + "loss": 0.602, + "step": 231 + }, + { + "epoch": 0.019229175300455863, + "grad_norm": 0.327216774225235, + "learning_rate": 4.904268545379197e-06, + "loss": 0.6564, + "step": 232 + }, + { + "epoch": 0.019312059676750933, + "grad_norm": 0.3228274881839752, + "learning_rate": 4.903854123497721e-06, + "loss": 0.6464, + "step": 233 + }, + { + "epoch": 0.019394944053046002, + "grad_norm": 0.2945462167263031, + "learning_rate": 4.9034397016162455e-06, + "loss": 0.5992, + "step": 234 + }, + { + "epoch": 0.01947782842934107, + "grad_norm": 0.2860357165336609, + "learning_rate": 4.903025279734771e-06, + "loss": 0.5836, + "step": 235 + }, + { + "epoch": 0.01956071280563614, + "grad_norm": 0.3317335844039917, + "learning_rate": 4.902610857853295e-06, + "loss": 0.6318, + "step": 236 + }, + { + "epoch": 0.019643597181931205, + "grad_norm": 0.330287367105484, + "learning_rate": 4.90219643597182e-06, + "loss": 0.6247, + "step": 237 + }, + { + "epoch": 0.019726481558226275, + "grad_norm": 0.34137266874313354, + "learning_rate": 4.901782014090344e-06, + "loss": 0.6311, + "step": 238 + }, + { + "epoch": 0.01980936593452134, + "grad_norm": 0.30789539217948914, + "learning_rate": 4.9013675922088685e-06, + "loss": 0.6696, + "step": 239 + }, + { + "epoch": 0.01989225031081641, + "grad_norm": 0.3023962676525116, + "learning_rate": 4.900953170327394e-06, + "loss": 0.6148, + "step": 240 + }, + { + "epoch": 0.01997513468711148, + "grad_norm": 0.2939088046550751, + "learning_rate": 4.900538748445918e-06, + "loss": 0.5945, + "step": 241 + }, + { + "epoch": 0.020058019063406547, + "grad_norm": 0.32027408480644226, + "learning_rate": 4.9001243265644435e-06, + "loss": 0.6518, + "step": 242 + }, + { + "epoch": 0.020140903439701617, + "grad_norm": 0.28385740518569946, + "learning_rate": 4.899709904682968e-06, + "loss": 0.5701, + "step": 243 + }, + { + "epoch": 0.020223787815996683, + "grad_norm": 0.3051215708255768, + "learning_rate": 4.899295482801492e-06, + "loss": 0.6354, + "step": 244 + }, + { + "epoch": 0.020306672192291753, + "grad_norm": 0.3225044012069702, + "learning_rate": 4.898881060920017e-06, + "loss": 0.7132, + "step": 245 + }, + { + "epoch": 0.020389556568586823, + "grad_norm": 0.2951192855834961, + "learning_rate": 4.898466639038541e-06, + "loss": 0.6597, + "step": 246 + }, + { + "epoch": 0.02047244094488189, + "grad_norm": 0.283244788646698, + "learning_rate": 4.898052217157066e-06, + "loss": 0.6204, + "step": 247 + }, + { + "epoch": 0.02055532532117696, + "grad_norm": 0.3243666887283325, + "learning_rate": 4.897637795275591e-06, + "loss": 0.6838, + "step": 248 + }, + { + "epoch": 0.020638209697472026, + "grad_norm": 0.2845022976398468, + "learning_rate": 4.897223373394115e-06, + "loss": 0.5907, + "step": 249 + }, + { + "epoch": 0.020721094073767096, + "grad_norm": 0.29496684670448303, + "learning_rate": 4.896808951512641e-06, + "loss": 0.6198, + "step": 250 + }, + { + "epoch": 0.020803978450062162, + "grad_norm": 0.31026995182037354, + "learning_rate": 4.896394529631165e-06, + "loss": 0.6156, + "step": 251 + }, + { + "epoch": 0.020886862826357232, + "grad_norm": 0.3020731508731842, + "learning_rate": 4.8959801077496895e-06, + "loss": 0.6898, + "step": 252 + }, + { + "epoch": 0.0209697472026523, + "grad_norm": 0.30239197611808777, + "learning_rate": 4.895565685868214e-06, + "loss": 0.5673, + "step": 253 + }, + { + "epoch": 0.021052631578947368, + "grad_norm": 0.3164243698120117, + "learning_rate": 4.895151263986738e-06, + "loss": 0.6912, + "step": 254 + }, + { + "epoch": 0.021135515955242438, + "grad_norm": 0.3250356912612915, + "learning_rate": 4.894736842105264e-06, + "loss": 0.6403, + "step": 255 + }, + { + "epoch": 0.021218400331537504, + "grad_norm": 0.30877047777175903, + "learning_rate": 4.894322420223788e-06, + "loss": 0.5803, + "step": 256 + }, + { + "epoch": 0.021301284707832574, + "grad_norm": 0.3121374845504761, + "learning_rate": 4.8939079983423125e-06, + "loss": 0.6368, + "step": 257 + }, + { + "epoch": 0.02138416908412764, + "grad_norm": 0.34734898805618286, + "learning_rate": 4.893493576460838e-06, + "loss": 0.657, + "step": 258 + }, + { + "epoch": 0.02146705346042271, + "grad_norm": 0.2923796474933624, + "learning_rate": 4.893079154579362e-06, + "loss": 0.6247, + "step": 259 + }, + { + "epoch": 0.02154993783671778, + "grad_norm": 0.33660832047462463, + "learning_rate": 4.892664732697887e-06, + "loss": 0.6437, + "step": 260 + }, + { + "epoch": 0.021632822213012846, + "grad_norm": 0.3240174353122711, + "learning_rate": 4.892250310816412e-06, + "loss": 0.6591, + "step": 261 + }, + { + "epoch": 0.021715706589307916, + "grad_norm": 0.2993757426738739, + "learning_rate": 4.891835888934936e-06, + "loss": 0.6319, + "step": 262 + }, + { + "epoch": 0.021798590965602983, + "grad_norm": 0.33616501092910767, + "learning_rate": 4.891421467053461e-06, + "loss": 0.6411, + "step": 263 + }, + { + "epoch": 0.021881475341898052, + "grad_norm": 0.28011152148246765, + "learning_rate": 4.891007045171985e-06, + "loss": 0.6199, + "step": 264 + }, + { + "epoch": 0.021964359718193122, + "grad_norm": 0.319423645734787, + "learning_rate": 4.89059262329051e-06, + "loss": 0.7031, + "step": 265 + }, + { + "epoch": 0.02204724409448819, + "grad_norm": 0.2669227421283722, + "learning_rate": 4.890178201409035e-06, + "loss": 0.5739, + "step": 266 + }, + { + "epoch": 0.02213012847078326, + "grad_norm": 0.30817216634750366, + "learning_rate": 4.889763779527559e-06, + "loss": 0.5984, + "step": 267 + }, + { + "epoch": 0.022213012847078325, + "grad_norm": 0.3059421479701996, + "learning_rate": 4.889349357646085e-06, + "loss": 0.6206, + "step": 268 + }, + { + "epoch": 0.022295897223373395, + "grad_norm": 0.321160227060318, + "learning_rate": 4.888934935764609e-06, + "loss": 0.6149, + "step": 269 + }, + { + "epoch": 0.02237878159966846, + "grad_norm": 0.37409359216690063, + "learning_rate": 4.8885205138831335e-06, + "loss": 0.6091, + "step": 270 + }, + { + "epoch": 0.02246166597596353, + "grad_norm": 0.3047485649585724, + "learning_rate": 4.888106092001658e-06, + "loss": 0.6017, + "step": 271 + }, + { + "epoch": 0.0225445503522586, + "grad_norm": 0.3171643614768982, + "learning_rate": 4.887691670120182e-06, + "loss": 0.626, + "step": 272 + }, + { + "epoch": 0.022627434728553667, + "grad_norm": 0.2963331937789917, + "learning_rate": 4.887277248238708e-06, + "loss": 0.5869, + "step": 273 + }, + { + "epoch": 0.022710319104848737, + "grad_norm": 0.30209407210350037, + "learning_rate": 4.886862826357232e-06, + "loss": 0.6315, + "step": 274 + }, + { + "epoch": 0.022793203481143803, + "grad_norm": 0.29272013902664185, + "learning_rate": 4.8864484044757566e-06, + "loss": 0.6624, + "step": 275 + }, + { + "epoch": 0.022876087857438873, + "grad_norm": 0.3921303451061249, + "learning_rate": 4.886033982594282e-06, + "loss": 0.6245, + "step": 276 + }, + { + "epoch": 0.02295897223373394, + "grad_norm": 0.2882024049758911, + "learning_rate": 4.885619560712806e-06, + "loss": 0.5938, + "step": 277 + }, + { + "epoch": 0.02304185661002901, + "grad_norm": 0.28235411643981934, + "learning_rate": 4.885205138831331e-06, + "loss": 0.634, + "step": 278 + }, + { + "epoch": 0.02312474098632408, + "grad_norm": 0.31357911229133606, + "learning_rate": 4.884790716949855e-06, + "loss": 0.6589, + "step": 279 + }, + { + "epoch": 0.023207625362619146, + "grad_norm": 0.2932921051979065, + "learning_rate": 4.8843762950683796e-06, + "loss": 0.5835, + "step": 280 + }, + { + "epoch": 0.023290509738914215, + "grad_norm": 0.292555570602417, + "learning_rate": 4.883961873186905e-06, + "loss": 0.6323, + "step": 281 + }, + { + "epoch": 0.023373394115209282, + "grad_norm": 0.3716922998428345, + "learning_rate": 4.883547451305429e-06, + "loss": 0.5944, + "step": 282 + }, + { + "epoch": 0.02345627849150435, + "grad_norm": 0.3041526973247528, + "learning_rate": 4.883133029423954e-06, + "loss": 0.6152, + "step": 283 + }, + { + "epoch": 0.02353916286779942, + "grad_norm": 0.29591459035873413, + "learning_rate": 4.882718607542479e-06, + "loss": 0.6702, + "step": 284 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 0.31460806727409363, + "learning_rate": 4.882304185661003e-06, + "loss": 0.6501, + "step": 285 + }, + { + "epoch": 0.023704931620389558, + "grad_norm": 0.3428550064563751, + "learning_rate": 4.881889763779528e-06, + "loss": 0.6231, + "step": 286 + }, + { + "epoch": 0.023787815996684624, + "grad_norm": 0.3053310513496399, + "learning_rate": 4.881475341898052e-06, + "loss": 0.6382, + "step": 287 + }, + { + "epoch": 0.023870700372979694, + "grad_norm": 0.29984110593795776, + "learning_rate": 4.881060920016577e-06, + "loss": 0.6843, + "step": 288 + }, + { + "epoch": 0.02395358474927476, + "grad_norm": 0.3051556348800659, + "learning_rate": 4.880646498135102e-06, + "loss": 0.6583, + "step": 289 + }, + { + "epoch": 0.02403646912556983, + "grad_norm": 0.32842400670051575, + "learning_rate": 4.880232076253626e-06, + "loss": 0.6383, + "step": 290 + }, + { + "epoch": 0.0241193535018649, + "grad_norm": 0.3182846009731293, + "learning_rate": 4.879817654372152e-06, + "loss": 0.6487, + "step": 291 + }, + { + "epoch": 0.024202237878159966, + "grad_norm": 0.3067048490047455, + "learning_rate": 4.879403232490676e-06, + "loss": 0.6408, + "step": 292 + }, + { + "epoch": 0.024285122254455036, + "grad_norm": 0.2941698133945465, + "learning_rate": 4.8789888106092006e-06, + "loss": 0.6183, + "step": 293 + }, + { + "epoch": 0.024368006630750103, + "grad_norm": 0.301107257604599, + "learning_rate": 4.878574388727725e-06, + "loss": 0.631, + "step": 294 + }, + { + "epoch": 0.024450891007045172, + "grad_norm": 0.2929108440876007, + "learning_rate": 4.8781599668462494e-06, + "loss": 0.6141, + "step": 295 + }, + { + "epoch": 0.02453377538334024, + "grad_norm": 0.2889519929885864, + "learning_rate": 4.877745544964775e-06, + "loss": 0.6456, + "step": 296 + }, + { + "epoch": 0.02461665975963531, + "grad_norm": 0.3041543960571289, + "learning_rate": 4.877331123083299e-06, + "loss": 0.6503, + "step": 297 + }, + { + "epoch": 0.02469954413593038, + "grad_norm": 0.3732686936855316, + "learning_rate": 4.8769167012018236e-06, + "loss": 0.623, + "step": 298 + }, + { + "epoch": 0.024782428512225445, + "grad_norm": 0.3293933868408203, + "learning_rate": 4.876502279320349e-06, + "loss": 0.6342, + "step": 299 + }, + { + "epoch": 0.024865312888520515, + "grad_norm": 0.3005862832069397, + "learning_rate": 4.876087857438873e-06, + "loss": 0.5853, + "step": 300 + }, + { + "epoch": 0.02494819726481558, + "grad_norm": 0.2854812741279602, + "learning_rate": 4.875673435557398e-06, + "loss": 0.6139, + "step": 301 + }, + { + "epoch": 0.02503108164111065, + "grad_norm": 0.3190535604953766, + "learning_rate": 4.875259013675923e-06, + "loss": 0.6582, + "step": 302 + }, + { + "epoch": 0.02511396601740572, + "grad_norm": 0.31284043192863464, + "learning_rate": 4.874844591794447e-06, + "loss": 0.6445, + "step": 303 + }, + { + "epoch": 0.025196850393700787, + "grad_norm": 0.3100276589393616, + "learning_rate": 4.874430169912972e-06, + "loss": 0.665, + "step": 304 + }, + { + "epoch": 0.025279734769995857, + "grad_norm": 0.2978265583515167, + "learning_rate": 4.874015748031496e-06, + "loss": 0.6342, + "step": 305 + }, + { + "epoch": 0.025362619146290923, + "grad_norm": 0.31898847222328186, + "learning_rate": 4.873601326150021e-06, + "loss": 0.6189, + "step": 306 + }, + { + "epoch": 0.025445503522585993, + "grad_norm": 0.3147716522216797, + "learning_rate": 4.873186904268546e-06, + "loss": 0.5964, + "step": 307 + }, + { + "epoch": 0.02552838789888106, + "grad_norm": 0.3025127649307251, + "learning_rate": 4.8727724823870704e-06, + "loss": 0.6185, + "step": 308 + }, + { + "epoch": 0.02561127227517613, + "grad_norm": 0.3042539954185486, + "learning_rate": 4.872358060505596e-06, + "loss": 0.6164, + "step": 309 + }, + { + "epoch": 0.0256941566514712, + "grad_norm": 0.2934959828853607, + "learning_rate": 4.87194363862412e-06, + "loss": 0.6652, + "step": 310 + }, + { + "epoch": 0.025777041027766266, + "grad_norm": 0.3180948793888092, + "learning_rate": 4.8715292167426446e-06, + "loss": 0.6766, + "step": 311 + }, + { + "epoch": 0.025859925404061335, + "grad_norm": 0.3000025749206543, + "learning_rate": 4.871114794861169e-06, + "loss": 0.6377, + "step": 312 + }, + { + "epoch": 0.0259428097803564, + "grad_norm": 0.2975216209888458, + "learning_rate": 4.8707003729796934e-06, + "loss": 0.6254, + "step": 313 + }, + { + "epoch": 0.02602569415665147, + "grad_norm": 0.2884451448917389, + "learning_rate": 4.870285951098218e-06, + "loss": 0.584, + "step": 314 + }, + { + "epoch": 0.026108578532946538, + "grad_norm": 0.3538367748260498, + "learning_rate": 4.869871529216743e-06, + "loss": 0.6523, + "step": 315 + }, + { + "epoch": 0.026191462909241608, + "grad_norm": 0.2856776714324951, + "learning_rate": 4.8694571073352676e-06, + "loss": 0.6364, + "step": 316 + }, + { + "epoch": 0.026274347285536678, + "grad_norm": 0.32078224420547485, + "learning_rate": 4.869042685453793e-06, + "loss": 0.6749, + "step": 317 + }, + { + "epoch": 0.026357231661831744, + "grad_norm": 0.3473566770553589, + "learning_rate": 4.868628263572317e-06, + "loss": 0.6772, + "step": 318 + }, + { + "epoch": 0.026440116038126814, + "grad_norm": 0.3199005126953125, + "learning_rate": 4.868213841690842e-06, + "loss": 0.6785, + "step": 319 + }, + { + "epoch": 0.02652300041442188, + "grad_norm": 0.3112602233886719, + "learning_rate": 4.867799419809366e-06, + "loss": 0.6597, + "step": 320 + }, + { + "epoch": 0.02660588479071695, + "grad_norm": 0.5189213156700134, + "learning_rate": 4.867384997927891e-06, + "loss": 0.6233, + "step": 321 + }, + { + "epoch": 0.02668876916701202, + "grad_norm": 0.3000120520591736, + "learning_rate": 4.866970576046416e-06, + "loss": 0.6819, + "step": 322 + }, + { + "epoch": 0.026771653543307086, + "grad_norm": 0.3178681433200836, + "learning_rate": 4.86655615416494e-06, + "loss": 0.6533, + "step": 323 + }, + { + "epoch": 0.026854537919602156, + "grad_norm": 0.30063506960868835, + "learning_rate": 4.866141732283465e-06, + "loss": 0.5773, + "step": 324 + }, + { + "epoch": 0.026937422295897222, + "grad_norm": 0.318668931722641, + "learning_rate": 4.86572731040199e-06, + "loss": 0.664, + "step": 325 + }, + { + "epoch": 0.027020306672192292, + "grad_norm": 0.3565739393234253, + "learning_rate": 4.8653128885205144e-06, + "loss": 0.6595, + "step": 326 + }, + { + "epoch": 0.02710319104848736, + "grad_norm": 0.2848496437072754, + "learning_rate": 4.864898466639039e-06, + "loss": 0.5806, + "step": 327 + }, + { + "epoch": 0.02718607542478243, + "grad_norm": 0.3243609368801117, + "learning_rate": 4.864484044757563e-06, + "loss": 0.5733, + "step": 328 + }, + { + "epoch": 0.0272689598010775, + "grad_norm": 0.30274537205696106, + "learning_rate": 4.864069622876088e-06, + "loss": 0.6398, + "step": 329 + }, + { + "epoch": 0.027351844177372565, + "grad_norm": 0.2830435037612915, + "learning_rate": 4.863655200994613e-06, + "loss": 0.573, + "step": 330 + }, + { + "epoch": 0.027434728553667635, + "grad_norm": 0.30771830677986145, + "learning_rate": 4.8632407791131374e-06, + "loss": 0.646, + "step": 331 + }, + { + "epoch": 0.0275176129299627, + "grad_norm": 0.2762799561023712, + "learning_rate": 4.862826357231662e-06, + "loss": 0.5693, + "step": 332 + }, + { + "epoch": 0.02760049730625777, + "grad_norm": 0.3032577931880951, + "learning_rate": 4.862411935350187e-06, + "loss": 0.6346, + "step": 333 + }, + { + "epoch": 0.027683381682552837, + "grad_norm": 0.49638327956199646, + "learning_rate": 4.861997513468712e-06, + "loss": 0.605, + "step": 334 + }, + { + "epoch": 0.027766266058847907, + "grad_norm": 0.33385494351387024, + "learning_rate": 4.861583091587236e-06, + "loss": 0.6251, + "step": 335 + }, + { + "epoch": 0.027849150435142977, + "grad_norm": 0.299856573343277, + "learning_rate": 4.8611686697057604e-06, + "loss": 0.5931, + "step": 336 + }, + { + "epoch": 0.027932034811438043, + "grad_norm": 0.3317852020263672, + "learning_rate": 4.860754247824285e-06, + "loss": 0.6046, + "step": 337 + }, + { + "epoch": 0.028014919187733113, + "grad_norm": 0.3349105715751648, + "learning_rate": 4.86033982594281e-06, + "loss": 0.6522, + "step": 338 + }, + { + "epoch": 0.02809780356402818, + "grad_norm": 0.3012475371360779, + "learning_rate": 4.859925404061335e-06, + "loss": 0.5979, + "step": 339 + }, + { + "epoch": 0.02818068794032325, + "grad_norm": 0.3019346296787262, + "learning_rate": 4.85951098217986e-06, + "loss": 0.5804, + "step": 340 + }, + { + "epoch": 0.02826357231661832, + "grad_norm": 0.3353364169597626, + "learning_rate": 4.859096560298384e-06, + "loss": 0.5972, + "step": 341 + }, + { + "epoch": 0.028346456692913385, + "grad_norm": 0.3266708254814148, + "learning_rate": 4.858682138416909e-06, + "loss": 0.6253, + "step": 342 + }, + { + "epoch": 0.028429341069208455, + "grad_norm": 0.32890817523002625, + "learning_rate": 4.858267716535434e-06, + "loss": 0.5827, + "step": 343 + }, + { + "epoch": 0.02851222544550352, + "grad_norm": 0.2913133203983307, + "learning_rate": 4.8578532946539584e-06, + "loss": 0.5729, + "step": 344 + }, + { + "epoch": 0.02859510982179859, + "grad_norm": 0.3419067859649658, + "learning_rate": 4.857438872772483e-06, + "loss": 0.6263, + "step": 345 + }, + { + "epoch": 0.028677994198093658, + "grad_norm": 0.3260219097137451, + "learning_rate": 4.857024450891007e-06, + "loss": 0.6024, + "step": 346 + }, + { + "epoch": 0.028760878574388728, + "grad_norm": 0.3224937915802002, + "learning_rate": 4.856610029009532e-06, + "loss": 0.6274, + "step": 347 + }, + { + "epoch": 0.028843762950683798, + "grad_norm": 0.28874996304512024, + "learning_rate": 4.856195607128057e-06, + "loss": 0.5732, + "step": 348 + }, + { + "epoch": 0.028926647326978864, + "grad_norm": 0.31316304206848145, + "learning_rate": 4.8557811852465814e-06, + "loss": 0.636, + "step": 349 + }, + { + "epoch": 0.029009531703273934, + "grad_norm": 0.3483511209487915, + "learning_rate": 4.855366763365106e-06, + "loss": 0.6297, + "step": 350 + }, + { + "epoch": 0.029092416079569, + "grad_norm": 0.26793965697288513, + "learning_rate": 4.854952341483631e-06, + "loss": 0.5578, + "step": 351 + }, + { + "epoch": 0.02917530045586407, + "grad_norm": 0.31637680530548096, + "learning_rate": 4.854537919602156e-06, + "loss": 0.6218, + "step": 352 + }, + { + "epoch": 0.029258184832159136, + "grad_norm": 0.2908051013946533, + "learning_rate": 4.85412349772068e-06, + "loss": 0.5761, + "step": 353 + }, + { + "epoch": 0.029341069208454206, + "grad_norm": 0.315992534160614, + "learning_rate": 4.8537090758392045e-06, + "loss": 0.6402, + "step": 354 + }, + { + "epoch": 0.029423953584749276, + "grad_norm": 0.2901885509490967, + "learning_rate": 4.853294653957729e-06, + "loss": 0.5732, + "step": 355 + }, + { + "epoch": 0.029506837961044342, + "grad_norm": 0.31255078315734863, + "learning_rate": 4.852880232076254e-06, + "loss": 0.6205, + "step": 356 + }, + { + "epoch": 0.029589722337339412, + "grad_norm": 0.31330594420433044, + "learning_rate": 4.852465810194779e-06, + "loss": 0.6157, + "step": 357 + }, + { + "epoch": 0.02967260671363448, + "grad_norm": 0.30091768503189087, + "learning_rate": 4.852051388313304e-06, + "loss": 0.6363, + "step": 358 + }, + { + "epoch": 0.02975549108992955, + "grad_norm": 0.2998180687427521, + "learning_rate": 4.851636966431828e-06, + "loss": 0.6147, + "step": 359 + }, + { + "epoch": 0.029838375466224618, + "grad_norm": 0.3030494153499603, + "learning_rate": 4.851222544550353e-06, + "loss": 0.5964, + "step": 360 + }, + { + "epoch": 0.029921259842519685, + "grad_norm": 0.2932775318622589, + "learning_rate": 4.850808122668877e-06, + "loss": 0.5818, + "step": 361 + }, + { + "epoch": 0.030004144218814754, + "grad_norm": 0.3140484690666199, + "learning_rate": 4.850393700787402e-06, + "loss": 0.6167, + "step": 362 + }, + { + "epoch": 0.03008702859510982, + "grad_norm": 0.3323778212070465, + "learning_rate": 4.849979278905927e-06, + "loss": 0.5844, + "step": 363 + }, + { + "epoch": 0.03016991297140489, + "grad_norm": 0.3219376802444458, + "learning_rate": 4.849564857024451e-06, + "loss": 0.6752, + "step": 364 + }, + { + "epoch": 0.030252797347699957, + "grad_norm": 0.4718853533267975, + "learning_rate": 4.849150435142976e-06, + "loss": 0.6083, + "step": 365 + }, + { + "epoch": 0.030335681723995027, + "grad_norm": 0.33655112981796265, + "learning_rate": 4.848736013261501e-06, + "loss": 0.6048, + "step": 366 + }, + { + "epoch": 0.030418566100290097, + "grad_norm": 0.29721203446388245, + "learning_rate": 4.8483215913800255e-06, + "loss": 0.5821, + "step": 367 + }, + { + "epoch": 0.030501450476585163, + "grad_norm": 0.3154633939266205, + "learning_rate": 4.84790716949855e-06, + "loss": 0.6123, + "step": 368 + }, + { + "epoch": 0.030584334852880233, + "grad_norm": 0.38699665665626526, + "learning_rate": 4.847492747617074e-06, + "loss": 0.5779, + "step": 369 + }, + { + "epoch": 0.0306672192291753, + "grad_norm": 0.2879536747932434, + "learning_rate": 4.847078325735599e-06, + "loss": 0.5886, + "step": 370 + }, + { + "epoch": 0.03075010360547037, + "grad_norm": 0.30335843563079834, + "learning_rate": 4.846663903854124e-06, + "loss": 0.6663, + "step": 371 + }, + { + "epoch": 0.030832987981765436, + "grad_norm": 0.29625287652015686, + "learning_rate": 4.8462494819726485e-06, + "loss": 0.6172, + "step": 372 + }, + { + "epoch": 0.030915872358060505, + "grad_norm": 0.28401675820350647, + "learning_rate": 4.845835060091173e-06, + "loss": 0.6227, + "step": 373 + }, + { + "epoch": 0.030998756734355575, + "grad_norm": 0.30357635021209717, + "learning_rate": 4.845420638209698e-06, + "loss": 0.6388, + "step": 374 + }, + { + "epoch": 0.03108164111065064, + "grad_norm": 0.28485170006752014, + "learning_rate": 4.845006216328223e-06, + "loss": 0.576, + "step": 375 + }, + { + "epoch": 0.03116452548694571, + "grad_norm": 0.28271275758743286, + "learning_rate": 4.844591794446747e-06, + "loss": 0.6133, + "step": 376 + }, + { + "epoch": 0.031247409863240778, + "grad_norm": 0.2855757772922516, + "learning_rate": 4.8441773725652715e-06, + "loss": 0.5336, + "step": 377 + }, + { + "epoch": 0.031330294239535844, + "grad_norm": 0.28049397468566895, + "learning_rate": 4.843762950683796e-06, + "loss": 0.602, + "step": 378 + }, + { + "epoch": 0.031413178615830914, + "grad_norm": 0.32106342911720276, + "learning_rate": 4.843348528802321e-06, + "loss": 0.5871, + "step": 379 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 0.3246700167655945, + "learning_rate": 4.842934106920846e-06, + "loss": 0.5619, + "step": 380 + }, + { + "epoch": 0.031578947368421054, + "grad_norm": 0.2797812223434448, + "learning_rate": 4.84251968503937e-06, + "loss": 0.5947, + "step": 381 + }, + { + "epoch": 0.031661831744716123, + "grad_norm": 0.2862568497657776, + "learning_rate": 4.842105263157895e-06, + "loss": 0.6172, + "step": 382 + }, + { + "epoch": 0.031744716121011186, + "grad_norm": 0.3507705628871918, + "learning_rate": 4.84169084127642e-06, + "loss": 0.6507, + "step": 383 + }, + { + "epoch": 0.031827600497306256, + "grad_norm": 0.29717594385147095, + "learning_rate": 4.841276419394945e-06, + "loss": 0.5767, + "step": 384 + }, + { + "epoch": 0.031910484873601326, + "grad_norm": 0.32763493061065674, + "learning_rate": 4.8408619975134695e-06, + "loss": 0.6541, + "step": 385 + }, + { + "epoch": 0.031993369249896396, + "grad_norm": 0.2796291410923004, + "learning_rate": 4.840447575631994e-06, + "loss": 0.5847, + "step": 386 + }, + { + "epoch": 0.032076253626191466, + "grad_norm": 0.2903553247451782, + "learning_rate": 4.840033153750518e-06, + "loss": 0.581, + "step": 387 + }, + { + "epoch": 0.03215913800248653, + "grad_norm": 0.3219704329967499, + "learning_rate": 4.839618731869043e-06, + "loss": 0.6076, + "step": 388 + }, + { + "epoch": 0.0322420223787816, + "grad_norm": 0.29998430609703064, + "learning_rate": 4.839204309987568e-06, + "loss": 0.6267, + "step": 389 + }, + { + "epoch": 0.03232490675507667, + "grad_norm": 0.3112008571624756, + "learning_rate": 4.8387898881060925e-06, + "loss": 0.5881, + "step": 390 + }, + { + "epoch": 0.03240779113137174, + "grad_norm": 0.28777971863746643, + "learning_rate": 4.838375466224617e-06, + "loss": 0.6105, + "step": 391 + }, + { + "epoch": 0.03249067550766681, + "grad_norm": 0.3269153833389282, + "learning_rate": 4.837961044343142e-06, + "loss": 0.6, + "step": 392 + }, + { + "epoch": 0.03257355988396187, + "grad_norm": 0.31380417943000793, + "learning_rate": 4.837546622461667e-06, + "loss": 0.6272, + "step": 393 + }, + { + "epoch": 0.03265644426025694, + "grad_norm": 0.30121591687202454, + "learning_rate": 4.837132200580191e-06, + "loss": 0.6209, + "step": 394 + }, + { + "epoch": 0.03273932863655201, + "grad_norm": 0.30607473850250244, + "learning_rate": 4.8367177786987155e-06, + "loss": 0.6322, + "step": 395 + }, + { + "epoch": 0.03282221301284708, + "grad_norm": 0.28972530364990234, + "learning_rate": 4.83630335681724e-06, + "loss": 0.5471, + "step": 396 + }, + { + "epoch": 0.03290509738914214, + "grad_norm": 0.3616604208946228, + "learning_rate": 4.835888934935765e-06, + "loss": 0.6033, + "step": 397 + }, + { + "epoch": 0.03298798176543721, + "grad_norm": 0.3226581811904907, + "learning_rate": 4.83547451305429e-06, + "loss": 0.6124, + "step": 398 + }, + { + "epoch": 0.03307086614173228, + "grad_norm": 0.2988527715206146, + "learning_rate": 4.835060091172814e-06, + "loss": 0.6266, + "step": 399 + }, + { + "epoch": 0.03315375051802735, + "grad_norm": 0.3277599513530731, + "learning_rate": 4.834645669291339e-06, + "loss": 0.5936, + "step": 400 + }, + { + "epoch": 0.03323663489432242, + "grad_norm": 0.327384352684021, + "learning_rate": 4.834231247409864e-06, + "loss": 0.6234, + "step": 401 + }, + { + "epoch": 0.033319519270617486, + "grad_norm": 0.29366305470466614, + "learning_rate": 4.833816825528388e-06, + "loss": 0.5729, + "step": 402 + }, + { + "epoch": 0.033402403646912555, + "grad_norm": 0.30287614464759827, + "learning_rate": 4.833402403646913e-06, + "loss": 0.6281, + "step": 403 + }, + { + "epoch": 0.033485288023207625, + "grad_norm": 0.28859275579452515, + "learning_rate": 4.832987981765437e-06, + "loss": 0.5778, + "step": 404 + }, + { + "epoch": 0.033568172399502695, + "grad_norm": 0.2870200574398041, + "learning_rate": 4.832573559883962e-06, + "loss": 0.5942, + "step": 405 + }, + { + "epoch": 0.033651056775797765, + "grad_norm": 0.3003169000148773, + "learning_rate": 4.832159138002487e-06, + "loss": 0.5794, + "step": 406 + }, + { + "epoch": 0.03373394115209283, + "grad_norm": 0.3087846338748932, + "learning_rate": 4.831744716121012e-06, + "loss": 0.5615, + "step": 407 + }, + { + "epoch": 0.0338168255283879, + "grad_norm": 0.3540176749229431, + "learning_rate": 4.8313302942395365e-06, + "loss": 0.6577, + "step": 408 + }, + { + "epoch": 0.03389970990468297, + "grad_norm": 0.27486833930015564, + "learning_rate": 4.830915872358061e-06, + "loss": 0.6229, + "step": 409 + }, + { + "epoch": 0.03398259428097804, + "grad_norm": 0.29682397842407227, + "learning_rate": 4.830501450476585e-06, + "loss": 0.6366, + "step": 410 + }, + { + "epoch": 0.03406547865727311, + "grad_norm": 0.3100220561027527, + "learning_rate": 4.83008702859511e-06, + "loss": 0.6705, + "step": 411 + }, + { + "epoch": 0.03414836303356817, + "grad_norm": 0.2926623821258545, + "learning_rate": 4.829672606713635e-06, + "loss": 0.5743, + "step": 412 + }, + { + "epoch": 0.03423124740986324, + "grad_norm": 0.3393244445323944, + "learning_rate": 4.8292581848321595e-06, + "loss": 0.6832, + "step": 413 + }, + { + "epoch": 0.03431413178615831, + "grad_norm": 0.3599708080291748, + "learning_rate": 4.828843762950684e-06, + "loss": 0.6313, + "step": 414 + }, + { + "epoch": 0.03439701616245338, + "grad_norm": 0.2997894287109375, + "learning_rate": 4.828429341069209e-06, + "loss": 0.5929, + "step": 415 + }, + { + "epoch": 0.03447990053874844, + "grad_norm": 0.3045388460159302, + "learning_rate": 4.828014919187734e-06, + "loss": 0.6069, + "step": 416 + }, + { + "epoch": 0.03456278491504351, + "grad_norm": 0.33078333735466003, + "learning_rate": 4.827600497306258e-06, + "loss": 0.5981, + "step": 417 + }, + { + "epoch": 0.03464566929133858, + "grad_norm": 0.30030298233032227, + "learning_rate": 4.8271860754247825e-06, + "loss": 0.585, + "step": 418 + }, + { + "epoch": 0.03472855366763365, + "grad_norm": 0.28453329205513, + "learning_rate": 4.826771653543308e-06, + "loss": 0.5798, + "step": 419 + }, + { + "epoch": 0.03481143804392872, + "grad_norm": 0.3277932405471802, + "learning_rate": 4.826357231661832e-06, + "loss": 0.6171, + "step": 420 + }, + { + "epoch": 0.034894322420223785, + "grad_norm": 0.2988298535346985, + "learning_rate": 4.825942809780357e-06, + "loss": 0.5969, + "step": 421 + }, + { + "epoch": 0.034977206796518855, + "grad_norm": 0.2770531177520752, + "learning_rate": 4.825528387898881e-06, + "loss": 0.5769, + "step": 422 + }, + { + "epoch": 0.035060091172813924, + "grad_norm": 0.3775043785572052, + "learning_rate": 4.825113966017406e-06, + "loss": 0.6135, + "step": 423 + }, + { + "epoch": 0.035142975549108994, + "grad_norm": 0.2991204261779785, + "learning_rate": 4.824699544135931e-06, + "loss": 0.6266, + "step": 424 + }, + { + "epoch": 0.035225859925404064, + "grad_norm": 0.29808318614959717, + "learning_rate": 4.824285122254456e-06, + "loss": 0.6188, + "step": 425 + }, + { + "epoch": 0.03530874430169913, + "grad_norm": 0.3075277805328369, + "learning_rate": 4.8238707003729805e-06, + "loss": 0.6064, + "step": 426 + }, + { + "epoch": 0.0353916286779942, + "grad_norm": 0.3068331182003021, + "learning_rate": 4.823456278491505e-06, + "loss": 0.6516, + "step": 427 + }, + { + "epoch": 0.03547451305428927, + "grad_norm": 0.338036447763443, + "learning_rate": 4.823041856610029e-06, + "loss": 0.6628, + "step": 428 + }, + { + "epoch": 0.03555739743058434, + "grad_norm": 0.3098353445529938, + "learning_rate": 4.822627434728554e-06, + "loss": 0.571, + "step": 429 + }, + { + "epoch": 0.035640281806879406, + "grad_norm": 0.29034554958343506, + "learning_rate": 4.822213012847079e-06, + "loss": 0.5984, + "step": 430 + }, + { + "epoch": 0.03572316618317447, + "grad_norm": 0.28866133093833923, + "learning_rate": 4.8217985909656035e-06, + "loss": 0.5645, + "step": 431 + }, + { + "epoch": 0.03580605055946954, + "grad_norm": 0.28175294399261475, + "learning_rate": 4.821384169084128e-06, + "loss": 0.5756, + "step": 432 + }, + { + "epoch": 0.03588893493576461, + "grad_norm": 0.2932020127773285, + "learning_rate": 4.820969747202653e-06, + "loss": 0.6116, + "step": 433 + }, + { + "epoch": 0.03597181931205968, + "grad_norm": 0.2918453812599182, + "learning_rate": 4.820555325321178e-06, + "loss": 0.6135, + "step": 434 + }, + { + "epoch": 0.03605470368835474, + "grad_norm": 0.27774935960769653, + "learning_rate": 4.820140903439702e-06, + "loss": 0.5822, + "step": 435 + }, + { + "epoch": 0.03613758806464981, + "grad_norm": 0.3003978133201599, + "learning_rate": 4.8197264815582265e-06, + "loss": 0.6183, + "step": 436 + }, + { + "epoch": 0.03622047244094488, + "grad_norm": 0.2974618077278137, + "learning_rate": 4.819312059676751e-06, + "loss": 0.5843, + "step": 437 + }, + { + "epoch": 0.03630335681723995, + "grad_norm": 0.2949097752571106, + "learning_rate": 4.818897637795276e-06, + "loss": 0.5793, + "step": 438 + }, + { + "epoch": 0.03638624119353502, + "grad_norm": 0.270171582698822, + "learning_rate": 4.818483215913801e-06, + "loss": 0.5762, + "step": 439 + }, + { + "epoch": 0.036469125569830084, + "grad_norm": 0.2933328449726105, + "learning_rate": 4.818068794032325e-06, + "loss": 0.5809, + "step": 440 + }, + { + "epoch": 0.036552009946125154, + "grad_norm": 0.3109627068042755, + "learning_rate": 4.81765437215085e-06, + "loss": 0.6242, + "step": 441 + }, + { + "epoch": 0.036634894322420224, + "grad_norm": 0.2919100522994995, + "learning_rate": 4.817239950269375e-06, + "loss": 0.6729, + "step": 442 + }, + { + "epoch": 0.036717778698715293, + "grad_norm": 0.27543893456459045, + "learning_rate": 4.816825528387899e-06, + "loss": 0.5982, + "step": 443 + }, + { + "epoch": 0.03680066307501036, + "grad_norm": 0.2785734534263611, + "learning_rate": 4.816411106506424e-06, + "loss": 0.5753, + "step": 444 + }, + { + "epoch": 0.036883547451305426, + "grad_norm": 0.297023206949234, + "learning_rate": 4.815996684624948e-06, + "loss": 0.6426, + "step": 445 + }, + { + "epoch": 0.036966431827600496, + "grad_norm": 0.2935127317905426, + "learning_rate": 4.815582262743473e-06, + "loss": 0.5668, + "step": 446 + }, + { + "epoch": 0.037049316203895566, + "grad_norm": 0.31613102555274963, + "learning_rate": 4.815167840861998e-06, + "loss": 0.6055, + "step": 447 + }, + { + "epoch": 0.037132200580190636, + "grad_norm": 0.29154452681541443, + "learning_rate": 4.814753418980523e-06, + "loss": 0.5789, + "step": 448 + }, + { + "epoch": 0.037215084956485706, + "grad_norm": 0.28714850544929504, + "learning_rate": 4.8143389970990475e-06, + "loss": 0.5532, + "step": 449 + }, + { + "epoch": 0.03729796933278077, + "grad_norm": 0.30648669600486755, + "learning_rate": 4.813924575217572e-06, + "loss": 0.633, + "step": 450 + }, + { + "epoch": 0.03738085370907584, + "grad_norm": 0.32399216294288635, + "learning_rate": 4.813510153336096e-06, + "loss": 0.6078, + "step": 451 + }, + { + "epoch": 0.03746373808537091, + "grad_norm": 0.303894966840744, + "learning_rate": 4.813095731454621e-06, + "loss": 0.6512, + "step": 452 + }, + { + "epoch": 0.03754662246166598, + "grad_norm": 0.28470051288604736, + "learning_rate": 4.812681309573145e-06, + "loss": 0.5841, + "step": 453 + }, + { + "epoch": 0.03762950683796104, + "grad_norm": 0.30272534489631653, + "learning_rate": 4.8122668876916705e-06, + "loss": 0.6248, + "step": 454 + }, + { + "epoch": 0.03771239121425611, + "grad_norm": 0.31582289934158325, + "learning_rate": 4.811852465810195e-06, + "loss": 0.6547, + "step": 455 + }, + { + "epoch": 0.03779527559055118, + "grad_norm": 0.2987618148326874, + "learning_rate": 4.81143804392872e-06, + "loss": 0.5952, + "step": 456 + }, + { + "epoch": 0.03787815996684625, + "grad_norm": 0.3239877223968506, + "learning_rate": 4.811023622047245e-06, + "loss": 0.621, + "step": 457 + }, + { + "epoch": 0.03796104434314132, + "grad_norm": 0.32741397619247437, + "learning_rate": 4.810609200165769e-06, + "loss": 0.5865, + "step": 458 + }, + { + "epoch": 0.03804392871943638, + "grad_norm": 0.286703497171402, + "learning_rate": 4.8101947782842935e-06, + "loss": 0.59, + "step": 459 + }, + { + "epoch": 0.03812681309573145, + "grad_norm": 0.2936321496963501, + "learning_rate": 4.809780356402819e-06, + "loss": 0.6355, + "step": 460 + }, + { + "epoch": 0.03820969747202652, + "grad_norm": 0.3076552450656891, + "learning_rate": 4.809365934521343e-06, + "loss": 0.5891, + "step": 461 + }, + { + "epoch": 0.03829258184832159, + "grad_norm": 0.28585752844810486, + "learning_rate": 4.808951512639868e-06, + "loss": 0.6026, + "step": 462 + }, + { + "epoch": 0.03837546622461666, + "grad_norm": 0.2727046608924866, + "learning_rate": 4.808537090758392e-06, + "loss": 0.544, + "step": 463 + }, + { + "epoch": 0.038458350600911725, + "grad_norm": 0.3382469117641449, + "learning_rate": 4.808122668876917e-06, + "loss": 0.6083, + "step": 464 + }, + { + "epoch": 0.038541234977206795, + "grad_norm": 0.2650943398475647, + "learning_rate": 4.807708246995442e-06, + "loss": 0.5314, + "step": 465 + }, + { + "epoch": 0.038624119353501865, + "grad_norm": 0.2746352255344391, + "learning_rate": 4.807293825113966e-06, + "loss": 0.5569, + "step": 466 + }, + { + "epoch": 0.038707003729796935, + "grad_norm": 0.6144485473632812, + "learning_rate": 4.8068794032324915e-06, + "loss": 0.5792, + "step": 467 + }, + { + "epoch": 0.038789888106092005, + "grad_norm": 0.2904563248157501, + "learning_rate": 4.806464981351016e-06, + "loss": 0.571, + "step": 468 + }, + { + "epoch": 0.03887277248238707, + "grad_norm": 0.28085461258888245, + "learning_rate": 4.80605055946954e-06, + "loss": 0.574, + "step": 469 + }, + { + "epoch": 0.03895565685868214, + "grad_norm": 0.2712056040763855, + "learning_rate": 4.805636137588065e-06, + "loss": 0.5789, + "step": 470 + }, + { + "epoch": 0.03903854123497721, + "grad_norm": 0.26822635531425476, + "learning_rate": 4.805221715706589e-06, + "loss": 0.5544, + "step": 471 + }, + { + "epoch": 0.03912142561127228, + "grad_norm": 0.3140382766723633, + "learning_rate": 4.8048072938251145e-06, + "loss": 0.6451, + "step": 472 + }, + { + "epoch": 0.03920430998756734, + "grad_norm": 0.287446528673172, + "learning_rate": 4.804392871943639e-06, + "loss": 0.5948, + "step": 473 + }, + { + "epoch": 0.03928719436386241, + "grad_norm": 0.2903457283973694, + "learning_rate": 4.803978450062164e-06, + "loss": 0.6115, + "step": 474 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 0.29095274209976196, + "learning_rate": 4.803564028180689e-06, + "loss": 0.5735, + "step": 475 + }, + { + "epoch": 0.03945296311645255, + "grad_norm": 0.3012527525424957, + "learning_rate": 4.803149606299213e-06, + "loss": 0.5563, + "step": 476 + }, + { + "epoch": 0.03953584749274762, + "grad_norm": 0.3317280113697052, + "learning_rate": 4.8027351844177375e-06, + "loss": 0.522, + "step": 477 + }, + { + "epoch": 0.03961873186904268, + "grad_norm": 0.33605796098709106, + "learning_rate": 4.802320762536262e-06, + "loss": 0.6053, + "step": 478 + }, + { + "epoch": 0.03970161624533775, + "grad_norm": 0.2998962104320526, + "learning_rate": 4.801906340654787e-06, + "loss": 0.5784, + "step": 479 + }, + { + "epoch": 0.03978450062163282, + "grad_norm": 0.3137480914592743, + "learning_rate": 4.801491918773312e-06, + "loss": 0.6216, + "step": 480 + }, + { + "epoch": 0.03986738499792789, + "grad_norm": 0.29823294281959534, + "learning_rate": 4.801077496891836e-06, + "loss": 0.6386, + "step": 481 + }, + { + "epoch": 0.03995026937422296, + "grad_norm": 0.27719417214393616, + "learning_rate": 4.800663075010361e-06, + "loss": 0.5952, + "step": 482 + }, + { + "epoch": 0.040033153750518025, + "grad_norm": 0.2955895960330963, + "learning_rate": 4.800248653128886e-06, + "loss": 0.6208, + "step": 483 + }, + { + "epoch": 0.040116038126813094, + "grad_norm": 0.2869635820388794, + "learning_rate": 4.79983423124741e-06, + "loss": 0.6198, + "step": 484 + }, + { + "epoch": 0.040198922503108164, + "grad_norm": 0.2871852219104767, + "learning_rate": 4.799419809365935e-06, + "loss": 0.6457, + "step": 485 + }, + { + "epoch": 0.040281806879403234, + "grad_norm": 0.25921034812927246, + "learning_rate": 4.799005387484459e-06, + "loss": 0.5862, + "step": 486 + }, + { + "epoch": 0.040364691255698304, + "grad_norm": 0.2841693162918091, + "learning_rate": 4.798590965602984e-06, + "loss": 0.5605, + "step": 487 + }, + { + "epoch": 0.04044757563199337, + "grad_norm": 0.28996825218200684, + "learning_rate": 4.798176543721509e-06, + "loss": 0.5837, + "step": 488 + }, + { + "epoch": 0.04053046000828844, + "grad_norm": 0.2926104664802551, + "learning_rate": 4.797762121840033e-06, + "loss": 0.6427, + "step": 489 + }, + { + "epoch": 0.04061334438458351, + "grad_norm": 0.32985857129096985, + "learning_rate": 4.7973476999585585e-06, + "loss": 0.588, + "step": 490 + }, + { + "epoch": 0.040696228760878576, + "grad_norm": 0.2935856878757477, + "learning_rate": 4.796933278077083e-06, + "loss": 0.5979, + "step": 491 + }, + { + "epoch": 0.040779113137173646, + "grad_norm": 0.3087214231491089, + "learning_rate": 4.796518856195607e-06, + "loss": 0.6003, + "step": 492 + }, + { + "epoch": 0.04086199751346871, + "grad_norm": 0.2761017084121704, + "learning_rate": 4.796104434314132e-06, + "loss": 0.5589, + "step": 493 + }, + { + "epoch": 0.04094488188976378, + "grad_norm": 0.31148457527160645, + "learning_rate": 4.795690012432656e-06, + "loss": 0.5899, + "step": 494 + }, + { + "epoch": 0.04102776626605885, + "grad_norm": 0.29657936096191406, + "learning_rate": 4.7952755905511815e-06, + "loss": 0.6038, + "step": 495 + }, + { + "epoch": 0.04111065064235392, + "grad_norm": 0.2751701772212982, + "learning_rate": 4.794861168669706e-06, + "loss": 0.5688, + "step": 496 + }, + { + "epoch": 0.04119353501864898, + "grad_norm": 0.37438392639160156, + "learning_rate": 4.794446746788231e-06, + "loss": 0.5206, + "step": 497 + }, + { + "epoch": 0.04127641939494405, + "grad_norm": 0.2836584150791168, + "learning_rate": 4.794032324906756e-06, + "loss": 0.5418, + "step": 498 + }, + { + "epoch": 0.04135930377123912, + "grad_norm": 0.30631956458091736, + "learning_rate": 4.79361790302528e-06, + "loss": 0.5975, + "step": 499 + }, + { + "epoch": 0.04144218814753419, + "grad_norm": 0.27450230717658997, + "learning_rate": 4.793203481143805e-06, + "loss": 0.5632, + "step": 500 + }, + { + "epoch": 0.04152507252382926, + "grad_norm": 0.3636972904205322, + "learning_rate": 4.79278905926233e-06, + "loss": 0.5801, + "step": 501 + }, + { + "epoch": 0.041607956900124324, + "grad_norm": 0.2846689224243164, + "learning_rate": 4.792374637380854e-06, + "loss": 0.5375, + "step": 502 + }, + { + "epoch": 0.041690841276419394, + "grad_norm": 0.3140423893928528, + "learning_rate": 4.791960215499379e-06, + "loss": 0.6149, + "step": 503 + }, + { + "epoch": 0.041773725652714463, + "grad_norm": 0.3084777593612671, + "learning_rate": 4.791545793617903e-06, + "loss": 0.5948, + "step": 504 + }, + { + "epoch": 0.04185661002900953, + "grad_norm": 0.2687724530696869, + "learning_rate": 4.791131371736428e-06, + "loss": 0.5704, + "step": 505 + }, + { + "epoch": 0.0419394944053046, + "grad_norm": 0.2888457775115967, + "learning_rate": 4.790716949854953e-06, + "loss": 0.6099, + "step": 506 + }, + { + "epoch": 0.042022378781599666, + "grad_norm": 0.30927976965904236, + "learning_rate": 4.790302527973477e-06, + "loss": 0.6537, + "step": 507 + }, + { + "epoch": 0.042105263157894736, + "grad_norm": 0.3024536967277527, + "learning_rate": 4.7898881060920025e-06, + "loss": 0.5636, + "step": 508 + }, + { + "epoch": 0.042188147534189806, + "grad_norm": 0.3090144693851471, + "learning_rate": 4.789473684210527e-06, + "loss": 0.5869, + "step": 509 + }, + { + "epoch": 0.042271031910484876, + "grad_norm": 0.3015652894973755, + "learning_rate": 4.789059262329051e-06, + "loss": 0.6497, + "step": 510 + }, + { + "epoch": 0.042353916286779945, + "grad_norm": 0.30465173721313477, + "learning_rate": 4.788644840447576e-06, + "loss": 0.6167, + "step": 511 + }, + { + "epoch": 0.04243680066307501, + "grad_norm": 0.2878759503364563, + "learning_rate": 4.7882304185661e-06, + "loss": 0.6118, + "step": 512 + }, + { + "epoch": 0.04251968503937008, + "grad_norm": 0.27850449085235596, + "learning_rate": 4.7878159966846255e-06, + "loss": 0.5673, + "step": 513 + }, + { + "epoch": 0.04260256941566515, + "grad_norm": 0.29175513982772827, + "learning_rate": 4.78740157480315e-06, + "loss": 0.6141, + "step": 514 + }, + { + "epoch": 0.04268545379196022, + "grad_norm": 0.3304097652435303, + "learning_rate": 4.786987152921675e-06, + "loss": 0.5958, + "step": 515 + }, + { + "epoch": 0.04276833816825528, + "grad_norm": 0.27558648586273193, + "learning_rate": 4.7865727310402e-06, + "loss": 0.5773, + "step": 516 + }, + { + "epoch": 0.04285122254455035, + "grad_norm": 0.28557613492012024, + "learning_rate": 4.786158309158724e-06, + "loss": 0.5859, + "step": 517 + }, + { + "epoch": 0.04293410692084542, + "grad_norm": 0.31380313634872437, + "learning_rate": 4.7857438872772485e-06, + "loss": 0.6497, + "step": 518 + }, + { + "epoch": 0.04301699129714049, + "grad_norm": 0.3501704931259155, + "learning_rate": 4.785329465395773e-06, + "loss": 0.595, + "step": 519 + }, + { + "epoch": 0.04309987567343556, + "grad_norm": 0.2842450737953186, + "learning_rate": 4.784915043514297e-06, + "loss": 0.5854, + "step": 520 + }, + { + "epoch": 0.04318276004973062, + "grad_norm": 0.2893671691417694, + "learning_rate": 4.784500621632823e-06, + "loss": 0.5424, + "step": 521 + }, + { + "epoch": 0.04326564442602569, + "grad_norm": 0.29267698526382446, + "learning_rate": 4.784086199751347e-06, + "loss": 0.6052, + "step": 522 + }, + { + "epoch": 0.04334852880232076, + "grad_norm": 0.2865925133228302, + "learning_rate": 4.783671777869872e-06, + "loss": 0.5921, + "step": 523 + }, + { + "epoch": 0.04343141317861583, + "grad_norm": 0.31299343705177307, + "learning_rate": 4.783257355988397e-06, + "loss": 0.6258, + "step": 524 + }, + { + "epoch": 0.0435142975549109, + "grad_norm": 0.2815278470516205, + "learning_rate": 4.782842934106921e-06, + "loss": 0.586, + "step": 525 + }, + { + "epoch": 0.043597181931205965, + "grad_norm": 0.2789231240749359, + "learning_rate": 4.782428512225446e-06, + "loss": 0.5617, + "step": 526 + }, + { + "epoch": 0.043680066307501035, + "grad_norm": 0.306830495595932, + "learning_rate": 4.78201409034397e-06, + "loss": 0.5649, + "step": 527 + }, + { + "epoch": 0.043762950683796105, + "grad_norm": 0.2792549431324005, + "learning_rate": 4.781599668462495e-06, + "loss": 0.6232, + "step": 528 + }, + { + "epoch": 0.043845835060091175, + "grad_norm": 0.3098222315311432, + "learning_rate": 4.78118524658102e-06, + "loss": 0.6219, + "step": 529 + }, + { + "epoch": 0.043928719436386245, + "grad_norm": 0.28481435775756836, + "learning_rate": 4.780770824699544e-06, + "loss": 0.6061, + "step": 530 + }, + { + "epoch": 0.04401160381268131, + "grad_norm": 0.2507385015487671, + "learning_rate": 4.7803564028180695e-06, + "loss": 0.5318, + "step": 531 + }, + { + "epoch": 0.04409448818897638, + "grad_norm": 0.33977583050727844, + "learning_rate": 4.779941980936594e-06, + "loss": 0.6655, + "step": 532 + }, + { + "epoch": 0.04417737256527145, + "grad_norm": 0.3023429811000824, + "learning_rate": 4.779527559055118e-06, + "loss": 0.6192, + "step": 533 + }, + { + "epoch": 0.04426025694156652, + "grad_norm": 0.286131888628006, + "learning_rate": 4.779113137173643e-06, + "loss": 0.5499, + "step": 534 + }, + { + "epoch": 0.04434314131786158, + "grad_norm": 0.29787036776542664, + "learning_rate": 4.778698715292167e-06, + "loss": 0.6437, + "step": 535 + }, + { + "epoch": 0.04442602569415665, + "grad_norm": 0.2988494634628296, + "learning_rate": 4.7782842934106925e-06, + "loss": 0.6108, + "step": 536 + }, + { + "epoch": 0.04450891007045172, + "grad_norm": 0.5352239608764648, + "learning_rate": 4.777869871529217e-06, + "loss": 0.5939, + "step": 537 + }, + { + "epoch": 0.04459179444674679, + "grad_norm": 0.31084302067756653, + "learning_rate": 4.777455449647741e-06, + "loss": 0.5446, + "step": 538 + }, + { + "epoch": 0.04467467882304186, + "grad_norm": 0.2634856700897217, + "learning_rate": 4.777041027766267e-06, + "loss": 0.5635, + "step": 539 + }, + { + "epoch": 0.04475756319933692, + "grad_norm": 0.2893992066383362, + "learning_rate": 4.776626605884791e-06, + "loss": 0.5624, + "step": 540 + }, + { + "epoch": 0.04484044757563199, + "grad_norm": 0.28337422013282776, + "learning_rate": 4.776212184003316e-06, + "loss": 0.6291, + "step": 541 + }, + { + "epoch": 0.04492333195192706, + "grad_norm": 0.2847951352596283, + "learning_rate": 4.775797762121841e-06, + "loss": 0.5899, + "step": 542 + }, + { + "epoch": 0.04500621632822213, + "grad_norm": 0.3260970413684845, + "learning_rate": 4.775383340240365e-06, + "loss": 0.6278, + "step": 543 + }, + { + "epoch": 0.0450891007045172, + "grad_norm": 0.3280145227909088, + "learning_rate": 4.77496891835889e-06, + "loss": 0.5993, + "step": 544 + }, + { + "epoch": 0.045171985080812264, + "grad_norm": 0.27709415555000305, + "learning_rate": 4.774554496477414e-06, + "loss": 0.5777, + "step": 545 + }, + { + "epoch": 0.045254869457107334, + "grad_norm": 0.276705265045166, + "learning_rate": 4.774140074595939e-06, + "loss": 0.5816, + "step": 546 + }, + { + "epoch": 0.045337753833402404, + "grad_norm": 0.2878621518611908, + "learning_rate": 4.773725652714464e-06, + "loss": 0.6022, + "step": 547 + }, + { + "epoch": 0.045420638209697474, + "grad_norm": 0.29199811816215515, + "learning_rate": 4.773311230832988e-06, + "loss": 0.5845, + "step": 548 + }, + { + "epoch": 0.045503522585992544, + "grad_norm": 0.31341320276260376, + "learning_rate": 4.7728968089515135e-06, + "loss": 0.6006, + "step": 549 + }, + { + "epoch": 0.04558640696228761, + "grad_norm": 0.28693369030952454, + "learning_rate": 4.772482387070038e-06, + "loss": 0.6031, + "step": 550 + }, + { + "epoch": 0.04566929133858268, + "grad_norm": 0.287815123796463, + "learning_rate": 4.772067965188562e-06, + "loss": 0.5629, + "step": 551 + }, + { + "epoch": 0.045752175714877746, + "grad_norm": 0.2793811857700348, + "learning_rate": 4.771653543307087e-06, + "loss": 0.5403, + "step": 552 + }, + { + "epoch": 0.045835060091172816, + "grad_norm": 0.28900647163391113, + "learning_rate": 4.771239121425611e-06, + "loss": 0.602, + "step": 553 + }, + { + "epoch": 0.04591794446746788, + "grad_norm": 0.2947494387626648, + "learning_rate": 4.7708246995441366e-06, + "loss": 0.5877, + "step": 554 + }, + { + "epoch": 0.04600082884376295, + "grad_norm": 0.3104610741138458, + "learning_rate": 4.770410277662661e-06, + "loss": 0.6439, + "step": 555 + }, + { + "epoch": 0.04608371322005802, + "grad_norm": 0.27342304587364197, + "learning_rate": 4.769995855781185e-06, + "loss": 0.5627, + "step": 556 + }, + { + "epoch": 0.04616659759635309, + "grad_norm": 0.301042377948761, + "learning_rate": 4.769581433899711e-06, + "loss": 0.5711, + "step": 557 + }, + { + "epoch": 0.04624948197264816, + "grad_norm": 0.31121426820755005, + "learning_rate": 4.769167012018235e-06, + "loss": 0.5518, + "step": 558 + }, + { + "epoch": 0.04633236634894322, + "grad_norm": 0.3043922483921051, + "learning_rate": 4.7687525901367596e-06, + "loss": 0.5964, + "step": 559 + }, + { + "epoch": 0.04641525072523829, + "grad_norm": 0.2939624488353729, + "learning_rate": 4.768338168255284e-06, + "loss": 0.6124, + "step": 560 + }, + { + "epoch": 0.04649813510153336, + "grad_norm": 0.2803196609020233, + "learning_rate": 4.767923746373808e-06, + "loss": 0.5849, + "step": 561 + }, + { + "epoch": 0.04658101947782843, + "grad_norm": 0.31197136640548706, + "learning_rate": 4.767509324492334e-06, + "loss": 0.5915, + "step": 562 + }, + { + "epoch": 0.0466639038541235, + "grad_norm": 0.3189193904399872, + "learning_rate": 4.767094902610858e-06, + "loss": 0.623, + "step": 563 + }, + { + "epoch": 0.046746788230418564, + "grad_norm": 0.2790443003177643, + "learning_rate": 4.766680480729383e-06, + "loss": 0.5676, + "step": 564 + }, + { + "epoch": 0.04682967260671363, + "grad_norm": 0.2832474112510681, + "learning_rate": 4.766266058847908e-06, + "loss": 0.5373, + "step": 565 + }, + { + "epoch": 0.0469125569830087, + "grad_norm": 0.2638501226902008, + "learning_rate": 4.765851636966432e-06, + "loss": 0.594, + "step": 566 + }, + { + "epoch": 0.04699544135930377, + "grad_norm": 0.274636447429657, + "learning_rate": 4.765437215084957e-06, + "loss": 0.5637, + "step": 567 + }, + { + "epoch": 0.04707832573559884, + "grad_norm": 0.30230262875556946, + "learning_rate": 4.765022793203481e-06, + "loss": 0.5857, + "step": 568 + }, + { + "epoch": 0.047161210111893906, + "grad_norm": 0.30139878392219543, + "learning_rate": 4.7646083713220056e-06, + "loss": 0.5706, + "step": 569 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 0.2995483875274658, + "learning_rate": 4.764193949440531e-06, + "loss": 0.6558, + "step": 570 + }, + { + "epoch": 0.047326978864484046, + "grad_norm": 0.29473790526390076, + "learning_rate": 4.763779527559055e-06, + "loss": 0.6199, + "step": 571 + }, + { + "epoch": 0.047409863240779115, + "grad_norm": 0.32881560921669006, + "learning_rate": 4.7633651056775806e-06, + "loss": 0.6158, + "step": 572 + }, + { + "epoch": 0.04749274761707418, + "grad_norm": 0.29223722219467163, + "learning_rate": 4.762950683796105e-06, + "loss": 0.6183, + "step": 573 + }, + { + "epoch": 0.04757563199336925, + "grad_norm": 0.30851301550865173, + "learning_rate": 4.762536261914629e-06, + "loss": 0.6105, + "step": 574 + }, + { + "epoch": 0.04765851636966432, + "grad_norm": 0.3168797791004181, + "learning_rate": 4.762121840033154e-06, + "loss": 0.6068, + "step": 575 + }, + { + "epoch": 0.04774140074595939, + "grad_norm": 0.2699555456638336, + "learning_rate": 4.761707418151678e-06, + "loss": 0.5744, + "step": 576 + }, + { + "epoch": 0.04782428512225446, + "grad_norm": 0.28627485036849976, + "learning_rate": 4.7612929962702036e-06, + "loss": 0.5839, + "step": 577 + }, + { + "epoch": 0.04790716949854952, + "grad_norm": 0.29764559864997864, + "learning_rate": 4.760878574388728e-06, + "loss": 0.6066, + "step": 578 + }, + { + "epoch": 0.04799005387484459, + "grad_norm": 0.2730076313018799, + "learning_rate": 4.7604641525072524e-06, + "loss": 0.6034, + "step": 579 + }, + { + "epoch": 0.04807293825113966, + "grad_norm": 0.2829475700855255, + "learning_rate": 4.760049730625778e-06, + "loss": 0.5806, + "step": 580 + }, + { + "epoch": 0.04815582262743473, + "grad_norm": 0.2871692478656769, + "learning_rate": 4.759635308744302e-06, + "loss": 0.6016, + "step": 581 + }, + { + "epoch": 0.0482387070037298, + "grad_norm": 0.2894265353679657, + "learning_rate": 4.759220886862827e-06, + "loss": 0.5883, + "step": 582 + }, + { + "epoch": 0.04832159138002486, + "grad_norm": 0.26680392026901245, + "learning_rate": 4.758806464981352e-06, + "loss": 0.5688, + "step": 583 + }, + { + "epoch": 0.04840447575631993, + "grad_norm": 0.2938423454761505, + "learning_rate": 4.758392043099876e-06, + "loss": 0.6264, + "step": 584 + }, + { + "epoch": 0.048487360132615, + "grad_norm": 0.31537848711013794, + "learning_rate": 4.757977621218401e-06, + "loss": 0.5679, + "step": 585 + }, + { + "epoch": 0.04857024450891007, + "grad_norm": 0.3046777844429016, + "learning_rate": 4.757563199336925e-06, + "loss": 0.5358, + "step": 586 + }, + { + "epoch": 0.04865312888520514, + "grad_norm": 0.269679456949234, + "learning_rate": 4.75714877745545e-06, + "loss": 0.5502, + "step": 587 + }, + { + "epoch": 0.048736013261500205, + "grad_norm": 0.2846774756908417, + "learning_rate": 4.756734355573975e-06, + "loss": 0.5333, + "step": 588 + }, + { + "epoch": 0.048818897637795275, + "grad_norm": 0.3204803764820099, + "learning_rate": 4.756319933692499e-06, + "loss": 0.6057, + "step": 589 + }, + { + "epoch": 0.048901782014090345, + "grad_norm": 0.2890942096710205, + "learning_rate": 4.7559055118110246e-06, + "loss": 0.5721, + "step": 590 + }, + { + "epoch": 0.048984666390385415, + "grad_norm": 0.3323914110660553, + "learning_rate": 4.755491089929549e-06, + "loss": 0.6049, + "step": 591 + }, + { + "epoch": 0.04906755076668048, + "grad_norm": 0.27530550956726074, + "learning_rate": 4.7550766680480734e-06, + "loss": 0.603, + "step": 592 + }, + { + "epoch": 0.04915043514297555, + "grad_norm": 0.3696022927761078, + "learning_rate": 4.754662246166598e-06, + "loss": 0.5789, + "step": 593 + }, + { + "epoch": 0.04923331951927062, + "grad_norm": 0.31728190183639526, + "learning_rate": 4.754247824285122e-06, + "loss": 0.579, + "step": 594 + }, + { + "epoch": 0.04931620389556569, + "grad_norm": 0.4074658751487732, + "learning_rate": 4.7538334024036476e-06, + "loss": 0.6252, + "step": 595 + }, + { + "epoch": 0.04939908827186076, + "grad_norm": 0.2788049280643463, + "learning_rate": 4.753418980522172e-06, + "loss": 0.566, + "step": 596 + }, + { + "epoch": 0.04948197264815582, + "grad_norm": 0.2875158190727234, + "learning_rate": 4.7530045586406964e-06, + "loss": 0.6074, + "step": 597 + }, + { + "epoch": 0.04956485702445089, + "grad_norm": 0.3081792891025543, + "learning_rate": 4.752590136759222e-06, + "loss": 0.5968, + "step": 598 + }, + { + "epoch": 0.04964774140074596, + "grad_norm": 0.2934686243534088, + "learning_rate": 4.752175714877746e-06, + "loss": 0.6098, + "step": 599 + }, + { + "epoch": 0.04973062577704103, + "grad_norm": 0.31635573506355286, + "learning_rate": 4.751761292996271e-06, + "loss": 0.6047, + "step": 600 + }, + { + "epoch": 0.0498135101533361, + "grad_norm": 0.4332502782344818, + "learning_rate": 4.751346871114795e-06, + "loss": 0.6226, + "step": 601 + }, + { + "epoch": 0.04989639452963116, + "grad_norm": 0.28606709837913513, + "learning_rate": 4.7509324492333194e-06, + "loss": 0.5726, + "step": 602 + }, + { + "epoch": 0.04997927890592623, + "grad_norm": 0.27568987011909485, + "learning_rate": 4.750518027351845e-06, + "loss": 0.5872, + "step": 603 + }, + { + "epoch": 0.0500621632822213, + "grad_norm": 0.2627095580101013, + "learning_rate": 4.750103605470369e-06, + "loss": 0.5334, + "step": 604 + }, + { + "epoch": 0.05014504765851637, + "grad_norm": 0.30160605907440186, + "learning_rate": 4.749689183588894e-06, + "loss": 0.5875, + "step": 605 + }, + { + "epoch": 0.05022793203481144, + "grad_norm": 0.30656933784484863, + "learning_rate": 4.749274761707419e-06, + "loss": 0.6016, + "step": 606 + }, + { + "epoch": 0.050310816411106504, + "grad_norm": 0.29025107622146606, + "learning_rate": 4.748860339825943e-06, + "loss": 0.5926, + "step": 607 + }, + { + "epoch": 0.050393700787401574, + "grad_norm": 0.2692035734653473, + "learning_rate": 4.748445917944468e-06, + "loss": 0.5464, + "step": 608 + }, + { + "epoch": 0.050476585163696644, + "grad_norm": 0.28356367349624634, + "learning_rate": 4.748031496062992e-06, + "loss": 0.5788, + "step": 609 + }, + { + "epoch": 0.050559469539991714, + "grad_norm": 0.30700305104255676, + "learning_rate": 4.747617074181517e-06, + "loss": 0.5967, + "step": 610 + }, + { + "epoch": 0.05064235391628678, + "grad_norm": 0.28306689858436584, + "learning_rate": 4.747202652300042e-06, + "loss": 0.5646, + "step": 611 + }, + { + "epoch": 0.050725238292581847, + "grad_norm": 0.2880675196647644, + "learning_rate": 4.746788230418566e-06, + "loss": 0.6055, + "step": 612 + }, + { + "epoch": 0.050808122668876916, + "grad_norm": 0.28912782669067383, + "learning_rate": 4.746373808537092e-06, + "loss": 0.5905, + "step": 613 + }, + { + "epoch": 0.050891007045171986, + "grad_norm": 0.2981110215187073, + "learning_rate": 4.745959386655616e-06, + "loss": 0.584, + "step": 614 + }, + { + "epoch": 0.050973891421467056, + "grad_norm": 0.30755335092544556, + "learning_rate": 4.7455449647741404e-06, + "loss": 0.6214, + "step": 615 + }, + { + "epoch": 0.05105677579776212, + "grad_norm": 0.2892230749130249, + "learning_rate": 4.745130542892665e-06, + "loss": 0.5846, + "step": 616 + }, + { + "epoch": 0.05113966017405719, + "grad_norm": 0.28250688314437866, + "learning_rate": 4.744716121011189e-06, + "loss": 0.5708, + "step": 617 + }, + { + "epoch": 0.05122254455035226, + "grad_norm": 0.28505319356918335, + "learning_rate": 4.744301699129715e-06, + "loss": 0.5623, + "step": 618 + }, + { + "epoch": 0.05130542892664733, + "grad_norm": 0.28148406744003296, + "learning_rate": 4.743887277248239e-06, + "loss": 0.5592, + "step": 619 + }, + { + "epoch": 0.0513883133029424, + "grad_norm": 0.28334563970565796, + "learning_rate": 4.7434728553667635e-06, + "loss": 0.5712, + "step": 620 + }, + { + "epoch": 0.05147119767923746, + "grad_norm": 0.27196812629699707, + "learning_rate": 4.743058433485289e-06, + "loss": 0.5767, + "step": 621 + }, + { + "epoch": 0.05155408205553253, + "grad_norm": 0.2702648341655731, + "learning_rate": 4.742644011603813e-06, + "loss": 0.5908, + "step": 622 + }, + { + "epoch": 0.0516369664318276, + "grad_norm": 0.33183133602142334, + "learning_rate": 4.742229589722338e-06, + "loss": 0.6093, + "step": 623 + }, + { + "epoch": 0.05171985080812267, + "grad_norm": 0.27887508273124695, + "learning_rate": 4.741815167840863e-06, + "loss": 0.5897, + "step": 624 + }, + { + "epoch": 0.05180273518441774, + "grad_norm": 0.29368123412132263, + "learning_rate": 4.741400745959387e-06, + "loss": 0.5727, + "step": 625 + }, + { + "epoch": 0.0518856195607128, + "grad_norm": 0.2809799909591675, + "learning_rate": 4.740986324077912e-06, + "loss": 0.5372, + "step": 626 + }, + { + "epoch": 0.05196850393700787, + "grad_norm": 0.3143068552017212, + "learning_rate": 4.740571902196436e-06, + "loss": 0.6124, + "step": 627 + }, + { + "epoch": 0.05205138831330294, + "grad_norm": 0.28145888447761536, + "learning_rate": 4.740157480314961e-06, + "loss": 0.5754, + "step": 628 + }, + { + "epoch": 0.05213427268959801, + "grad_norm": 0.2878841459751129, + "learning_rate": 4.739743058433486e-06, + "loss": 0.647, + "step": 629 + }, + { + "epoch": 0.052217157065893076, + "grad_norm": 0.29968464374542236, + "learning_rate": 4.73932863655201e-06, + "loss": 0.6371, + "step": 630 + }, + { + "epoch": 0.052300041442188146, + "grad_norm": 0.2979857325553894, + "learning_rate": 4.738914214670536e-06, + "loss": 0.6008, + "step": 631 + }, + { + "epoch": 0.052382925818483216, + "grad_norm": 0.29787713289260864, + "learning_rate": 4.73849979278906e-06, + "loss": 0.5647, + "step": 632 + }, + { + "epoch": 0.052465810194778285, + "grad_norm": 0.2878720462322235, + "learning_rate": 4.7380853709075845e-06, + "loss": 0.6069, + "step": 633 + }, + { + "epoch": 0.052548694571073355, + "grad_norm": 0.2838108241558075, + "learning_rate": 4.737670949026109e-06, + "loss": 0.5825, + "step": 634 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 0.2760535776615143, + "learning_rate": 4.737256527144633e-06, + "loss": 0.64, + "step": 635 + }, + { + "epoch": 0.05271446332366349, + "grad_norm": 0.29690688848495483, + "learning_rate": 4.736842105263158e-06, + "loss": 0.6094, + "step": 636 + }, + { + "epoch": 0.05279734769995856, + "grad_norm": 0.2988104224205017, + "learning_rate": 4.736427683381683e-06, + "loss": 0.6178, + "step": 637 + }, + { + "epoch": 0.05288023207625363, + "grad_norm": 0.36195307970046997, + "learning_rate": 4.7360132615002075e-06, + "loss": 0.6035, + "step": 638 + }, + { + "epoch": 0.0529631164525487, + "grad_norm": 0.2988924980163574, + "learning_rate": 4.735598839618733e-06, + "loss": 0.6297, + "step": 639 + }, + { + "epoch": 0.05304600082884376, + "grad_norm": 0.28392377495765686, + "learning_rate": 4.735184417737257e-06, + "loss": 0.6208, + "step": 640 + }, + { + "epoch": 0.05312888520513883, + "grad_norm": 0.29688411951065063, + "learning_rate": 4.734769995855782e-06, + "loss": 0.5927, + "step": 641 + }, + { + "epoch": 0.0532117695814339, + "grad_norm": 0.27135735750198364, + "learning_rate": 4.734355573974306e-06, + "loss": 0.5855, + "step": 642 + }, + { + "epoch": 0.05329465395772897, + "grad_norm": 0.3031463325023651, + "learning_rate": 4.7339411520928305e-06, + "loss": 0.6212, + "step": 643 + }, + { + "epoch": 0.05337753833402404, + "grad_norm": 0.31926149129867554, + "learning_rate": 4.733526730211356e-06, + "loss": 0.5967, + "step": 644 + }, + { + "epoch": 0.0534604227103191, + "grad_norm": 0.2780073881149292, + "learning_rate": 4.73311230832988e-06, + "loss": 0.5789, + "step": 645 + }, + { + "epoch": 0.05354330708661417, + "grad_norm": 0.2806304693222046, + "learning_rate": 4.732697886448405e-06, + "loss": 0.5772, + "step": 646 + }, + { + "epoch": 0.05362619146290924, + "grad_norm": 0.2919198274612427, + "learning_rate": 4.73228346456693e-06, + "loss": 0.6224, + "step": 647 + }, + { + "epoch": 0.05370907583920431, + "grad_norm": 0.28547602891921997, + "learning_rate": 4.731869042685454e-06, + "loss": 0.6125, + "step": 648 + }, + { + "epoch": 0.053791960215499375, + "grad_norm": 0.27969181537628174, + "learning_rate": 4.731454620803979e-06, + "loss": 0.5839, + "step": 649 + }, + { + "epoch": 0.053874844591794445, + "grad_norm": 0.3313753604888916, + "learning_rate": 4.731040198922503e-06, + "loss": 0.6095, + "step": 650 + }, + { + "epoch": 0.053957728968089515, + "grad_norm": 0.3085145652294159, + "learning_rate": 4.730625777041028e-06, + "loss": 0.5855, + "step": 651 + }, + { + "epoch": 0.054040613344384585, + "grad_norm": 0.30488210916519165, + "learning_rate": 4.730211355159553e-06, + "loss": 0.5388, + "step": 652 + }, + { + "epoch": 0.054123497720679654, + "grad_norm": 0.2849658131599426, + "learning_rate": 4.729796933278077e-06, + "loss": 0.6047, + "step": 653 + }, + { + "epoch": 0.05420638209697472, + "grad_norm": 0.3015500009059906, + "learning_rate": 4.729382511396602e-06, + "loss": 0.6429, + "step": 654 + }, + { + "epoch": 0.05428926647326979, + "grad_norm": 0.3005625605583191, + "learning_rate": 4.728968089515127e-06, + "loss": 0.6055, + "step": 655 + }, + { + "epoch": 0.05437215084956486, + "grad_norm": 0.30080413818359375, + "learning_rate": 4.7285536676336515e-06, + "loss": 0.5519, + "step": 656 + }, + { + "epoch": 0.05445503522585993, + "grad_norm": 0.2865578830242157, + "learning_rate": 4.728139245752176e-06, + "loss": 0.5755, + "step": 657 + }, + { + "epoch": 0.054537919602155, + "grad_norm": 0.30484580993652344, + "learning_rate": 4.727724823870701e-06, + "loss": 0.598, + "step": 658 + }, + { + "epoch": 0.05462080397845006, + "grad_norm": 0.27940499782562256, + "learning_rate": 4.727310401989226e-06, + "loss": 0.5914, + "step": 659 + }, + { + "epoch": 0.05470368835474513, + "grad_norm": 0.3197372555732727, + "learning_rate": 4.72689598010775e-06, + "loss": 0.6132, + "step": 660 + }, + { + "epoch": 0.0547865727310402, + "grad_norm": 0.27767205238342285, + "learning_rate": 4.7264815582262745e-06, + "loss": 0.5807, + "step": 661 + }, + { + "epoch": 0.05486945710733527, + "grad_norm": 0.27581068873405457, + "learning_rate": 4.7260671363448e-06, + "loss": 0.5704, + "step": 662 + }, + { + "epoch": 0.05495234148363034, + "grad_norm": 0.28246158361434937, + "learning_rate": 4.725652714463324e-06, + "loss": 0.5905, + "step": 663 + }, + { + "epoch": 0.0550352258599254, + "grad_norm": 0.3189125061035156, + "learning_rate": 4.725238292581849e-06, + "loss": 0.6606, + "step": 664 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 0.2783942222595215, + "learning_rate": 4.724823870700374e-06, + "loss": 0.6161, + "step": 665 + }, + { + "epoch": 0.05520099461251554, + "grad_norm": 0.31047749519348145, + "learning_rate": 4.724409448818898e-06, + "loss": 0.6255, + "step": 666 + }, + { + "epoch": 0.05528387898881061, + "grad_norm": 0.2932157516479492, + "learning_rate": 4.723995026937423e-06, + "loss": 0.6077, + "step": 667 + }, + { + "epoch": 0.055366763365105674, + "grad_norm": 0.2810492515563965, + "learning_rate": 4.723580605055947e-06, + "loss": 0.5339, + "step": 668 + }, + { + "epoch": 0.055449647741400744, + "grad_norm": 0.30648115277290344, + "learning_rate": 4.723166183174472e-06, + "loss": 0.6272, + "step": 669 + }, + { + "epoch": 0.055532532117695814, + "grad_norm": 0.2825585901737213, + "learning_rate": 4.722751761292997e-06, + "loss": 0.6042, + "step": 670 + }, + { + "epoch": 0.055615416493990884, + "grad_norm": 0.28231656551361084, + "learning_rate": 4.722337339411521e-06, + "loss": 0.6201, + "step": 671 + }, + { + "epoch": 0.055698300870285954, + "grad_norm": 0.29449692368507385, + "learning_rate": 4.721922917530046e-06, + "loss": 0.5795, + "step": 672 + }, + { + "epoch": 0.055781185246581017, + "grad_norm": 0.2750587463378906, + "learning_rate": 4.721508495648571e-06, + "loss": 0.5803, + "step": 673 + }, + { + "epoch": 0.055864069622876086, + "grad_norm": 0.28550606966018677, + "learning_rate": 4.7210940737670955e-06, + "loss": 0.5727, + "step": 674 + }, + { + "epoch": 0.055946953999171156, + "grad_norm": 0.2701563537120819, + "learning_rate": 4.72067965188562e-06, + "loss": 0.5721, + "step": 675 + }, + { + "epoch": 0.056029838375466226, + "grad_norm": 0.28714612126350403, + "learning_rate": 4.720265230004144e-06, + "loss": 0.6018, + "step": 676 + }, + { + "epoch": 0.056112722751761296, + "grad_norm": 0.2727581858634949, + "learning_rate": 4.719850808122669e-06, + "loss": 0.5545, + "step": 677 + }, + { + "epoch": 0.05619560712805636, + "grad_norm": 0.268539160490036, + "learning_rate": 4.719436386241194e-06, + "loss": 0.5717, + "step": 678 + }, + { + "epoch": 0.05627849150435143, + "grad_norm": 0.26155027747154236, + "learning_rate": 4.7190219643597185e-06, + "loss": 0.5462, + "step": 679 + }, + { + "epoch": 0.0563613758806465, + "grad_norm": 0.2979479134082794, + "learning_rate": 4.718607542478244e-06, + "loss": 0.6244, + "step": 680 + }, + { + "epoch": 0.05644426025694157, + "grad_norm": 0.2811577022075653, + "learning_rate": 4.718193120596768e-06, + "loss": 0.5875, + "step": 681 + }, + { + "epoch": 0.05652714463323664, + "grad_norm": 0.2753483057022095, + "learning_rate": 4.717778698715293e-06, + "loss": 0.5748, + "step": 682 + }, + { + "epoch": 0.0566100290095317, + "grad_norm": 0.2816849946975708, + "learning_rate": 4.717364276833817e-06, + "loss": 0.5777, + "step": 683 + }, + { + "epoch": 0.05669291338582677, + "grad_norm": 0.28655385971069336, + "learning_rate": 4.7169498549523415e-06, + "loss": 0.6067, + "step": 684 + }, + { + "epoch": 0.05677579776212184, + "grad_norm": 0.29260894656181335, + "learning_rate": 4.716535433070866e-06, + "loss": 0.5848, + "step": 685 + }, + { + "epoch": 0.05685868213841691, + "grad_norm": 0.2643453776836395, + "learning_rate": 4.716121011189391e-06, + "loss": 0.6089, + "step": 686 + }, + { + "epoch": 0.05694156651471197, + "grad_norm": 0.26736336946487427, + "learning_rate": 4.715706589307916e-06, + "loss": 0.5959, + "step": 687 + }, + { + "epoch": 0.05702445089100704, + "grad_norm": 0.2886108160018921, + "learning_rate": 4.715292167426441e-06, + "loss": 0.5904, + "step": 688 + }, + { + "epoch": 0.05710733526730211, + "grad_norm": 0.3401309549808502, + "learning_rate": 4.714877745544965e-06, + "loss": 0.6046, + "step": 689 + }, + { + "epoch": 0.05719021964359718, + "grad_norm": 0.2635171413421631, + "learning_rate": 4.71446332366349e-06, + "loss": 0.5432, + "step": 690 + }, + { + "epoch": 0.05727310401989225, + "grad_norm": 0.2854808270931244, + "learning_rate": 4.714048901782014e-06, + "loss": 0.6021, + "step": 691 + }, + { + "epoch": 0.057355988396187316, + "grad_norm": 0.28850722312927246, + "learning_rate": 4.713634479900539e-06, + "loss": 0.5769, + "step": 692 + }, + { + "epoch": 0.057438872772482386, + "grad_norm": 0.29525479674339294, + "learning_rate": 4.713220058019064e-06, + "loss": 0.621, + "step": 693 + }, + { + "epoch": 0.057521757148777455, + "grad_norm": 0.27906334400177, + "learning_rate": 4.712805636137588e-06, + "loss": 0.5848, + "step": 694 + }, + { + "epoch": 0.057604641525072525, + "grad_norm": 0.28389859199523926, + "learning_rate": 4.712391214256113e-06, + "loss": 0.6222, + "step": 695 + }, + { + "epoch": 0.057687525901367595, + "grad_norm": 0.2901958227157593, + "learning_rate": 4.711976792374638e-06, + "loss": 0.5698, + "step": 696 + }, + { + "epoch": 0.05777041027766266, + "grad_norm": 0.27840450406074524, + "learning_rate": 4.7115623704931625e-06, + "loss": 0.5823, + "step": 697 + }, + { + "epoch": 0.05785329465395773, + "grad_norm": 0.3001241385936737, + "learning_rate": 4.711147948611687e-06, + "loss": 0.6164, + "step": 698 + }, + { + "epoch": 0.0579361790302528, + "grad_norm": 0.27966898679733276, + "learning_rate": 4.710733526730212e-06, + "loss": 0.5688, + "step": 699 + }, + { + "epoch": 0.05801906340654787, + "grad_norm": 0.2837158143520355, + "learning_rate": 4.710319104848737e-06, + "loss": 0.5562, + "step": 700 + }, + { + "epoch": 0.05810194778284294, + "grad_norm": 0.29079553484916687, + "learning_rate": 4.709904682967261e-06, + "loss": 0.5926, + "step": 701 + }, + { + "epoch": 0.058184832159138, + "grad_norm": 0.31498923897743225, + "learning_rate": 4.7094902610857855e-06, + "loss": 0.6477, + "step": 702 + }, + { + "epoch": 0.05826771653543307, + "grad_norm": 0.2968491017818451, + "learning_rate": 4.70907583920431e-06, + "loss": 0.595, + "step": 703 + }, + { + "epoch": 0.05835060091172814, + "grad_norm": 0.29314061999320984, + "learning_rate": 4.708661417322835e-06, + "loss": 0.587, + "step": 704 + }, + { + "epoch": 0.05843348528802321, + "grad_norm": 0.2855537235736847, + "learning_rate": 4.70824699544136e-06, + "loss": 0.5684, + "step": 705 + }, + { + "epoch": 0.05851636966431827, + "grad_norm": 0.28321772813796997, + "learning_rate": 4.707832573559885e-06, + "loss": 0.5997, + "step": 706 + }, + { + "epoch": 0.05859925404061334, + "grad_norm": 0.2829386293888092, + "learning_rate": 4.707418151678409e-06, + "loss": 0.5823, + "step": 707 + }, + { + "epoch": 0.05868213841690841, + "grad_norm": 0.33031824231147766, + "learning_rate": 4.707003729796934e-06, + "loss": 0.6021, + "step": 708 + }, + { + "epoch": 0.05876502279320348, + "grad_norm": 0.27947840094566345, + "learning_rate": 4.706589307915458e-06, + "loss": 0.5681, + "step": 709 + }, + { + "epoch": 0.05884790716949855, + "grad_norm": 0.3129144608974457, + "learning_rate": 4.706174886033983e-06, + "loss": 0.5959, + "step": 710 + }, + { + "epoch": 0.058930791545793615, + "grad_norm": 0.31047725677490234, + "learning_rate": 4.705760464152508e-06, + "loss": 0.5863, + "step": 711 + }, + { + "epoch": 0.059013675922088685, + "grad_norm": 0.2833392918109894, + "learning_rate": 4.705346042271032e-06, + "loss": 0.5814, + "step": 712 + }, + { + "epoch": 0.059096560298383755, + "grad_norm": 0.3248955309391022, + "learning_rate": 4.704931620389557e-06, + "loss": 0.5839, + "step": 713 + }, + { + "epoch": 0.059179444674678824, + "grad_norm": 0.28459981083869934, + "learning_rate": 4.704517198508082e-06, + "loss": 0.5943, + "step": 714 + }, + { + "epoch": 0.059262329050973894, + "grad_norm": 0.298292875289917, + "learning_rate": 4.7041027766266065e-06, + "loss": 0.5869, + "step": 715 + }, + { + "epoch": 0.05934521342726896, + "grad_norm": 0.30408865213394165, + "learning_rate": 4.703688354745131e-06, + "loss": 0.6295, + "step": 716 + }, + { + "epoch": 0.05942809780356403, + "grad_norm": 0.29634520411491394, + "learning_rate": 4.703273932863655e-06, + "loss": 0.5936, + "step": 717 + }, + { + "epoch": 0.0595109821798591, + "grad_norm": 0.3025098741054535, + "learning_rate": 4.70285951098218e-06, + "loss": 0.6176, + "step": 718 + }, + { + "epoch": 0.05959386655615417, + "grad_norm": 0.28513050079345703, + "learning_rate": 4.702445089100705e-06, + "loss": 0.5028, + "step": 719 + }, + { + "epoch": 0.059676750932449237, + "grad_norm": 0.28136369585990906, + "learning_rate": 4.7020306672192295e-06, + "loss": 0.5715, + "step": 720 + }, + { + "epoch": 0.0597596353087443, + "grad_norm": 0.29227086901664734, + "learning_rate": 4.701616245337754e-06, + "loss": 0.5453, + "step": 721 + }, + { + "epoch": 0.05984251968503937, + "grad_norm": 0.30276867747306824, + "learning_rate": 4.701201823456279e-06, + "loss": 0.5953, + "step": 722 + }, + { + "epoch": 0.05992540406133444, + "grad_norm": 0.2928871810436249, + "learning_rate": 4.700787401574804e-06, + "loss": 0.5434, + "step": 723 + }, + { + "epoch": 0.06000828843762951, + "grad_norm": 0.2779698073863983, + "learning_rate": 4.700372979693328e-06, + "loss": 0.5843, + "step": 724 + }, + { + "epoch": 0.06009117281392457, + "grad_norm": 0.27578920125961304, + "learning_rate": 4.6999585578118525e-06, + "loss": 0.58, + "step": 725 + }, + { + "epoch": 0.06017405719021964, + "grad_norm": 0.2727469801902771, + "learning_rate": 4.699544135930377e-06, + "loss": 0.5872, + "step": 726 + }, + { + "epoch": 0.06025694156651471, + "grad_norm": 0.32149773836135864, + "learning_rate": 4.699129714048902e-06, + "loss": 0.5999, + "step": 727 + }, + { + "epoch": 0.06033982594280978, + "grad_norm": 0.27034470438957214, + "learning_rate": 4.698715292167427e-06, + "loss": 0.5785, + "step": 728 + }, + { + "epoch": 0.06042271031910485, + "grad_norm": 0.29294249415397644, + "learning_rate": 4.698300870285952e-06, + "loss": 0.567, + "step": 729 + }, + { + "epoch": 0.060505594695399914, + "grad_norm": 0.30745118856430054, + "learning_rate": 4.697886448404476e-06, + "loss": 0.5819, + "step": 730 + }, + { + "epoch": 0.060588479071694984, + "grad_norm": 0.26490283012390137, + "learning_rate": 4.697472026523001e-06, + "loss": 0.5277, + "step": 731 + }, + { + "epoch": 0.060671363447990054, + "grad_norm": 0.33292704820632935, + "learning_rate": 4.697057604641525e-06, + "loss": 0.5728, + "step": 732 + }, + { + "epoch": 0.060754247824285124, + "grad_norm": 0.35005074739456177, + "learning_rate": 4.69664318276005e-06, + "loss": 0.5773, + "step": 733 + }, + { + "epoch": 0.06083713220058019, + "grad_norm": 0.30568936467170715, + "learning_rate": 4.696228760878575e-06, + "loss": 0.5836, + "step": 734 + }, + { + "epoch": 0.060920016576875256, + "grad_norm": 0.28714996576309204, + "learning_rate": 4.695814338997099e-06, + "loss": 0.6226, + "step": 735 + }, + { + "epoch": 0.061002900953170326, + "grad_norm": 0.27018114924430847, + "learning_rate": 4.695399917115624e-06, + "loss": 0.5558, + "step": 736 + }, + { + "epoch": 0.061085785329465396, + "grad_norm": 0.2913294732570648, + "learning_rate": 4.694985495234149e-06, + "loss": 0.6141, + "step": 737 + }, + { + "epoch": 0.061168669705760466, + "grad_norm": 0.2698752284049988, + "learning_rate": 4.6945710733526735e-06, + "loss": 0.5104, + "step": 738 + }, + { + "epoch": 0.061251554082055536, + "grad_norm": 0.2825221121311188, + "learning_rate": 4.694156651471198e-06, + "loss": 0.5853, + "step": 739 + }, + { + "epoch": 0.0613344384583506, + "grad_norm": 0.3014468848705292, + "learning_rate": 4.693742229589723e-06, + "loss": 0.6025, + "step": 740 + }, + { + "epoch": 0.06141732283464567, + "grad_norm": 0.28394216299057007, + "learning_rate": 4.693327807708248e-06, + "loss": 0.5939, + "step": 741 + }, + { + "epoch": 0.06150020721094074, + "grad_norm": 0.3066372275352478, + "learning_rate": 4.692913385826772e-06, + "loss": 0.6031, + "step": 742 + }, + { + "epoch": 0.06158309158723581, + "grad_norm": 0.2848980724811554, + "learning_rate": 4.6924989639452965e-06, + "loss": 0.5641, + "step": 743 + }, + { + "epoch": 0.06166597596353087, + "grad_norm": 0.27768176794052124, + "learning_rate": 4.692084542063821e-06, + "loss": 0.5654, + "step": 744 + }, + { + "epoch": 0.06174886033982594, + "grad_norm": 0.2884795069694519, + "learning_rate": 4.691670120182346e-06, + "loss": 0.5933, + "step": 745 + }, + { + "epoch": 0.06183174471612101, + "grad_norm": 0.30791938304901123, + "learning_rate": 4.691255698300871e-06, + "loss": 0.6367, + "step": 746 + }, + { + "epoch": 0.06191462909241608, + "grad_norm": 0.6058428883552551, + "learning_rate": 4.690841276419396e-06, + "loss": 0.5954, + "step": 747 + }, + { + "epoch": 0.06199751346871115, + "grad_norm": 0.29314690828323364, + "learning_rate": 4.69042685453792e-06, + "loss": 0.6245, + "step": 748 + }, + { + "epoch": 0.06208039784500621, + "grad_norm": 0.2781945466995239, + "learning_rate": 4.690012432656445e-06, + "loss": 0.5652, + "step": 749 + }, + { + "epoch": 0.06216328222130128, + "grad_norm": 0.28007906675338745, + "learning_rate": 4.689598010774969e-06, + "loss": 0.5715, + "step": 750 + }, + { + "epoch": 0.06224616659759635, + "grad_norm": 0.3084630072116852, + "learning_rate": 4.689183588893494e-06, + "loss": 0.5972, + "step": 751 + }, + { + "epoch": 0.06232905097389142, + "grad_norm": 0.2996090054512024, + "learning_rate": 4.688769167012018e-06, + "loss": 0.6363, + "step": 752 + }, + { + "epoch": 0.06241193535018649, + "grad_norm": 0.2746064066886902, + "learning_rate": 4.688354745130543e-06, + "loss": 0.5683, + "step": 753 + }, + { + "epoch": 0.062494819726481556, + "grad_norm": 0.2953972816467285, + "learning_rate": 4.687940323249068e-06, + "loss": 0.6378, + "step": 754 + }, + { + "epoch": 0.06257770410277663, + "grad_norm": 0.2725238800048828, + "learning_rate": 4.687525901367593e-06, + "loss": 0.5607, + "step": 755 + }, + { + "epoch": 0.06266058847907169, + "grad_norm": 0.2657369077205658, + "learning_rate": 4.6871114794861175e-06, + "loss": 0.5888, + "step": 756 + }, + { + "epoch": 0.06274347285536676, + "grad_norm": 0.2874719500541687, + "learning_rate": 4.686697057604642e-06, + "loss": 0.5624, + "step": 757 + }, + { + "epoch": 0.06282635723166183, + "grad_norm": 0.2981370687484741, + "learning_rate": 4.686282635723166e-06, + "loss": 0.5829, + "step": 758 + }, + { + "epoch": 0.0629092416079569, + "grad_norm": 0.31170526146888733, + "learning_rate": 4.685868213841691e-06, + "loss": 0.6264, + "step": 759 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 0.29102808237075806, + "learning_rate": 4.685453791960216e-06, + "loss": 0.5379, + "step": 760 + }, + { + "epoch": 0.06307501036054704, + "grad_norm": 0.27517038583755493, + "learning_rate": 4.6850393700787405e-06, + "loss": 0.5981, + "step": 761 + }, + { + "epoch": 0.06315789473684211, + "grad_norm": 0.26553791761398315, + "learning_rate": 4.684624948197265e-06, + "loss": 0.5667, + "step": 762 + }, + { + "epoch": 0.06324077911313718, + "grad_norm": 0.290865957736969, + "learning_rate": 4.68421052631579e-06, + "loss": 0.5799, + "step": 763 + }, + { + "epoch": 0.06332366348943225, + "grad_norm": 0.29525673389434814, + "learning_rate": 4.683796104434315e-06, + "loss": 0.5703, + "step": 764 + }, + { + "epoch": 0.06340654786572732, + "grad_norm": 0.3198724687099457, + "learning_rate": 4.683381682552839e-06, + "loss": 0.5583, + "step": 765 + }, + { + "epoch": 0.06348943224202237, + "grad_norm": 0.2965705692768097, + "learning_rate": 4.6829672606713635e-06, + "loss": 0.5653, + "step": 766 + }, + { + "epoch": 0.06357231661831744, + "grad_norm": 0.2679207921028137, + "learning_rate": 4.682552838789888e-06, + "loss": 0.577, + "step": 767 + }, + { + "epoch": 0.06365520099461251, + "grad_norm": 0.28753477334976196, + "learning_rate": 4.682138416908413e-06, + "loss": 0.5833, + "step": 768 + }, + { + "epoch": 0.06373808537090758, + "grad_norm": 0.6845914721488953, + "learning_rate": 4.681723995026938e-06, + "loss": 0.5759, + "step": 769 + }, + { + "epoch": 0.06382096974720265, + "grad_norm": 0.3146698772907257, + "learning_rate": 4.681309573145462e-06, + "loss": 0.5677, + "step": 770 + }, + { + "epoch": 0.06390385412349772, + "grad_norm": 0.33692091703414917, + "learning_rate": 4.680895151263987e-06, + "loss": 0.6569, + "step": 771 + }, + { + "epoch": 0.06398673849979279, + "grad_norm": 0.2617219388484955, + "learning_rate": 4.680480729382512e-06, + "loss": 0.5501, + "step": 772 + }, + { + "epoch": 0.06406962287608786, + "grad_norm": 0.28823381662368774, + "learning_rate": 4.680066307501036e-06, + "loss": 0.5941, + "step": 773 + }, + { + "epoch": 0.06415250725238293, + "grad_norm": 0.2830592393875122, + "learning_rate": 4.679651885619561e-06, + "loss": 0.5897, + "step": 774 + }, + { + "epoch": 0.06423539162867799, + "grad_norm": 0.26408645510673523, + "learning_rate": 4.679237463738085e-06, + "loss": 0.5369, + "step": 775 + }, + { + "epoch": 0.06431827600497306, + "grad_norm": 0.2706126570701599, + "learning_rate": 4.67882304185661e-06, + "loss": 0.5677, + "step": 776 + }, + { + "epoch": 0.06440116038126813, + "grad_norm": 0.28374284505844116, + "learning_rate": 4.678408619975135e-06, + "loss": 0.6084, + "step": 777 + }, + { + "epoch": 0.0644840447575632, + "grad_norm": 0.2860717177391052, + "learning_rate": 4.67799419809366e-06, + "loss": 0.6156, + "step": 778 + }, + { + "epoch": 0.06456692913385827, + "grad_norm": 0.29182013869285583, + "learning_rate": 4.6775797762121845e-06, + "loss": 0.5468, + "step": 779 + }, + { + "epoch": 0.06464981351015334, + "grad_norm": 0.3115129768848419, + "learning_rate": 4.677165354330709e-06, + "loss": 0.6168, + "step": 780 + }, + { + "epoch": 0.0647326978864484, + "grad_norm": 0.27825674414634705, + "learning_rate": 4.676750932449234e-06, + "loss": 0.6018, + "step": 781 + }, + { + "epoch": 0.06481558226274348, + "grad_norm": 0.2646465599536896, + "learning_rate": 4.676336510567759e-06, + "loss": 0.5814, + "step": 782 + }, + { + "epoch": 0.06489846663903855, + "grad_norm": 0.28828758001327515, + "learning_rate": 4.675922088686283e-06, + "loss": 0.5545, + "step": 783 + }, + { + "epoch": 0.06498135101533362, + "grad_norm": 0.280521035194397, + "learning_rate": 4.6755076668048075e-06, + "loss": 0.6112, + "step": 784 + }, + { + "epoch": 0.06506423539162867, + "grad_norm": 0.579290509223938, + "learning_rate": 4.675093244923332e-06, + "loss": 0.5925, + "step": 785 + }, + { + "epoch": 0.06514711976792374, + "grad_norm": 0.2924567461013794, + "learning_rate": 4.674678823041857e-06, + "loss": 0.6264, + "step": 786 + }, + { + "epoch": 0.06523000414421881, + "grad_norm": 0.39812782406806946, + "learning_rate": 4.674264401160382e-06, + "loss": 0.5831, + "step": 787 + }, + { + "epoch": 0.06531288852051388, + "grad_norm": 0.6982154250144958, + "learning_rate": 4.673849979278906e-06, + "loss": 0.5918, + "step": 788 + }, + { + "epoch": 0.06539577289680895, + "grad_norm": 0.29502007365226746, + "learning_rate": 4.673435557397431e-06, + "loss": 0.612, + "step": 789 + }, + { + "epoch": 0.06547865727310402, + "grad_norm": 0.31935253739356995, + "learning_rate": 4.673021135515956e-06, + "loss": 0.5813, + "step": 790 + }, + { + "epoch": 0.06556154164939909, + "grad_norm": 0.2991914451122284, + "learning_rate": 4.67260671363448e-06, + "loss": 0.6005, + "step": 791 + }, + { + "epoch": 0.06564442602569416, + "grad_norm": 0.2707747220993042, + "learning_rate": 4.672192291753005e-06, + "loss": 0.5466, + "step": 792 + }, + { + "epoch": 0.06572731040198923, + "grad_norm": 0.26387956738471985, + "learning_rate": 4.671777869871529e-06, + "loss": 0.5699, + "step": 793 + }, + { + "epoch": 0.06581019477828429, + "grad_norm": 0.3066766560077667, + "learning_rate": 4.671363447990054e-06, + "loss": 0.6839, + "step": 794 + }, + { + "epoch": 0.06589307915457936, + "grad_norm": 0.2787393629550934, + "learning_rate": 4.670949026108579e-06, + "loss": 0.6025, + "step": 795 + }, + { + "epoch": 0.06597596353087443, + "grad_norm": 0.28337380290031433, + "learning_rate": 4.670534604227104e-06, + "loss": 0.5114, + "step": 796 + }, + { + "epoch": 0.0660588479071695, + "grad_norm": 0.28823143243789673, + "learning_rate": 4.6701201823456285e-06, + "loss": 0.5989, + "step": 797 + }, + { + "epoch": 0.06614173228346457, + "grad_norm": 0.28884807229042053, + "learning_rate": 4.669705760464153e-06, + "loss": 0.6495, + "step": 798 + }, + { + "epoch": 0.06622461665975964, + "grad_norm": 0.3149169981479645, + "learning_rate": 4.669291338582677e-06, + "loss": 0.6047, + "step": 799 + }, + { + "epoch": 0.0663075010360547, + "grad_norm": 0.3364529013633728, + "learning_rate": 4.668876916701202e-06, + "loss": 0.5922, + "step": 800 + }, + { + "epoch": 0.06639038541234978, + "grad_norm": 0.30812951922416687, + "learning_rate": 4.668462494819727e-06, + "loss": 0.6195, + "step": 801 + }, + { + "epoch": 0.06647326978864485, + "grad_norm": 0.27544528245925903, + "learning_rate": 4.6680480729382515e-06, + "loss": 0.574, + "step": 802 + }, + { + "epoch": 0.06655615416493992, + "grad_norm": 0.27388808131217957, + "learning_rate": 4.667633651056776e-06, + "loss": 0.593, + "step": 803 + }, + { + "epoch": 0.06663903854123497, + "grad_norm": 0.2912275195121765, + "learning_rate": 4.667219229175301e-06, + "loss": 0.563, + "step": 804 + }, + { + "epoch": 0.06672192291753004, + "grad_norm": 0.3050595819950104, + "learning_rate": 4.666804807293826e-06, + "loss": 0.6467, + "step": 805 + }, + { + "epoch": 0.06680480729382511, + "grad_norm": 0.25503629446029663, + "learning_rate": 4.66639038541235e-06, + "loss": 0.553, + "step": 806 + }, + { + "epoch": 0.06688769167012018, + "grad_norm": 0.30509495735168457, + "learning_rate": 4.6659759635308745e-06, + "loss": 0.6186, + "step": 807 + }, + { + "epoch": 0.06697057604641525, + "grad_norm": 0.26507607102394104, + "learning_rate": 4.665561541649399e-06, + "loss": 0.5537, + "step": 808 + }, + { + "epoch": 0.06705346042271032, + "grad_norm": 0.28209972381591797, + "learning_rate": 4.665147119767924e-06, + "loss": 0.5952, + "step": 809 + }, + { + "epoch": 0.06713634479900539, + "grad_norm": 0.2693888247013092, + "learning_rate": 4.664732697886449e-06, + "loss": 0.5432, + "step": 810 + }, + { + "epoch": 0.06721922917530046, + "grad_norm": 0.2712569534778595, + "learning_rate": 4.664318276004973e-06, + "loss": 0.5276, + "step": 811 + }, + { + "epoch": 0.06730211355159553, + "grad_norm": 0.2586735486984253, + "learning_rate": 4.663903854123498e-06, + "loss": 0.5876, + "step": 812 + }, + { + "epoch": 0.06738499792789059, + "grad_norm": 0.2894429564476013, + "learning_rate": 4.663489432242023e-06, + "loss": 0.6057, + "step": 813 + }, + { + "epoch": 0.06746788230418566, + "grad_norm": 0.29059886932373047, + "learning_rate": 4.663075010360547e-06, + "loss": 0.5958, + "step": 814 + }, + { + "epoch": 0.06755076668048073, + "grad_norm": 0.2671761214733124, + "learning_rate": 4.662660588479072e-06, + "loss": 0.5488, + "step": 815 + }, + { + "epoch": 0.0676336510567758, + "grad_norm": 0.2899112403392792, + "learning_rate": 4.662246166597596e-06, + "loss": 0.5682, + "step": 816 + }, + { + "epoch": 0.06771653543307087, + "grad_norm": 0.2771666944026947, + "learning_rate": 4.661831744716121e-06, + "loss": 0.5804, + "step": 817 + }, + { + "epoch": 0.06779941980936594, + "grad_norm": 0.27950766682624817, + "learning_rate": 4.661417322834646e-06, + "loss": 0.5786, + "step": 818 + }, + { + "epoch": 0.067882304185661, + "grad_norm": 0.26945096254348755, + "learning_rate": 4.661002900953171e-06, + "loss": 0.5961, + "step": 819 + }, + { + "epoch": 0.06796518856195607, + "grad_norm": 0.3177087604999542, + "learning_rate": 4.6605884790716955e-06, + "loss": 0.6059, + "step": 820 + }, + { + "epoch": 0.06804807293825114, + "grad_norm": 0.3108621835708618, + "learning_rate": 4.66017405719022e-06, + "loss": 0.5975, + "step": 821 + }, + { + "epoch": 0.06813095731454621, + "grad_norm": 0.3185555040836334, + "learning_rate": 4.659759635308745e-06, + "loss": 0.6243, + "step": 822 + }, + { + "epoch": 0.06821384169084127, + "grad_norm": 0.2570492923259735, + "learning_rate": 4.65934521342727e-06, + "loss": 0.5182, + "step": 823 + }, + { + "epoch": 0.06829672606713634, + "grad_norm": 0.3114747405052185, + "learning_rate": 4.658930791545794e-06, + "loss": 0.6152, + "step": 824 + }, + { + "epoch": 0.06837961044343141, + "grad_norm": 0.29350051283836365, + "learning_rate": 4.6585163696643186e-06, + "loss": 0.5576, + "step": 825 + }, + { + "epoch": 0.06846249481972648, + "grad_norm": 0.26231256127357483, + "learning_rate": 4.658101947782843e-06, + "loss": 0.572, + "step": 826 + }, + { + "epoch": 0.06854537919602155, + "grad_norm": 0.29008403420448303, + "learning_rate": 4.657687525901368e-06, + "loss": 0.6082, + "step": 827 + }, + { + "epoch": 0.06862826357231662, + "grad_norm": 0.26747050881385803, + "learning_rate": 4.657273104019893e-06, + "loss": 0.5705, + "step": 828 + }, + { + "epoch": 0.06871114794861169, + "grad_norm": 0.3138808012008667, + "learning_rate": 4.656858682138417e-06, + "loss": 0.5851, + "step": 829 + }, + { + "epoch": 0.06879403232490676, + "grad_norm": 0.28958410024642944, + "learning_rate": 4.656444260256942e-06, + "loss": 0.6205, + "step": 830 + }, + { + "epoch": 0.06887691670120183, + "grad_norm": 0.3138085603713989, + "learning_rate": 4.656029838375467e-06, + "loss": 0.5638, + "step": 831 + }, + { + "epoch": 0.06895980107749689, + "grad_norm": 0.28301113843917847, + "learning_rate": 4.655615416493991e-06, + "loss": 0.5584, + "step": 832 + }, + { + "epoch": 0.06904268545379195, + "grad_norm": 0.3166961371898651, + "learning_rate": 4.655200994612516e-06, + "loss": 0.619, + "step": 833 + }, + { + "epoch": 0.06912556983008702, + "grad_norm": 0.2631582021713257, + "learning_rate": 4.65478657273104e-06, + "loss": 0.5649, + "step": 834 + }, + { + "epoch": 0.0692084542063821, + "grad_norm": 0.3018805980682373, + "learning_rate": 4.654372150849565e-06, + "loss": 0.6024, + "step": 835 + }, + { + "epoch": 0.06929133858267716, + "grad_norm": 0.28160175681114197, + "learning_rate": 4.65395772896809e-06, + "loss": 0.5992, + "step": 836 + }, + { + "epoch": 0.06937422295897223, + "grad_norm": 0.29403626918792725, + "learning_rate": 4.653543307086614e-06, + "loss": 0.5833, + "step": 837 + }, + { + "epoch": 0.0694571073352673, + "grad_norm": 0.2720746695995331, + "learning_rate": 4.6531288852051396e-06, + "loss": 0.582, + "step": 838 + }, + { + "epoch": 0.06953999171156237, + "grad_norm": 0.2988300323486328, + "learning_rate": 4.652714463323664e-06, + "loss": 0.6025, + "step": 839 + }, + { + "epoch": 0.06962287608785744, + "grad_norm": 0.2717914283275604, + "learning_rate": 4.652300041442188e-06, + "loss": 0.5843, + "step": 840 + }, + { + "epoch": 0.06970576046415251, + "grad_norm": 0.2766326069831848, + "learning_rate": 4.651885619560713e-06, + "loss": 0.5977, + "step": 841 + }, + { + "epoch": 0.06978864484044757, + "grad_norm": 0.2628249526023865, + "learning_rate": 4.651471197679237e-06, + "loss": 0.5443, + "step": 842 + }, + { + "epoch": 0.06987152921674264, + "grad_norm": 0.3041544258594513, + "learning_rate": 4.6510567757977626e-06, + "loss": 0.5974, + "step": 843 + }, + { + "epoch": 0.06995441359303771, + "grad_norm": 0.2895543575286865, + "learning_rate": 4.650642353916287e-06, + "loss": 0.533, + "step": 844 + }, + { + "epoch": 0.07003729796933278, + "grad_norm": 0.3073027431964874, + "learning_rate": 4.650227932034812e-06, + "loss": 0.5613, + "step": 845 + }, + { + "epoch": 0.07012018234562785, + "grad_norm": 0.284402996301651, + "learning_rate": 4.649813510153337e-06, + "loss": 0.57, + "step": 846 + }, + { + "epoch": 0.07020306672192292, + "grad_norm": 0.306367427110672, + "learning_rate": 4.649399088271861e-06, + "loss": 0.5877, + "step": 847 + }, + { + "epoch": 0.07028595109821799, + "grad_norm": 0.27843910455703735, + "learning_rate": 4.6489846663903856e-06, + "loss": 0.5521, + "step": 848 + }, + { + "epoch": 0.07036883547451306, + "grad_norm": 0.2685913145542145, + "learning_rate": 4.64857024450891e-06, + "loss": 0.5531, + "step": 849 + }, + { + "epoch": 0.07045171985080813, + "grad_norm": 0.2856747508049011, + "learning_rate": 4.648155822627435e-06, + "loss": 0.5333, + "step": 850 + }, + { + "epoch": 0.07053460422710318, + "grad_norm": 0.2754206657409668, + "learning_rate": 4.64774140074596e-06, + "loss": 0.534, + "step": 851 + }, + { + "epoch": 0.07061748860339825, + "grad_norm": 0.2726350724697113, + "learning_rate": 4.647326978864484e-06, + "loss": 0.5717, + "step": 852 + }, + { + "epoch": 0.07070037297969332, + "grad_norm": 0.28771349787712097, + "learning_rate": 4.646912556983009e-06, + "loss": 0.5928, + "step": 853 + }, + { + "epoch": 0.0707832573559884, + "grad_norm": 0.29740050435066223, + "learning_rate": 4.646498135101534e-06, + "loss": 0.5799, + "step": 854 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 0.2713727056980133, + "learning_rate": 4.646083713220058e-06, + "loss": 0.5455, + "step": 855 + }, + { + "epoch": 0.07094902610857853, + "grad_norm": 0.2672290503978729, + "learning_rate": 4.645669291338583e-06, + "loss": 0.6026, + "step": 856 + }, + { + "epoch": 0.0710319104848736, + "grad_norm": 0.2939911186695099, + "learning_rate": 4.645254869457108e-06, + "loss": 0.6348, + "step": 857 + }, + { + "epoch": 0.07111479486116867, + "grad_norm": 0.25928667187690735, + "learning_rate": 4.6448404475756324e-06, + "loss": 0.5555, + "step": 858 + }, + { + "epoch": 0.07119767923746374, + "grad_norm": 0.2910548448562622, + "learning_rate": 4.644426025694157e-06, + "loss": 0.6037, + "step": 859 + }, + { + "epoch": 0.07128056361375881, + "grad_norm": 0.2954711616039276, + "learning_rate": 4.644011603812681e-06, + "loss": 0.611, + "step": 860 + }, + { + "epoch": 0.07136344799005387, + "grad_norm": 0.27958351373672485, + "learning_rate": 4.6435971819312066e-06, + "loss": 0.6158, + "step": 861 + }, + { + "epoch": 0.07144633236634894, + "grad_norm": 0.3107668161392212, + "learning_rate": 4.643182760049731e-06, + "loss": 0.6169, + "step": 862 + }, + { + "epoch": 0.07152921674264401, + "grad_norm": 0.26710301637649536, + "learning_rate": 4.642768338168256e-06, + "loss": 0.5501, + "step": 863 + }, + { + "epoch": 0.07161210111893908, + "grad_norm": 0.2909033000469208, + "learning_rate": 4.642353916286781e-06, + "loss": 0.5757, + "step": 864 + }, + { + "epoch": 0.07169498549523415, + "grad_norm": 0.29376739263534546, + "learning_rate": 4.641939494405305e-06, + "loss": 0.6655, + "step": 865 + }, + { + "epoch": 0.07177786987152922, + "grad_norm": 0.30144259333610535, + "learning_rate": 4.6415250725238296e-06, + "loss": 0.6237, + "step": 866 + }, + { + "epoch": 0.07186075424782429, + "grad_norm": 0.3186337351799011, + "learning_rate": 4.641110650642354e-06, + "loss": 0.5916, + "step": 867 + }, + { + "epoch": 0.07194363862411936, + "grad_norm": 0.31938549876213074, + "learning_rate": 4.640696228760879e-06, + "loss": 0.5674, + "step": 868 + }, + { + "epoch": 0.07202652300041443, + "grad_norm": 0.2961660623550415, + "learning_rate": 4.640281806879404e-06, + "loss": 0.5744, + "step": 869 + }, + { + "epoch": 0.07210940737670948, + "grad_norm": 0.2849061191082001, + "learning_rate": 4.639867384997928e-06, + "loss": 0.5862, + "step": 870 + }, + { + "epoch": 0.07219229175300455, + "grad_norm": 0.3009960353374481, + "learning_rate": 4.6394529631164534e-06, + "loss": 0.5952, + "step": 871 + }, + { + "epoch": 0.07227517612929962, + "grad_norm": 0.2673109471797943, + "learning_rate": 4.639038541234978e-06, + "loss": 0.5688, + "step": 872 + }, + { + "epoch": 0.07235806050559469, + "grad_norm": 0.2695416510105133, + "learning_rate": 4.638624119353502e-06, + "loss": 0.535, + "step": 873 + }, + { + "epoch": 0.07244094488188976, + "grad_norm": 0.30445408821105957, + "learning_rate": 4.638209697472027e-06, + "loss": 0.5857, + "step": 874 + }, + { + "epoch": 0.07252382925818483, + "grad_norm": 0.284464567899704, + "learning_rate": 4.637795275590551e-06, + "loss": 0.5467, + "step": 875 + }, + { + "epoch": 0.0726067136344799, + "grad_norm": 0.28305649757385254, + "learning_rate": 4.6373808537090764e-06, + "loss": 0.6116, + "step": 876 + }, + { + "epoch": 0.07268959801077497, + "grad_norm": 0.30277690291404724, + "learning_rate": 4.636966431827601e-06, + "loss": 0.5984, + "step": 877 + }, + { + "epoch": 0.07277248238707004, + "grad_norm": 0.2971579432487488, + "learning_rate": 4.636552009946125e-06, + "loss": 0.5983, + "step": 878 + }, + { + "epoch": 0.07285536676336511, + "grad_norm": 0.3041172921657562, + "learning_rate": 4.636137588064651e-06, + "loss": 0.6139, + "step": 879 + }, + { + "epoch": 0.07293825113966017, + "grad_norm": 0.276118665933609, + "learning_rate": 4.635723166183175e-06, + "loss": 0.5651, + "step": 880 + }, + { + "epoch": 0.07302113551595524, + "grad_norm": 0.29044827818870544, + "learning_rate": 4.6353087443016994e-06, + "loss": 0.5771, + "step": 881 + }, + { + "epoch": 0.07310401989225031, + "grad_norm": 0.2886148691177368, + "learning_rate": 4.634894322420224e-06, + "loss": 0.6004, + "step": 882 + }, + { + "epoch": 0.07318690426854538, + "grad_norm": 0.27924251556396484, + "learning_rate": 4.634479900538748e-06, + "loss": 0.5733, + "step": 883 + }, + { + "epoch": 0.07326978864484045, + "grad_norm": 0.2806059420108795, + "learning_rate": 4.634065478657274e-06, + "loss": 0.592, + "step": 884 + }, + { + "epoch": 0.07335267302113552, + "grad_norm": 0.2916703522205353, + "learning_rate": 4.633651056775798e-06, + "loss": 0.5768, + "step": 885 + }, + { + "epoch": 0.07343555739743059, + "grad_norm": 0.27948641777038574, + "learning_rate": 4.633236634894323e-06, + "loss": 0.5724, + "step": 886 + }, + { + "epoch": 0.07351844177372566, + "grad_norm": 0.2653394639492035, + "learning_rate": 4.632822213012848e-06, + "loss": 0.5533, + "step": 887 + }, + { + "epoch": 0.07360132615002073, + "grad_norm": 0.268749862909317, + "learning_rate": 4.632407791131372e-06, + "loss": 0.6012, + "step": 888 + }, + { + "epoch": 0.07368421052631578, + "grad_norm": 0.27196136116981506, + "learning_rate": 4.631993369249897e-06, + "loss": 0.5932, + "step": 889 + }, + { + "epoch": 0.07376709490261085, + "grad_norm": 0.28965434432029724, + "learning_rate": 4.631578947368421e-06, + "loss": 0.5928, + "step": 890 + }, + { + "epoch": 0.07384997927890592, + "grad_norm": 0.2617194950580597, + "learning_rate": 4.6311645254869455e-06, + "loss": 0.5013, + "step": 891 + }, + { + "epoch": 0.07393286365520099, + "grad_norm": 0.27590808272361755, + "learning_rate": 4.630750103605471e-06, + "loss": 0.5727, + "step": 892 + }, + { + "epoch": 0.07401574803149606, + "grad_norm": 0.31232619285583496, + "learning_rate": 4.630335681723995e-06, + "loss": 0.582, + "step": 893 + }, + { + "epoch": 0.07409863240779113, + "grad_norm": 0.28596141934394836, + "learning_rate": 4.6299212598425204e-06, + "loss": 0.5318, + "step": 894 + }, + { + "epoch": 0.0741815167840862, + "grad_norm": 0.28568896651268005, + "learning_rate": 4.629506837961045e-06, + "loss": 0.5716, + "step": 895 + }, + { + "epoch": 0.07426440116038127, + "grad_norm": 0.2786010503768921, + "learning_rate": 4.629092416079569e-06, + "loss": 0.5857, + "step": 896 + }, + { + "epoch": 0.07434728553667634, + "grad_norm": 0.29684361815452576, + "learning_rate": 4.628677994198094e-06, + "loss": 0.5823, + "step": 897 + }, + { + "epoch": 0.07443016991297141, + "grad_norm": 0.29418399930000305, + "learning_rate": 4.628263572316619e-06, + "loss": 0.6107, + "step": 898 + }, + { + "epoch": 0.07451305428926647, + "grad_norm": 0.2867834270000458, + "learning_rate": 4.6278491504351434e-06, + "loss": 0.5453, + "step": 899 + }, + { + "epoch": 0.07459593866556154, + "grad_norm": 0.28534021973609924, + "learning_rate": 4.627434728553668e-06, + "loss": 0.5488, + "step": 900 + }, + { + "epoch": 0.0746788230418566, + "grad_norm": 0.267385333776474, + "learning_rate": 4.627020306672192e-06, + "loss": 0.5429, + "step": 901 + }, + { + "epoch": 0.07476170741815168, + "grad_norm": 0.2833807170391083, + "learning_rate": 4.626605884790718e-06, + "loss": 0.6044, + "step": 902 + }, + { + "epoch": 0.07484459179444675, + "grad_norm": 0.25868263840675354, + "learning_rate": 4.626191462909242e-06, + "loss": 0.5254, + "step": 903 + }, + { + "epoch": 0.07492747617074182, + "grad_norm": 0.3075624108314514, + "learning_rate": 4.6257770410277665e-06, + "loss": 0.6086, + "step": 904 + }, + { + "epoch": 0.07501036054703689, + "grad_norm": 0.28035736083984375, + "learning_rate": 4.625362619146292e-06, + "loss": 0.5609, + "step": 905 + }, + { + "epoch": 0.07509324492333196, + "grad_norm": 0.26101911067962646, + "learning_rate": 4.624948197264816e-06, + "loss": 0.5453, + "step": 906 + }, + { + "epoch": 0.07517612929962703, + "grad_norm": 0.2764147222042084, + "learning_rate": 4.624533775383341e-06, + "loss": 0.6041, + "step": 907 + }, + { + "epoch": 0.07525901367592208, + "grad_norm": 0.286871999502182, + "learning_rate": 4.624119353501865e-06, + "loss": 0.573, + "step": 908 + }, + { + "epoch": 0.07534189805221715, + "grad_norm": 0.28192052245140076, + "learning_rate": 4.6237049316203895e-06, + "loss": 0.5881, + "step": 909 + }, + { + "epoch": 0.07542478242851222, + "grad_norm": 0.28322073817253113, + "learning_rate": 4.623290509738915e-06, + "loss": 0.6515, + "step": 910 + }, + { + "epoch": 0.07550766680480729, + "grad_norm": 0.28578972816467285, + "learning_rate": 4.622876087857439e-06, + "loss": 0.6155, + "step": 911 + }, + { + "epoch": 0.07559055118110236, + "grad_norm": 0.2836241126060486, + "learning_rate": 4.6224616659759645e-06, + "loss": 0.5832, + "step": 912 + }, + { + "epoch": 0.07567343555739743, + "grad_norm": 0.26934218406677246, + "learning_rate": 4.622047244094489e-06, + "loss": 0.566, + "step": 913 + }, + { + "epoch": 0.0757563199336925, + "grad_norm": 0.3163556754589081, + "learning_rate": 4.621632822213013e-06, + "loss": 0.5789, + "step": 914 + }, + { + "epoch": 0.07583920430998757, + "grad_norm": 0.2839760184288025, + "learning_rate": 4.621218400331538e-06, + "loss": 0.6251, + "step": 915 + }, + { + "epoch": 0.07592208868628264, + "grad_norm": 0.2643094062805176, + "learning_rate": 4.620803978450062e-06, + "loss": 0.5092, + "step": 916 + }, + { + "epoch": 0.07600497306257771, + "grad_norm": 0.2878752648830414, + "learning_rate": 4.6203895565685875e-06, + "loss": 0.5778, + "step": 917 + }, + { + "epoch": 0.07608785743887277, + "grad_norm": 0.29493191838264465, + "learning_rate": 4.619975134687112e-06, + "loss": 0.5896, + "step": 918 + }, + { + "epoch": 0.07617074181516784, + "grad_norm": 0.3022734522819519, + "learning_rate": 4.619560712805636e-06, + "loss": 0.5854, + "step": 919 + }, + { + "epoch": 0.0762536261914629, + "grad_norm": 0.26750433444976807, + "learning_rate": 4.619146290924162e-06, + "loss": 0.524, + "step": 920 + }, + { + "epoch": 0.07633651056775798, + "grad_norm": 0.29440945386886597, + "learning_rate": 4.618731869042686e-06, + "loss": 0.5875, + "step": 921 + }, + { + "epoch": 0.07641939494405305, + "grad_norm": 0.3142056167125702, + "learning_rate": 4.6183174471612105e-06, + "loss": 0.577, + "step": 922 + }, + { + "epoch": 0.07650227932034812, + "grad_norm": 0.3178985118865967, + "learning_rate": 4.617903025279735e-06, + "loss": 0.601, + "step": 923 + }, + { + "epoch": 0.07658516369664319, + "grad_norm": 0.2976849675178528, + "learning_rate": 4.617488603398259e-06, + "loss": 0.5471, + "step": 924 + }, + { + "epoch": 0.07666804807293826, + "grad_norm": 0.2682495415210724, + "learning_rate": 4.617074181516785e-06, + "loss": 0.5715, + "step": 925 + }, + { + "epoch": 0.07675093244923333, + "grad_norm": 0.27683454751968384, + "learning_rate": 4.616659759635309e-06, + "loss": 0.5491, + "step": 926 + }, + { + "epoch": 0.07683381682552838, + "grad_norm": 0.28415948152542114, + "learning_rate": 4.6162453377538335e-06, + "loss": 0.557, + "step": 927 + }, + { + "epoch": 0.07691670120182345, + "grad_norm": 0.2801585793495178, + "learning_rate": 4.615830915872359e-06, + "loss": 0.5505, + "step": 928 + }, + { + "epoch": 0.07699958557811852, + "grad_norm": 0.2777591645717621, + "learning_rate": 4.615416493990883e-06, + "loss": 0.6074, + "step": 929 + }, + { + "epoch": 0.07708246995441359, + "grad_norm": 0.29386720061302185, + "learning_rate": 4.615002072109408e-06, + "loss": 0.5885, + "step": 930 + }, + { + "epoch": 0.07716535433070866, + "grad_norm": 0.2794555425643921, + "learning_rate": 4.614587650227932e-06, + "loss": 0.5842, + "step": 931 + }, + { + "epoch": 0.07724823870700373, + "grad_norm": 0.2642279267311096, + "learning_rate": 4.6141732283464565e-06, + "loss": 0.5904, + "step": 932 + }, + { + "epoch": 0.0773311230832988, + "grad_norm": 0.27595770359039307, + "learning_rate": 4.613758806464982e-06, + "loss": 0.6152, + "step": 933 + }, + { + "epoch": 0.07741400745959387, + "grad_norm": 0.2756257951259613, + "learning_rate": 4.613344384583506e-06, + "loss": 0.5643, + "step": 934 + }, + { + "epoch": 0.07749689183588894, + "grad_norm": 0.2855170667171478, + "learning_rate": 4.6129299627020315e-06, + "loss": 0.5453, + "step": 935 + }, + { + "epoch": 0.07757977621218401, + "grad_norm": 0.2877998650074005, + "learning_rate": 4.612515540820556e-06, + "loss": 0.5931, + "step": 936 + }, + { + "epoch": 0.07766266058847907, + "grad_norm": 0.3092268705368042, + "learning_rate": 4.61210111893908e-06, + "loss": 0.5867, + "step": 937 + }, + { + "epoch": 0.07774554496477414, + "grad_norm": 0.29507628083229065, + "learning_rate": 4.611686697057606e-06, + "loss": 0.582, + "step": 938 + }, + { + "epoch": 0.0778284293410692, + "grad_norm": 0.2855828106403351, + "learning_rate": 4.61127227517613e-06, + "loss": 0.5671, + "step": 939 + }, + { + "epoch": 0.07791131371736428, + "grad_norm": 0.2724294066429138, + "learning_rate": 4.6108578532946545e-06, + "loss": 0.611, + "step": 940 + }, + { + "epoch": 0.07799419809365934, + "grad_norm": 0.31807631254196167, + "learning_rate": 4.610443431413179e-06, + "loss": 0.606, + "step": 941 + }, + { + "epoch": 0.07807708246995441, + "grad_norm": 0.26481735706329346, + "learning_rate": 4.610029009531703e-06, + "loss": 0.5526, + "step": 942 + }, + { + "epoch": 0.07815996684624948, + "grad_norm": 0.28358331322669983, + "learning_rate": 4.609614587650229e-06, + "loss": 0.5867, + "step": 943 + }, + { + "epoch": 0.07824285122254455, + "grad_norm": 0.2728070914745331, + "learning_rate": 4.609200165768753e-06, + "loss": 0.5966, + "step": 944 + }, + { + "epoch": 0.07832573559883962, + "grad_norm": 0.3131295442581177, + "learning_rate": 4.6087857438872775e-06, + "loss": 0.6205, + "step": 945 + }, + { + "epoch": 0.07840861997513468, + "grad_norm": 0.27144384384155273, + "learning_rate": 4.608371322005803e-06, + "loss": 0.5518, + "step": 946 + }, + { + "epoch": 0.07849150435142975, + "grad_norm": 0.2774192690849304, + "learning_rate": 4.607956900124327e-06, + "loss": 0.5836, + "step": 947 + }, + { + "epoch": 0.07857438872772482, + "grad_norm": 0.29335522651672363, + "learning_rate": 4.607542478242852e-06, + "loss": 0.5974, + "step": 948 + }, + { + "epoch": 0.07865727310401989, + "grad_norm": 0.2898538410663605, + "learning_rate": 4.607128056361376e-06, + "loss": 0.5777, + "step": 949 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 0.27504467964172363, + "learning_rate": 4.6067136344799005e-06, + "loss": 0.5729, + "step": 950 + }, + { + "epoch": 0.07882304185661003, + "grad_norm": 0.3004942536354065, + "learning_rate": 4.606299212598426e-06, + "loss": 0.627, + "step": 951 + }, + { + "epoch": 0.0789059262329051, + "grad_norm": 0.25771564245224, + "learning_rate": 4.60588479071695e-06, + "loss": 0.5324, + "step": 952 + }, + { + "epoch": 0.07898881060920017, + "grad_norm": 0.2826637327671051, + "learning_rate": 4.6054703688354755e-06, + "loss": 0.5688, + "step": 953 + }, + { + "epoch": 0.07907169498549524, + "grad_norm": 0.25336191058158875, + "learning_rate": 4.605055946954e-06, + "loss": 0.5369, + "step": 954 + }, + { + "epoch": 0.07915457936179031, + "grad_norm": 0.284369558095932, + "learning_rate": 4.604641525072524e-06, + "loss": 0.5965, + "step": 955 + }, + { + "epoch": 0.07923746373808536, + "grad_norm": 0.28321734070777893, + "learning_rate": 4.604227103191049e-06, + "loss": 0.5869, + "step": 956 + }, + { + "epoch": 0.07932034811438043, + "grad_norm": 0.2602824568748474, + "learning_rate": 4.603812681309573e-06, + "loss": 0.5676, + "step": 957 + }, + { + "epoch": 0.0794032324906755, + "grad_norm": 0.2768510580062866, + "learning_rate": 4.603398259428098e-06, + "loss": 0.5589, + "step": 958 + }, + { + "epoch": 0.07948611686697057, + "grad_norm": 0.26958176493644714, + "learning_rate": 4.602983837546623e-06, + "loss": 0.5712, + "step": 959 + }, + { + "epoch": 0.07956900124326564, + "grad_norm": 0.29090556502342224, + "learning_rate": 4.602569415665147e-06, + "loss": 0.5713, + "step": 960 + }, + { + "epoch": 0.07965188561956071, + "grad_norm": 0.2768706679344177, + "learning_rate": 4.602154993783673e-06, + "loss": 0.5726, + "step": 961 + }, + { + "epoch": 0.07973476999585578, + "grad_norm": 0.2653220593929291, + "learning_rate": 4.601740571902197e-06, + "loss": 0.5945, + "step": 962 + }, + { + "epoch": 0.07981765437215085, + "grad_norm": 0.2717601954936981, + "learning_rate": 4.6013261500207215e-06, + "loss": 0.5258, + "step": 963 + }, + { + "epoch": 0.07990053874844592, + "grad_norm": 0.310914009809494, + "learning_rate": 4.600911728139246e-06, + "loss": 0.6313, + "step": 964 + }, + { + "epoch": 0.079983423124741, + "grad_norm": 0.264925092458725, + "learning_rate": 4.60049730625777e-06, + "loss": 0.5284, + "step": 965 + }, + { + "epoch": 0.08006630750103605, + "grad_norm": 0.2913883328437805, + "learning_rate": 4.600082884376296e-06, + "loss": 0.5795, + "step": 966 + }, + { + "epoch": 0.08014919187733112, + "grad_norm": 0.2996979057788849, + "learning_rate": 4.59966846249482e-06, + "loss": 0.5519, + "step": 967 + }, + { + "epoch": 0.08023207625362619, + "grad_norm": 0.3101538419723511, + "learning_rate": 4.5992540406133445e-06, + "loss": 0.5931, + "step": 968 + }, + { + "epoch": 0.08031496062992126, + "grad_norm": 0.27796417474746704, + "learning_rate": 4.59883961873187e-06, + "loss": 0.5767, + "step": 969 + }, + { + "epoch": 0.08039784500621633, + "grad_norm": 0.30436813831329346, + "learning_rate": 4.598425196850394e-06, + "loss": 0.6166, + "step": 970 + }, + { + "epoch": 0.0804807293825114, + "grad_norm": 0.2926459014415741, + "learning_rate": 4.598010774968919e-06, + "loss": 0.6215, + "step": 971 + }, + { + "epoch": 0.08056361375880647, + "grad_norm": 0.3253946900367737, + "learning_rate": 4.597596353087443e-06, + "loss": 0.6237, + "step": 972 + }, + { + "epoch": 0.08064649813510154, + "grad_norm": 0.2841174006462097, + "learning_rate": 4.5971819312059675e-06, + "loss": 0.5705, + "step": 973 + }, + { + "epoch": 0.08072938251139661, + "grad_norm": 0.2749116122722626, + "learning_rate": 4.596767509324493e-06, + "loss": 0.5814, + "step": 974 + }, + { + "epoch": 0.08081226688769166, + "grad_norm": 0.321963906288147, + "learning_rate": 4.596353087443017e-06, + "loss": 0.5331, + "step": 975 + }, + { + "epoch": 0.08089515126398673, + "grad_norm": 0.28339359164237976, + "learning_rate": 4.595938665561542e-06, + "loss": 0.5519, + "step": 976 + }, + { + "epoch": 0.0809780356402818, + "grad_norm": 0.3047540783882141, + "learning_rate": 4.595524243680067e-06, + "loss": 0.5385, + "step": 977 + }, + { + "epoch": 0.08106092001657687, + "grad_norm": 0.29049691557884216, + "learning_rate": 4.595109821798591e-06, + "loss": 0.5892, + "step": 978 + }, + { + "epoch": 0.08114380439287194, + "grad_norm": 0.2776451110839844, + "learning_rate": 4.594695399917117e-06, + "loss": 0.5399, + "step": 979 + }, + { + "epoch": 0.08122668876916701, + "grad_norm": 0.28646716475486755, + "learning_rate": 4.594280978035641e-06, + "loss": 0.5675, + "step": 980 + }, + { + "epoch": 0.08130957314546208, + "grad_norm": 0.3048925995826721, + "learning_rate": 4.5938665561541655e-06, + "loss": 0.6071, + "step": 981 + }, + { + "epoch": 0.08139245752175715, + "grad_norm": 0.29181942343711853, + "learning_rate": 4.59345213427269e-06, + "loss": 0.5585, + "step": 982 + }, + { + "epoch": 0.08147534189805222, + "grad_norm": 0.2846011817455292, + "learning_rate": 4.593037712391214e-06, + "loss": 0.5712, + "step": 983 + }, + { + "epoch": 0.08155822627434729, + "grad_norm": 0.3035436272621155, + "learning_rate": 4.59262329050974e-06, + "loss": 0.5449, + "step": 984 + }, + { + "epoch": 0.08164111065064235, + "grad_norm": 0.3049437701702118, + "learning_rate": 4.592208868628264e-06, + "loss": 0.5569, + "step": 985 + }, + { + "epoch": 0.08172399502693742, + "grad_norm": 0.27982205152511597, + "learning_rate": 4.5917944467467885e-06, + "loss": 0.6024, + "step": 986 + }, + { + "epoch": 0.08180687940323249, + "grad_norm": 0.2851247787475586, + "learning_rate": 4.591380024865314e-06, + "loss": 0.5726, + "step": 987 + }, + { + "epoch": 0.08188976377952756, + "grad_norm": 0.28597524762153625, + "learning_rate": 4.590965602983838e-06, + "loss": 0.5591, + "step": 988 + }, + { + "epoch": 0.08197264815582263, + "grad_norm": 0.28866276144981384, + "learning_rate": 4.590551181102363e-06, + "loss": 0.5844, + "step": 989 + }, + { + "epoch": 0.0820555325321177, + "grad_norm": 0.3233027458190918, + "learning_rate": 4.590136759220887e-06, + "loss": 0.6175, + "step": 990 + }, + { + "epoch": 0.08213841690841277, + "grad_norm": 0.2642940580844879, + "learning_rate": 4.5897223373394115e-06, + "loss": 0.5395, + "step": 991 + }, + { + "epoch": 0.08222130128470784, + "grad_norm": 0.2673892378807068, + "learning_rate": 4.589307915457937e-06, + "loss": 0.5654, + "step": 992 + }, + { + "epoch": 0.08230418566100291, + "grad_norm": 0.31682756543159485, + "learning_rate": 4.588893493576461e-06, + "loss": 0.626, + "step": 993 + }, + { + "epoch": 0.08238707003729796, + "grad_norm": 0.2774866223335266, + "learning_rate": 4.588479071694986e-06, + "loss": 0.5731, + "step": 994 + }, + { + "epoch": 0.08246995441359303, + "grad_norm": 0.27920541167259216, + "learning_rate": 4.588064649813511e-06, + "loss": 0.5893, + "step": 995 + }, + { + "epoch": 0.0825528387898881, + "grad_norm": 0.2686271071434021, + "learning_rate": 4.587650227932035e-06, + "loss": 0.6299, + "step": 996 + }, + { + "epoch": 0.08263572316618317, + "grad_norm": 0.2720228135585785, + "learning_rate": 4.58723580605056e-06, + "loss": 0.5875, + "step": 997 + }, + { + "epoch": 0.08271860754247824, + "grad_norm": 0.27736878395080566, + "learning_rate": 4.586821384169084e-06, + "loss": 0.5505, + "step": 998 + }, + { + "epoch": 0.08280149191877331, + "grad_norm": 0.27739274501800537, + "learning_rate": 4.586406962287609e-06, + "loss": 0.5355, + "step": 999 + }, + { + "epoch": 0.08288437629506838, + "grad_norm": 0.26536324620246887, + "learning_rate": 4.585992540406134e-06, + "loss": 0.5168, + "step": 1000 + }, + { + "epoch": 0.08296726067136345, + "grad_norm": 0.2846812903881073, + "learning_rate": 4.585578118524658e-06, + "loss": 0.551, + "step": 1001 + }, + { + "epoch": 0.08305014504765852, + "grad_norm": 0.3005164563655853, + "learning_rate": 4.585163696643184e-06, + "loss": 0.6272, + "step": 1002 + }, + { + "epoch": 0.08313302942395359, + "grad_norm": 0.25957050919532776, + "learning_rate": 4.584749274761708e-06, + "loss": 0.5423, + "step": 1003 + }, + { + "epoch": 0.08321591380024865, + "grad_norm": 0.27568385004997253, + "learning_rate": 4.5843348528802325e-06, + "loss": 0.5638, + "step": 1004 + }, + { + "epoch": 0.08329879817654372, + "grad_norm": 0.2991975247859955, + "learning_rate": 4.583920430998757e-06, + "loss": 0.5643, + "step": 1005 + }, + { + "epoch": 0.08338168255283879, + "grad_norm": 0.29007628560066223, + "learning_rate": 4.583506009117281e-06, + "loss": 0.5901, + "step": 1006 + }, + { + "epoch": 0.08346456692913386, + "grad_norm": 0.27763497829437256, + "learning_rate": 4.583091587235806e-06, + "loss": 0.5551, + "step": 1007 + }, + { + "epoch": 0.08354745130542893, + "grad_norm": 0.28060728311538696, + "learning_rate": 4.582677165354331e-06, + "loss": 0.6016, + "step": 1008 + }, + { + "epoch": 0.083630335681724, + "grad_norm": 0.28669673204421997, + "learning_rate": 4.5822627434728555e-06, + "loss": 0.5516, + "step": 1009 + }, + { + "epoch": 0.08371322005801907, + "grad_norm": 0.29246529936790466, + "learning_rate": 4.581848321591381e-06, + "loss": 0.6494, + "step": 1010 + }, + { + "epoch": 0.08379610443431414, + "grad_norm": 0.2832469344139099, + "learning_rate": 4.581433899709905e-06, + "loss": 0.5814, + "step": 1011 + }, + { + "epoch": 0.0838789888106092, + "grad_norm": 0.3135397434234619, + "learning_rate": 4.58101947782843e-06, + "loss": 0.5806, + "step": 1012 + }, + { + "epoch": 0.08396187318690426, + "grad_norm": 0.2837733030319214, + "learning_rate": 4.580605055946954e-06, + "loss": 0.561, + "step": 1013 + }, + { + "epoch": 0.08404475756319933, + "grad_norm": 0.28442996740341187, + "learning_rate": 4.5801906340654785e-06, + "loss": 0.5838, + "step": 1014 + }, + { + "epoch": 0.0841276419394944, + "grad_norm": 0.27904921770095825, + "learning_rate": 4.579776212184004e-06, + "loss": 0.5803, + "step": 1015 + }, + { + "epoch": 0.08421052631578947, + "grad_norm": 0.2706582546234131, + "learning_rate": 4.579361790302528e-06, + "loss": 0.5515, + "step": 1016 + }, + { + "epoch": 0.08429341069208454, + "grad_norm": 0.2632027566432953, + "learning_rate": 4.578947368421053e-06, + "loss": 0.5016, + "step": 1017 + }, + { + "epoch": 0.08437629506837961, + "grad_norm": 0.30025342106819153, + "learning_rate": 4.578532946539578e-06, + "loss": 0.6257, + "step": 1018 + }, + { + "epoch": 0.08445917944467468, + "grad_norm": 0.2795947790145874, + "learning_rate": 4.578118524658102e-06, + "loss": 0.5739, + "step": 1019 + }, + { + "epoch": 0.08454206382096975, + "grad_norm": 0.2858903408050537, + "learning_rate": 4.577704102776628e-06, + "loss": 0.6064, + "step": 1020 + }, + { + "epoch": 0.08462494819726482, + "grad_norm": 0.24879910051822662, + "learning_rate": 4.577289680895152e-06, + "loss": 0.5109, + "step": 1021 + }, + { + "epoch": 0.08470783257355989, + "grad_norm": 0.2831918001174927, + "learning_rate": 4.5768752590136765e-06, + "loss": 0.5658, + "step": 1022 + }, + { + "epoch": 0.08479071694985495, + "grad_norm": 0.29785382747650146, + "learning_rate": 4.576460837132201e-06, + "loss": 0.5658, + "step": 1023 + }, + { + "epoch": 0.08487360132615002, + "grad_norm": 0.26766541600227356, + "learning_rate": 4.576046415250725e-06, + "loss": 0.5597, + "step": 1024 + }, + { + "epoch": 0.08495648570244509, + "grad_norm": 0.2588466703891754, + "learning_rate": 4.57563199336925e-06, + "loss": 0.5196, + "step": 1025 + }, + { + "epoch": 0.08503937007874016, + "grad_norm": 0.28120097517967224, + "learning_rate": 4.575217571487775e-06, + "loss": 0.5522, + "step": 1026 + }, + { + "epoch": 0.08512225445503523, + "grad_norm": 0.315326064825058, + "learning_rate": 4.5748031496062995e-06, + "loss": 0.5901, + "step": 1027 + }, + { + "epoch": 0.0852051388313303, + "grad_norm": 0.2508055567741394, + "learning_rate": 4.574388727724825e-06, + "loss": 0.5235, + "step": 1028 + }, + { + "epoch": 0.08528802320762537, + "grad_norm": 0.31172192096710205, + "learning_rate": 4.573974305843349e-06, + "loss": 0.61, + "step": 1029 + }, + { + "epoch": 0.08537090758392044, + "grad_norm": 0.26646772027015686, + "learning_rate": 4.573559883961874e-06, + "loss": 0.5762, + "step": 1030 + }, + { + "epoch": 0.0854537919602155, + "grad_norm": 0.30158179998397827, + "learning_rate": 4.573145462080398e-06, + "loss": 0.5604, + "step": 1031 + }, + { + "epoch": 0.08553667633651056, + "grad_norm": 0.269031822681427, + "learning_rate": 4.5727310401989225e-06, + "loss": 0.5499, + "step": 1032 + }, + { + "epoch": 0.08561956071280563, + "grad_norm": 0.2630411982536316, + "learning_rate": 4.572316618317448e-06, + "loss": 0.5441, + "step": 1033 + }, + { + "epoch": 0.0857024450891007, + "grad_norm": 0.29011544585227966, + "learning_rate": 4.571902196435972e-06, + "loss": 0.6039, + "step": 1034 + }, + { + "epoch": 0.08578532946539577, + "grad_norm": 0.2786533832550049, + "learning_rate": 4.571487774554497e-06, + "loss": 0.5992, + "step": 1035 + }, + { + "epoch": 0.08586821384169084, + "grad_norm": 0.2769410014152527, + "learning_rate": 4.571073352673022e-06, + "loss": 0.5781, + "step": 1036 + }, + { + "epoch": 0.08595109821798591, + "grad_norm": 0.2790321707725525, + "learning_rate": 4.570658930791546e-06, + "loss": 0.5749, + "step": 1037 + }, + { + "epoch": 0.08603398259428098, + "grad_norm": 0.28105148673057556, + "learning_rate": 4.570244508910071e-06, + "loss": 0.597, + "step": 1038 + }, + { + "epoch": 0.08611686697057605, + "grad_norm": 0.30630236864089966, + "learning_rate": 4.569830087028595e-06, + "loss": 0.5933, + "step": 1039 + }, + { + "epoch": 0.08619975134687112, + "grad_norm": 0.2883754074573517, + "learning_rate": 4.56941566514712e-06, + "loss": 0.6293, + "step": 1040 + }, + { + "epoch": 0.08628263572316619, + "grad_norm": 0.2791236340999603, + "learning_rate": 4.569001243265645e-06, + "loss": 0.5779, + "step": 1041 + }, + { + "epoch": 0.08636552009946125, + "grad_norm": 0.26828500628471375, + "learning_rate": 4.568586821384169e-06, + "loss": 0.6041, + "step": 1042 + }, + { + "epoch": 0.08644840447575632, + "grad_norm": 0.24561476707458496, + "learning_rate": 4.568172399502694e-06, + "loss": 0.5361, + "step": 1043 + }, + { + "epoch": 0.08653128885205139, + "grad_norm": 0.267452210187912, + "learning_rate": 4.567757977621219e-06, + "loss": 0.5419, + "step": 1044 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 0.2671384811401367, + "learning_rate": 4.5673435557397435e-06, + "loss": 0.5435, + "step": 1045 + }, + { + "epoch": 0.08669705760464153, + "grad_norm": 0.2944224774837494, + "learning_rate": 4.566929133858268e-06, + "loss": 0.5959, + "step": 1046 + }, + { + "epoch": 0.0867799419809366, + "grad_norm": 0.26553580164909363, + "learning_rate": 4.566514711976792e-06, + "loss": 0.5441, + "step": 1047 + }, + { + "epoch": 0.08686282635723167, + "grad_norm": 0.27637356519699097, + "learning_rate": 4.566100290095317e-06, + "loss": 0.6013, + "step": 1048 + }, + { + "epoch": 0.08694571073352673, + "grad_norm": 0.26142922043800354, + "learning_rate": 4.565685868213842e-06, + "loss": 0.5856, + "step": 1049 + }, + { + "epoch": 0.0870285951098218, + "grad_norm": 0.26578739285469055, + "learning_rate": 4.5652714463323665e-06, + "loss": 0.5707, + "step": 1050 + }, + { + "epoch": 0.08711147948611686, + "grad_norm": 0.2637694776058197, + "learning_rate": 4.564857024450892e-06, + "loss": 0.577, + "step": 1051 + }, + { + "epoch": 0.08719436386241193, + "grad_norm": 0.2739333212375641, + "learning_rate": 4.564442602569416e-06, + "loss": 0.6245, + "step": 1052 + }, + { + "epoch": 0.087277248238707, + "grad_norm": 0.2861277461051941, + "learning_rate": 4.564028180687941e-06, + "loss": 0.5996, + "step": 1053 + }, + { + "epoch": 0.08736013261500207, + "grad_norm": 0.28271612524986267, + "learning_rate": 4.563613758806465e-06, + "loss": 0.5414, + "step": 1054 + }, + { + "epoch": 0.08744301699129714, + "grad_norm": 0.2722260057926178, + "learning_rate": 4.5631993369249895e-06, + "loss": 0.5298, + "step": 1055 + }, + { + "epoch": 0.08752590136759221, + "grad_norm": 0.2621467709541321, + "learning_rate": 4.562784915043515e-06, + "loss": 0.5014, + "step": 1056 + }, + { + "epoch": 0.08760878574388728, + "grad_norm": 0.2813078761100769, + "learning_rate": 4.562370493162039e-06, + "loss": 0.5313, + "step": 1057 + }, + { + "epoch": 0.08769167012018235, + "grad_norm": 0.2677317261695862, + "learning_rate": 4.561956071280564e-06, + "loss": 0.5355, + "step": 1058 + }, + { + "epoch": 0.08777455449647742, + "grad_norm": 0.2688811719417572, + "learning_rate": 4.561541649399089e-06, + "loss": 0.5507, + "step": 1059 + }, + { + "epoch": 0.08785743887277249, + "grad_norm": 0.288449227809906, + "learning_rate": 4.561127227517613e-06, + "loss": 0.5586, + "step": 1060 + }, + { + "epoch": 0.08794032324906755, + "grad_norm": 0.3154754340648651, + "learning_rate": 4.560712805636138e-06, + "loss": 0.56, + "step": 1061 + }, + { + "epoch": 0.08802320762536262, + "grad_norm": 0.2727501392364502, + "learning_rate": 4.560298383754663e-06, + "loss": 0.5395, + "step": 1062 + }, + { + "epoch": 0.08810609200165768, + "grad_norm": 0.31795254349708557, + "learning_rate": 4.5598839618731875e-06, + "loss": 0.6231, + "step": 1063 + }, + { + "epoch": 0.08818897637795275, + "grad_norm": 0.2832825183868408, + "learning_rate": 4.559469539991712e-06, + "loss": 0.5804, + "step": 1064 + }, + { + "epoch": 0.08827186075424782, + "grad_norm": 0.4116249680519104, + "learning_rate": 4.559055118110236e-06, + "loss": 0.5857, + "step": 1065 + }, + { + "epoch": 0.0883547451305429, + "grad_norm": 0.26260659098625183, + "learning_rate": 4.558640696228761e-06, + "loss": 0.5306, + "step": 1066 + }, + { + "epoch": 0.08843762950683796, + "grad_norm": 0.27621713280677795, + "learning_rate": 4.558226274347286e-06, + "loss": 0.5626, + "step": 1067 + }, + { + "epoch": 0.08852051388313303, + "grad_norm": 0.29249587655067444, + "learning_rate": 4.5578118524658105e-06, + "loss": 0.5851, + "step": 1068 + }, + { + "epoch": 0.0886033982594281, + "grad_norm": 0.2757376432418823, + "learning_rate": 4.557397430584336e-06, + "loss": 0.5687, + "step": 1069 + }, + { + "epoch": 0.08868628263572316, + "grad_norm": 0.28210553526878357, + "learning_rate": 4.55698300870286e-06, + "loss": 0.5803, + "step": 1070 + }, + { + "epoch": 0.08876916701201823, + "grad_norm": 0.26478761434555054, + "learning_rate": 4.556568586821385e-06, + "loss": 0.5324, + "step": 1071 + }, + { + "epoch": 0.0888520513883133, + "grad_norm": 0.31786659359931946, + "learning_rate": 4.556154164939909e-06, + "loss": 0.5882, + "step": 1072 + }, + { + "epoch": 0.08893493576460837, + "grad_norm": 0.26940441131591797, + "learning_rate": 4.5557397430584335e-06, + "loss": 0.5673, + "step": 1073 + }, + { + "epoch": 0.08901782014090344, + "grad_norm": 0.27700111269950867, + "learning_rate": 4.555325321176958e-06, + "loss": 0.5776, + "step": 1074 + }, + { + "epoch": 0.08910070451719851, + "grad_norm": 0.2676769196987152, + "learning_rate": 4.554910899295483e-06, + "loss": 0.5899, + "step": 1075 + }, + { + "epoch": 0.08918358889349358, + "grad_norm": 0.28364020586013794, + "learning_rate": 4.554496477414008e-06, + "loss": 0.548, + "step": 1076 + }, + { + "epoch": 0.08926647326978865, + "grad_norm": 0.31934329867362976, + "learning_rate": 4.554082055532533e-06, + "loss": 0.6553, + "step": 1077 + }, + { + "epoch": 0.08934935764608372, + "grad_norm": 0.293793261051178, + "learning_rate": 4.553667633651057e-06, + "loss": 0.5582, + "step": 1078 + }, + { + "epoch": 0.08943224202237879, + "grad_norm": 0.25466522574424744, + "learning_rate": 4.553253211769582e-06, + "loss": 0.5325, + "step": 1079 + }, + { + "epoch": 0.08951512639867384, + "grad_norm": 0.28412508964538574, + "learning_rate": 4.552838789888106e-06, + "loss": 0.571, + "step": 1080 + }, + { + "epoch": 0.08959801077496891, + "grad_norm": 0.2735932767391205, + "learning_rate": 4.552424368006631e-06, + "loss": 0.5749, + "step": 1081 + }, + { + "epoch": 0.08968089515126398, + "grad_norm": 0.28208327293395996, + "learning_rate": 4.552009946125156e-06, + "loss": 0.612, + "step": 1082 + }, + { + "epoch": 0.08976377952755905, + "grad_norm": 0.2711701989173889, + "learning_rate": 4.55159552424368e-06, + "loss": 0.5779, + "step": 1083 + }, + { + "epoch": 0.08984666390385412, + "grad_norm": 0.2650095224380493, + "learning_rate": 4.551181102362205e-06, + "loss": 0.5296, + "step": 1084 + }, + { + "epoch": 0.0899295482801492, + "grad_norm": 0.27746346592903137, + "learning_rate": 4.55076668048073e-06, + "loss": 0.558, + "step": 1085 + }, + { + "epoch": 0.09001243265644426, + "grad_norm": 0.27674880623817444, + "learning_rate": 4.5503522585992545e-06, + "loss": 0.5649, + "step": 1086 + }, + { + "epoch": 0.09009531703273933, + "grad_norm": 0.2958088517189026, + "learning_rate": 4.549937836717779e-06, + "loss": 0.5759, + "step": 1087 + }, + { + "epoch": 0.0901782014090344, + "grad_norm": 0.2672654390335083, + "learning_rate": 4.549523414836303e-06, + "loss": 0.5623, + "step": 1088 + }, + { + "epoch": 0.09026108578532946, + "grad_norm": 0.27991530299186707, + "learning_rate": 4.549108992954828e-06, + "loss": 0.5709, + "step": 1089 + }, + { + "epoch": 0.09034397016162453, + "grad_norm": 0.25761741399765015, + "learning_rate": 4.548694571073353e-06, + "loss": 0.531, + "step": 1090 + }, + { + "epoch": 0.0904268545379196, + "grad_norm": 0.2998071014881134, + "learning_rate": 4.5482801491918776e-06, + "loss": 0.5972, + "step": 1091 + }, + { + "epoch": 0.09050973891421467, + "grad_norm": 0.2751041650772095, + "learning_rate": 4.547865727310402e-06, + "loss": 0.601, + "step": 1092 + }, + { + "epoch": 0.09059262329050974, + "grad_norm": 0.3101927638053894, + "learning_rate": 4.547451305428927e-06, + "loss": 0.6169, + "step": 1093 + }, + { + "epoch": 0.09067550766680481, + "grad_norm": 0.2871125638484955, + "learning_rate": 4.547036883547452e-06, + "loss": 0.5566, + "step": 1094 + }, + { + "epoch": 0.09075839204309988, + "grad_norm": 0.30005136132240295, + "learning_rate": 4.546622461665976e-06, + "loss": 0.6396, + "step": 1095 + }, + { + "epoch": 0.09084127641939495, + "grad_norm": 0.320975124835968, + "learning_rate": 4.546208039784501e-06, + "loss": 0.6102, + "step": 1096 + }, + { + "epoch": 0.09092416079569002, + "grad_norm": 0.27394038438796997, + "learning_rate": 4.545793617903026e-06, + "loss": 0.5703, + "step": 1097 + }, + { + "epoch": 0.09100704517198509, + "grad_norm": 0.29858899116516113, + "learning_rate": 4.54537919602155e-06, + "loss": 0.594, + "step": 1098 + }, + { + "epoch": 0.09108992954828014, + "grad_norm": 0.262188583612442, + "learning_rate": 4.544964774140075e-06, + "loss": 0.5684, + "step": 1099 + }, + { + "epoch": 0.09117281392457521, + "grad_norm": 0.27356046438217163, + "learning_rate": 4.5445503522586e-06, + "loss": 0.6287, + "step": 1100 + }, + { + "epoch": 0.09125569830087028, + "grad_norm": 0.2462005764245987, + "learning_rate": 4.544135930377124e-06, + "loss": 0.5076, + "step": 1101 + }, + { + "epoch": 0.09133858267716535, + "grad_norm": 0.2958691418170929, + "learning_rate": 4.543721508495649e-06, + "loss": 0.594, + "step": 1102 + }, + { + "epoch": 0.09142146705346042, + "grad_norm": 0.27876174449920654, + "learning_rate": 4.543307086614174e-06, + "loss": 0.5746, + "step": 1103 + }, + { + "epoch": 0.09150435142975549, + "grad_norm": 0.27788370847702026, + "learning_rate": 4.5428926647326986e-06, + "loss": 0.5441, + "step": 1104 + }, + { + "epoch": 0.09158723580605056, + "grad_norm": 0.2708550989627838, + "learning_rate": 4.542478242851223e-06, + "loss": 0.5762, + "step": 1105 + }, + { + "epoch": 0.09167012018234563, + "grad_norm": 0.2842223048210144, + "learning_rate": 4.542063820969747e-06, + "loss": 0.5776, + "step": 1106 + }, + { + "epoch": 0.0917530045586407, + "grad_norm": 0.3003634810447693, + "learning_rate": 4.541649399088272e-06, + "loss": 0.5988, + "step": 1107 + }, + { + "epoch": 0.09183588893493576, + "grad_norm": 0.27952998876571655, + "learning_rate": 4.541234977206797e-06, + "loss": 0.5565, + "step": 1108 + }, + { + "epoch": 0.09191877331123083, + "grad_norm": 0.28364428877830505, + "learning_rate": 4.5408205553253216e-06, + "loss": 0.5475, + "step": 1109 + }, + { + "epoch": 0.0920016576875259, + "grad_norm": 0.31848129630088806, + "learning_rate": 4.540406133443846e-06, + "loss": 0.6077, + "step": 1110 + }, + { + "epoch": 0.09208454206382097, + "grad_norm": 0.28897523880004883, + "learning_rate": 4.539991711562371e-06, + "loss": 0.5591, + "step": 1111 + }, + { + "epoch": 0.09216742644011604, + "grad_norm": 0.27914515137672424, + "learning_rate": 4.539577289680896e-06, + "loss": 0.5834, + "step": 1112 + }, + { + "epoch": 0.09225031081641111, + "grad_norm": 0.28321313858032227, + "learning_rate": 4.53916286779942e-06, + "loss": 0.6179, + "step": 1113 + }, + { + "epoch": 0.09233319519270618, + "grad_norm": 0.29470911622047424, + "learning_rate": 4.5387484459179446e-06, + "loss": 0.5949, + "step": 1114 + }, + { + "epoch": 0.09241607956900125, + "grad_norm": 0.30550917983055115, + "learning_rate": 4.538334024036469e-06, + "loss": 0.6034, + "step": 1115 + }, + { + "epoch": 0.09249896394529632, + "grad_norm": 0.2884824275970459, + "learning_rate": 4.537919602154994e-06, + "loss": 0.5906, + "step": 1116 + }, + { + "epoch": 0.09258184832159139, + "grad_norm": 0.3159334361553192, + "learning_rate": 4.537505180273519e-06, + "loss": 0.5755, + "step": 1117 + }, + { + "epoch": 0.09266473269788644, + "grad_norm": 0.26967501640319824, + "learning_rate": 4.537090758392044e-06, + "loss": 0.547, + "step": 1118 + }, + { + "epoch": 0.09274761707418151, + "grad_norm": 0.27433112263679504, + "learning_rate": 4.536676336510568e-06, + "loss": 0.53, + "step": 1119 + }, + { + "epoch": 0.09283050145047658, + "grad_norm": 0.26114922761917114, + "learning_rate": 4.536261914629093e-06, + "loss": 0.5922, + "step": 1120 + }, + { + "epoch": 0.09291338582677165, + "grad_norm": 0.2872995436191559, + "learning_rate": 4.535847492747617e-06, + "loss": 0.5746, + "step": 1121 + }, + { + "epoch": 0.09299627020306672, + "grad_norm": 0.2645280063152313, + "learning_rate": 4.535433070866142e-06, + "loss": 0.5544, + "step": 1122 + }, + { + "epoch": 0.09307915457936179, + "grad_norm": 0.2967548370361328, + "learning_rate": 4.535018648984667e-06, + "loss": 0.5972, + "step": 1123 + }, + { + "epoch": 0.09316203895565686, + "grad_norm": 0.29996487498283386, + "learning_rate": 4.5346042271031914e-06, + "loss": 0.5641, + "step": 1124 + }, + { + "epoch": 0.09324492333195193, + "grad_norm": 0.2973060607910156, + "learning_rate": 4.534189805221716e-06, + "loss": 0.6175, + "step": 1125 + }, + { + "epoch": 0.093327807708247, + "grad_norm": 0.2722616493701935, + "learning_rate": 4.533775383340241e-06, + "loss": 0.5313, + "step": 1126 + }, + { + "epoch": 0.09341069208454206, + "grad_norm": 0.2834742069244385, + "learning_rate": 4.5333609614587656e-06, + "loss": 0.5926, + "step": 1127 + }, + { + "epoch": 0.09349357646083713, + "grad_norm": 0.2555971145629883, + "learning_rate": 4.53294653957729e-06, + "loss": 0.5626, + "step": 1128 + }, + { + "epoch": 0.0935764608371322, + "grad_norm": 0.28886839747428894, + "learning_rate": 4.5325321176958144e-06, + "loss": 0.5909, + "step": 1129 + }, + { + "epoch": 0.09365934521342727, + "grad_norm": 0.2674325406551361, + "learning_rate": 4.532117695814339e-06, + "loss": 0.5547, + "step": 1130 + }, + { + "epoch": 0.09374222958972234, + "grad_norm": 0.32185298204421997, + "learning_rate": 4.531703273932864e-06, + "loss": 0.5521, + "step": 1131 + }, + { + "epoch": 0.0938251139660174, + "grad_norm": 0.2602789103984833, + "learning_rate": 4.5312888520513886e-06, + "loss": 0.5613, + "step": 1132 + }, + { + "epoch": 0.09390799834231248, + "grad_norm": 0.2801956236362457, + "learning_rate": 4.530874430169913e-06, + "loss": 0.605, + "step": 1133 + }, + { + "epoch": 0.09399088271860755, + "grad_norm": 0.2564257085323334, + "learning_rate": 4.530460008288438e-06, + "loss": 0.5961, + "step": 1134 + }, + { + "epoch": 0.09407376709490262, + "grad_norm": 0.2712496519088745, + "learning_rate": 4.530045586406963e-06, + "loss": 0.5589, + "step": 1135 + }, + { + "epoch": 0.09415665147119769, + "grad_norm": 0.2907867431640625, + "learning_rate": 4.529631164525487e-06, + "loss": 0.5897, + "step": 1136 + }, + { + "epoch": 0.09423953584749274, + "grad_norm": 0.2477366328239441, + "learning_rate": 4.5292167426440124e-06, + "loss": 0.5397, + "step": 1137 + }, + { + "epoch": 0.09432242022378781, + "grad_norm": 0.29849764704704285, + "learning_rate": 4.528802320762537e-06, + "loss": 0.6163, + "step": 1138 + }, + { + "epoch": 0.09440530460008288, + "grad_norm": 0.28370431065559387, + "learning_rate": 4.528387898881061e-06, + "loss": 0.5845, + "step": 1139 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 0.26540735363960266, + "learning_rate": 4.527973476999586e-06, + "loss": 0.5455, + "step": 1140 + }, + { + "epoch": 0.09457107335267302, + "grad_norm": 0.25845369696617126, + "learning_rate": 4.52755905511811e-06, + "loss": 0.5384, + "step": 1141 + }, + { + "epoch": 0.09465395772896809, + "grad_norm": 0.3100195825099945, + "learning_rate": 4.5271446332366354e-06, + "loss": 0.5981, + "step": 1142 + }, + { + "epoch": 0.09473684210526316, + "grad_norm": 0.27154430747032166, + "learning_rate": 4.52673021135516e-06, + "loss": 0.5746, + "step": 1143 + }, + { + "epoch": 0.09481972648155823, + "grad_norm": 0.3085050880908966, + "learning_rate": 4.526315789473685e-06, + "loss": 0.6221, + "step": 1144 + }, + { + "epoch": 0.0949026108578533, + "grad_norm": 0.2903475761413574, + "learning_rate": 4.5259013675922096e-06, + "loss": 0.5845, + "step": 1145 + }, + { + "epoch": 0.09498549523414836, + "grad_norm": 0.30678510665893555, + "learning_rate": 4.525486945710734e-06, + "loss": 0.5744, + "step": 1146 + }, + { + "epoch": 0.09506837961044343, + "grad_norm": 0.27839425206184387, + "learning_rate": 4.5250725238292584e-06, + "loss": 0.5648, + "step": 1147 + }, + { + "epoch": 0.0951512639867385, + "grad_norm": 0.2899670898914337, + "learning_rate": 4.524658101947783e-06, + "loss": 0.5714, + "step": 1148 + }, + { + "epoch": 0.09523414836303357, + "grad_norm": 0.2847357988357544, + "learning_rate": 4.524243680066308e-06, + "loss": 0.5953, + "step": 1149 + }, + { + "epoch": 0.09531703273932864, + "grad_norm": 0.27972713112831116, + "learning_rate": 4.523829258184833e-06, + "loss": 0.5818, + "step": 1150 + }, + { + "epoch": 0.0953999171156237, + "grad_norm": 0.2508236765861511, + "learning_rate": 4.523414836303357e-06, + "loss": 0.5697, + "step": 1151 + }, + { + "epoch": 0.09548280149191878, + "grad_norm": 0.2892949879169464, + "learning_rate": 4.523000414421882e-06, + "loss": 0.5677, + "step": 1152 + }, + { + "epoch": 0.09556568586821385, + "grad_norm": 0.30133727192878723, + "learning_rate": 4.522585992540407e-06, + "loss": 0.5887, + "step": 1153 + }, + { + "epoch": 0.09564857024450892, + "grad_norm": 0.2758755683898926, + "learning_rate": 4.522171570658931e-06, + "loss": 0.5655, + "step": 1154 + }, + { + "epoch": 0.09573145462080399, + "grad_norm": 0.25881046056747437, + "learning_rate": 4.521757148777456e-06, + "loss": 0.553, + "step": 1155 + }, + { + "epoch": 0.09581433899709904, + "grad_norm": 0.2798384428024292, + "learning_rate": 4.52134272689598e-06, + "loss": 0.5546, + "step": 1156 + }, + { + "epoch": 0.09589722337339411, + "grad_norm": 0.28568652272224426, + "learning_rate": 4.520928305014505e-06, + "loss": 0.6141, + "step": 1157 + }, + { + "epoch": 0.09598010774968918, + "grad_norm": 0.2705405056476593, + "learning_rate": 4.52051388313303e-06, + "loss": 0.5746, + "step": 1158 + }, + { + "epoch": 0.09606299212598425, + "grad_norm": 0.27869680523872375, + "learning_rate": 4.520099461251554e-06, + "loss": 0.5543, + "step": 1159 + }, + { + "epoch": 0.09614587650227932, + "grad_norm": 0.27170437574386597, + "learning_rate": 4.5196850393700794e-06, + "loss": 0.5734, + "step": 1160 + }, + { + "epoch": 0.09622876087857439, + "grad_norm": 0.28404176235198975, + "learning_rate": 4.519270617488604e-06, + "loss": 0.5638, + "step": 1161 + }, + { + "epoch": 0.09631164525486946, + "grad_norm": 0.28037136793136597, + "learning_rate": 4.518856195607128e-06, + "loss": 0.5784, + "step": 1162 + }, + { + "epoch": 0.09639452963116453, + "grad_norm": 0.31098395586013794, + "learning_rate": 4.518441773725653e-06, + "loss": 0.5828, + "step": 1163 + }, + { + "epoch": 0.0964774140074596, + "grad_norm": 0.35060983896255493, + "learning_rate": 4.518027351844177e-06, + "loss": 0.5767, + "step": 1164 + }, + { + "epoch": 0.09656029838375466, + "grad_norm": 0.27058881521224976, + "learning_rate": 4.5176129299627024e-06, + "loss": 0.57, + "step": 1165 + }, + { + "epoch": 0.09664318276004973, + "grad_norm": 0.28838035464286804, + "learning_rate": 4.517198508081227e-06, + "loss": 0.5793, + "step": 1166 + }, + { + "epoch": 0.0967260671363448, + "grad_norm": 0.28020212054252625, + "learning_rate": 4.516784086199752e-06, + "loss": 0.575, + "step": 1167 + }, + { + "epoch": 0.09680895151263987, + "grad_norm": 0.26314809918403625, + "learning_rate": 4.516369664318277e-06, + "loss": 0.557, + "step": 1168 + }, + { + "epoch": 0.09689183588893494, + "grad_norm": 0.2826824486255646, + "learning_rate": 4.515955242436801e-06, + "loss": 0.5725, + "step": 1169 + }, + { + "epoch": 0.09697472026523, + "grad_norm": 0.2918686866760254, + "learning_rate": 4.5155408205553255e-06, + "loss": 0.6102, + "step": 1170 + }, + { + "epoch": 0.09705760464152507, + "grad_norm": 0.28776758909225464, + "learning_rate": 4.51512639867385e-06, + "loss": 0.5567, + "step": 1171 + }, + { + "epoch": 0.09714048901782014, + "grad_norm": 0.25951462984085083, + "learning_rate": 4.514711976792375e-06, + "loss": 0.5553, + "step": 1172 + }, + { + "epoch": 0.09722337339411521, + "grad_norm": 0.25680282711982727, + "learning_rate": 4.5142975549109e-06, + "loss": 0.5343, + "step": 1173 + }, + { + "epoch": 0.09730625777041028, + "grad_norm": 0.33559226989746094, + "learning_rate": 4.513883133029424e-06, + "loss": 0.6141, + "step": 1174 + }, + { + "epoch": 0.09738914214670534, + "grad_norm": 0.29495519399642944, + "learning_rate": 4.513468711147949e-06, + "loss": 0.5956, + "step": 1175 + }, + { + "epoch": 0.09747202652300041, + "grad_norm": 0.3058073818683624, + "learning_rate": 4.513054289266474e-06, + "loss": 0.6028, + "step": 1176 + }, + { + "epoch": 0.09755491089929548, + "grad_norm": 0.2898581922054291, + "learning_rate": 4.512639867384998e-06, + "loss": 0.5612, + "step": 1177 + }, + { + "epoch": 0.09763779527559055, + "grad_norm": 0.27400290966033936, + "learning_rate": 4.5122254455035234e-06, + "loss": 0.5414, + "step": 1178 + }, + { + "epoch": 0.09772067965188562, + "grad_norm": 0.28356269001960754, + "learning_rate": 4.511811023622048e-06, + "loss": 0.5663, + "step": 1179 + }, + { + "epoch": 0.09780356402818069, + "grad_norm": 0.26828888058662415, + "learning_rate": 4.511396601740572e-06, + "loss": 0.5247, + "step": 1180 + }, + { + "epoch": 0.09788644840447576, + "grad_norm": 0.26564741134643555, + "learning_rate": 4.510982179859097e-06, + "loss": 0.5179, + "step": 1181 + }, + { + "epoch": 0.09796933278077083, + "grad_norm": 0.2763494849205017, + "learning_rate": 4.510567757977621e-06, + "loss": 0.5627, + "step": 1182 + }, + { + "epoch": 0.0980522171570659, + "grad_norm": 0.2993323504924774, + "learning_rate": 4.5101533360961465e-06, + "loss": 0.5562, + "step": 1183 + }, + { + "epoch": 0.09813510153336095, + "grad_norm": 0.27075767517089844, + "learning_rate": 4.509738914214671e-06, + "loss": 0.5664, + "step": 1184 + }, + { + "epoch": 0.09821798590965602, + "grad_norm": 0.2780574858188629, + "learning_rate": 4.509324492333196e-06, + "loss": 0.566, + "step": 1185 + }, + { + "epoch": 0.0983008702859511, + "grad_norm": 0.2831488847732544, + "learning_rate": 4.508910070451721e-06, + "loss": 0.5917, + "step": 1186 + }, + { + "epoch": 0.09838375466224616, + "grad_norm": 0.3007243573665619, + "learning_rate": 4.508495648570245e-06, + "loss": 0.6082, + "step": 1187 + }, + { + "epoch": 0.09846663903854123, + "grad_norm": 0.2817358374595642, + "learning_rate": 4.5080812266887695e-06, + "loss": 0.5948, + "step": 1188 + }, + { + "epoch": 0.0985495234148363, + "grad_norm": 0.2820909321308136, + "learning_rate": 4.507666804807294e-06, + "loss": 0.5543, + "step": 1189 + }, + { + "epoch": 0.09863240779113137, + "grad_norm": 0.2779269814491272, + "learning_rate": 4.507252382925819e-06, + "loss": 0.576, + "step": 1190 + }, + { + "epoch": 0.09871529216742644, + "grad_norm": 0.2628866732120514, + "learning_rate": 4.506837961044344e-06, + "loss": 0.5568, + "step": 1191 + }, + { + "epoch": 0.09879817654372151, + "grad_norm": 0.2944766879081726, + "learning_rate": 4.506423539162868e-06, + "loss": 0.606, + "step": 1192 + }, + { + "epoch": 0.09888106092001658, + "grad_norm": 0.2575654089450836, + "learning_rate": 4.506009117281393e-06, + "loss": 0.5169, + "step": 1193 + }, + { + "epoch": 0.09896394529631164, + "grad_norm": 0.307539701461792, + "learning_rate": 4.505594695399918e-06, + "loss": 0.5073, + "step": 1194 + }, + { + "epoch": 0.09904682967260671, + "grad_norm": 0.2620030641555786, + "learning_rate": 4.505180273518442e-06, + "loss": 0.5397, + "step": 1195 + }, + { + "epoch": 0.09912971404890178, + "grad_norm": 0.2816161513328552, + "learning_rate": 4.504765851636967e-06, + "loss": 0.5527, + "step": 1196 + }, + { + "epoch": 0.09921259842519685, + "grad_norm": 0.2792831361293793, + "learning_rate": 4.504351429755491e-06, + "loss": 0.5504, + "step": 1197 + }, + { + "epoch": 0.09929548280149192, + "grad_norm": 0.27923107147216797, + "learning_rate": 4.503937007874016e-06, + "loss": 0.562, + "step": 1198 + }, + { + "epoch": 0.09937836717778699, + "grad_norm": 0.2834323048591614, + "learning_rate": 4.503522585992541e-06, + "loss": 0.6092, + "step": 1199 + }, + { + "epoch": 0.09946125155408206, + "grad_norm": 0.28433486819267273, + "learning_rate": 4.503108164111065e-06, + "loss": 0.5659, + "step": 1200 + }, + { + "epoch": 0.09954413593037713, + "grad_norm": 0.28686895966529846, + "learning_rate": 4.5026937422295905e-06, + "loss": 0.5545, + "step": 1201 + }, + { + "epoch": 0.0996270203066722, + "grad_norm": 0.2881747782230377, + "learning_rate": 4.502279320348115e-06, + "loss": 0.538, + "step": 1202 + }, + { + "epoch": 0.09970990468296725, + "grad_norm": 0.2764453887939453, + "learning_rate": 4.501864898466639e-06, + "loss": 0.5581, + "step": 1203 + }, + { + "epoch": 0.09979278905926232, + "grad_norm": 0.30853500962257385, + "learning_rate": 4.501450476585164e-06, + "loss": 0.6339, + "step": 1204 + }, + { + "epoch": 0.0998756734355574, + "grad_norm": 0.2907939851284027, + "learning_rate": 4.501036054703688e-06, + "loss": 0.6076, + "step": 1205 + }, + { + "epoch": 0.09995855781185246, + "grad_norm": 0.302060067653656, + "learning_rate": 4.5006216328222135e-06, + "loss": 0.5832, + "step": 1206 + }, + { + "epoch": 0.10004144218814753, + "grad_norm": 0.28499895334243774, + "learning_rate": 4.500207210940738e-06, + "loss": 0.5543, + "step": 1207 + }, + { + "epoch": 0.1001243265644426, + "grad_norm": 0.28235045075416565, + "learning_rate": 4.499792789059262e-06, + "loss": 0.5526, + "step": 1208 + }, + { + "epoch": 0.10020721094073767, + "grad_norm": 0.27799031138420105, + "learning_rate": 4.499378367177788e-06, + "loss": 0.5403, + "step": 1209 + }, + { + "epoch": 0.10029009531703274, + "grad_norm": 0.2957574129104614, + "learning_rate": 4.498963945296312e-06, + "loss": 0.577, + "step": 1210 + }, + { + "epoch": 0.10037297969332781, + "grad_norm": 0.29267942905426025, + "learning_rate": 4.4985495234148365e-06, + "loss": 0.5514, + "step": 1211 + }, + { + "epoch": 0.10045586406962288, + "grad_norm": 0.3260367214679718, + "learning_rate": 4.498135101533361e-06, + "loss": 0.5739, + "step": 1212 + }, + { + "epoch": 0.10053874844591794, + "grad_norm": 0.2852626144886017, + "learning_rate": 4.497720679651885e-06, + "loss": 0.5823, + "step": 1213 + }, + { + "epoch": 0.10062163282221301, + "grad_norm": 0.27347832918167114, + "learning_rate": 4.497306257770411e-06, + "loss": 0.5671, + "step": 1214 + }, + { + "epoch": 0.10070451719850808, + "grad_norm": 0.2735826373100281, + "learning_rate": 4.496891835888935e-06, + "loss": 0.5604, + "step": 1215 + }, + { + "epoch": 0.10078740157480315, + "grad_norm": 0.289502888917923, + "learning_rate": 4.49647741400746e-06, + "loss": 0.5277, + "step": 1216 + }, + { + "epoch": 0.10087028595109822, + "grad_norm": 0.2696060836315155, + "learning_rate": 4.496062992125985e-06, + "loss": 0.5138, + "step": 1217 + }, + { + "epoch": 0.10095317032739329, + "grad_norm": 0.2632719874382019, + "learning_rate": 4.495648570244509e-06, + "loss": 0.5718, + "step": 1218 + }, + { + "epoch": 0.10103605470368836, + "grad_norm": 0.2979549467563629, + "learning_rate": 4.4952341483630345e-06, + "loss": 0.6075, + "step": 1219 + }, + { + "epoch": 0.10111893907998343, + "grad_norm": 0.29637011885643005, + "learning_rate": 4.494819726481559e-06, + "loss": 0.5978, + "step": 1220 + }, + { + "epoch": 0.1012018234562785, + "grad_norm": 0.2861822843551636, + "learning_rate": 4.494405304600083e-06, + "loss": 0.5513, + "step": 1221 + }, + { + "epoch": 0.10128470783257355, + "grad_norm": 0.2883451282978058, + "learning_rate": 4.493990882718608e-06, + "loss": 0.5758, + "step": 1222 + }, + { + "epoch": 0.10136759220886862, + "grad_norm": 0.2816309630870819, + "learning_rate": 4.493576460837132e-06, + "loss": 0.5666, + "step": 1223 + }, + { + "epoch": 0.10145047658516369, + "grad_norm": 0.29393309354782104, + "learning_rate": 4.4931620389556575e-06, + "loss": 0.5643, + "step": 1224 + }, + { + "epoch": 0.10153336096145876, + "grad_norm": 0.293396919965744, + "learning_rate": 4.492747617074182e-06, + "loss": 0.6319, + "step": 1225 + }, + { + "epoch": 0.10161624533775383, + "grad_norm": 0.27858009934425354, + "learning_rate": 4.492333195192706e-06, + "loss": 0.5983, + "step": 1226 + }, + { + "epoch": 0.1016991297140489, + "grad_norm": 0.29866108298301697, + "learning_rate": 4.491918773311232e-06, + "loss": 0.5616, + "step": 1227 + }, + { + "epoch": 0.10178201409034397, + "grad_norm": 0.2803502380847931, + "learning_rate": 4.491504351429756e-06, + "loss": 0.55, + "step": 1228 + }, + { + "epoch": 0.10186489846663904, + "grad_norm": 0.28791725635528564, + "learning_rate": 4.4910899295482805e-06, + "loss": 0.5188, + "step": 1229 + }, + { + "epoch": 0.10194778284293411, + "grad_norm": 0.29803526401519775, + "learning_rate": 4.490675507666805e-06, + "loss": 0.6098, + "step": 1230 + }, + { + "epoch": 0.10203066721922918, + "grad_norm": 0.26283136010169983, + "learning_rate": 4.490261085785329e-06, + "loss": 0.5519, + "step": 1231 + }, + { + "epoch": 0.10211355159552424, + "grad_norm": 0.28881463408470154, + "learning_rate": 4.489846663903855e-06, + "loss": 0.5933, + "step": 1232 + }, + { + "epoch": 0.10219643597181931, + "grad_norm": 0.2808476686477661, + "learning_rate": 4.489432242022379e-06, + "loss": 0.5869, + "step": 1233 + }, + { + "epoch": 0.10227932034811438, + "grad_norm": 0.2790272831916809, + "learning_rate": 4.489017820140904e-06, + "loss": 0.5628, + "step": 1234 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 0.27363938093185425, + "learning_rate": 4.488603398259429e-06, + "loss": 0.5658, + "step": 1235 + }, + { + "epoch": 0.10244508910070452, + "grad_norm": 0.28898462653160095, + "learning_rate": 4.488188976377953e-06, + "loss": 0.6133, + "step": 1236 + }, + { + "epoch": 0.10252797347699959, + "grad_norm": 0.2550761103630066, + "learning_rate": 4.487774554496478e-06, + "loss": 0.5567, + "step": 1237 + }, + { + "epoch": 0.10261085785329466, + "grad_norm": 0.27230104804039, + "learning_rate": 4.487360132615002e-06, + "loss": 0.5505, + "step": 1238 + }, + { + "epoch": 0.10269374222958973, + "grad_norm": 0.31176063418388367, + "learning_rate": 4.486945710733527e-06, + "loss": 0.6151, + "step": 1239 + }, + { + "epoch": 0.1027766266058848, + "grad_norm": 0.28337663412094116, + "learning_rate": 4.486531288852052e-06, + "loss": 0.5669, + "step": 1240 + }, + { + "epoch": 0.10285951098217985, + "grad_norm": 0.29904067516326904, + "learning_rate": 4.486116866970576e-06, + "loss": 0.5931, + "step": 1241 + }, + { + "epoch": 0.10294239535847492, + "grad_norm": 0.2761286497116089, + "learning_rate": 4.4857024450891015e-06, + "loss": 0.5572, + "step": 1242 + }, + { + "epoch": 0.10302527973476999, + "grad_norm": 0.272738516330719, + "learning_rate": 4.485288023207626e-06, + "loss": 0.5748, + "step": 1243 + }, + { + "epoch": 0.10310816411106506, + "grad_norm": 0.25870612263679504, + "learning_rate": 4.48487360132615e-06, + "loss": 0.5734, + "step": 1244 + }, + { + "epoch": 0.10319104848736013, + "grad_norm": 0.26987653970718384, + "learning_rate": 4.484459179444675e-06, + "loss": 0.5545, + "step": 1245 + }, + { + "epoch": 0.1032739328636552, + "grad_norm": 0.2886475622653961, + "learning_rate": 4.484044757563199e-06, + "loss": 0.5658, + "step": 1246 + }, + { + "epoch": 0.10335681723995027, + "grad_norm": 0.2814441919326782, + "learning_rate": 4.4836303356817245e-06, + "loss": 0.5584, + "step": 1247 + }, + { + "epoch": 0.10343970161624534, + "grad_norm": 0.2735563814640045, + "learning_rate": 4.483215913800249e-06, + "loss": 0.6044, + "step": 1248 + }, + { + "epoch": 0.10352258599254041, + "grad_norm": 0.2799004316329956, + "learning_rate": 4.482801491918773e-06, + "loss": 0.6264, + "step": 1249 + }, + { + "epoch": 0.10360547036883548, + "grad_norm": 0.29487448930740356, + "learning_rate": 4.482387070037299e-06, + "loss": 0.626, + "step": 1250 + }, + { + "epoch": 0.10368835474513054, + "grad_norm": 0.2584938704967499, + "learning_rate": 4.481972648155823e-06, + "loss": 0.5714, + "step": 1251 + }, + { + "epoch": 0.1037712391214256, + "grad_norm": 0.2826838195323944, + "learning_rate": 4.4815582262743475e-06, + "loss": 0.5352, + "step": 1252 + }, + { + "epoch": 0.10385412349772068, + "grad_norm": 0.2591056823730469, + "learning_rate": 4.481143804392872e-06, + "loss": 0.5035, + "step": 1253 + }, + { + "epoch": 0.10393700787401575, + "grad_norm": 0.2916572690010071, + "learning_rate": 4.480729382511396e-06, + "loss": 0.5806, + "step": 1254 + }, + { + "epoch": 0.10401989225031082, + "grad_norm": 0.2760732173919678, + "learning_rate": 4.480314960629922e-06, + "loss": 0.5583, + "step": 1255 + }, + { + "epoch": 0.10410277662660589, + "grad_norm": 0.2557518482208252, + "learning_rate": 4.479900538748446e-06, + "loss": 0.5379, + "step": 1256 + }, + { + "epoch": 0.10418566100290096, + "grad_norm": 0.2551189661026001, + "learning_rate": 4.479486116866971e-06, + "loss": 0.5059, + "step": 1257 + }, + { + "epoch": 0.10426854537919603, + "grad_norm": 0.2905339002609253, + "learning_rate": 4.479071694985496e-06, + "loss": 0.6238, + "step": 1258 + }, + { + "epoch": 0.1043514297554911, + "grad_norm": 0.31372153759002686, + "learning_rate": 4.47865727310402e-06, + "loss": 0.5707, + "step": 1259 + }, + { + "epoch": 0.10443431413178615, + "grad_norm": 0.2790207266807556, + "learning_rate": 4.4782428512225455e-06, + "loss": 0.541, + "step": 1260 + }, + { + "epoch": 0.10451719850808122, + "grad_norm": 0.2743843197822571, + "learning_rate": 4.47782842934107e-06, + "loss": 0.5342, + "step": 1261 + }, + { + "epoch": 0.10460008288437629, + "grad_norm": 0.2828810513019562, + "learning_rate": 4.477414007459594e-06, + "loss": 0.5282, + "step": 1262 + }, + { + "epoch": 0.10468296726067136, + "grad_norm": 0.4352573752403259, + "learning_rate": 4.476999585578119e-06, + "loss": 0.5363, + "step": 1263 + }, + { + "epoch": 0.10476585163696643, + "grad_norm": 0.28512144088745117, + "learning_rate": 4.476585163696643e-06, + "loss": 0.6123, + "step": 1264 + }, + { + "epoch": 0.1048487360132615, + "grad_norm": 0.26069334149360657, + "learning_rate": 4.4761707418151685e-06, + "loss": 0.5776, + "step": 1265 + }, + { + "epoch": 0.10493162038955657, + "grad_norm": 0.2713608145713806, + "learning_rate": 4.475756319933693e-06, + "loss": 0.5187, + "step": 1266 + }, + { + "epoch": 0.10501450476585164, + "grad_norm": 0.3297128677368164, + "learning_rate": 4.475341898052217e-06, + "loss": 0.5717, + "step": 1267 + }, + { + "epoch": 0.10509738914214671, + "grad_norm": 0.2836648225784302, + "learning_rate": 4.474927476170743e-06, + "loss": 0.629, + "step": 1268 + }, + { + "epoch": 0.10518027351844178, + "grad_norm": 0.2710469663143158, + "learning_rate": 4.474513054289267e-06, + "loss": 0.5753, + "step": 1269 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 0.27301645278930664, + "learning_rate": 4.4740986324077915e-06, + "loss": 0.5305, + "step": 1270 + }, + { + "epoch": 0.1053460422710319, + "grad_norm": 0.274749755859375, + "learning_rate": 4.473684210526316e-06, + "loss": 0.551, + "step": 1271 + }, + { + "epoch": 0.10542892664732698, + "grad_norm": 0.33061015605926514, + "learning_rate": 4.47326978864484e-06, + "loss": 0.6136, + "step": 1272 + }, + { + "epoch": 0.10551181102362205, + "grad_norm": 0.2797594964504242, + "learning_rate": 4.472855366763366e-06, + "loss": 0.5514, + "step": 1273 + }, + { + "epoch": 0.10559469539991712, + "grad_norm": 0.2636263966560364, + "learning_rate": 4.47244094488189e-06, + "loss": 0.5901, + "step": 1274 + }, + { + "epoch": 0.10567757977621219, + "grad_norm": 0.27710703015327454, + "learning_rate": 4.4720265230004145e-06, + "loss": 0.5579, + "step": 1275 + }, + { + "epoch": 0.10576046415250726, + "grad_norm": 0.2834899127483368, + "learning_rate": 4.47161210111894e-06, + "loss": 0.5406, + "step": 1276 + }, + { + "epoch": 0.10584334852880233, + "grad_norm": 0.29359129071235657, + "learning_rate": 4.471197679237464e-06, + "loss": 0.5733, + "step": 1277 + }, + { + "epoch": 0.1059262329050974, + "grad_norm": 0.2727518677711487, + "learning_rate": 4.470783257355989e-06, + "loss": 0.5895, + "step": 1278 + }, + { + "epoch": 0.10600911728139245, + "grad_norm": 0.2826988697052002, + "learning_rate": 4.470368835474513e-06, + "loss": 0.5718, + "step": 1279 + }, + { + "epoch": 0.10609200165768752, + "grad_norm": 0.29438790678977966, + "learning_rate": 4.4699544135930375e-06, + "loss": 0.5403, + "step": 1280 + }, + { + "epoch": 0.10617488603398259, + "grad_norm": 0.2633553445339203, + "learning_rate": 4.469539991711563e-06, + "loss": 0.5607, + "step": 1281 + }, + { + "epoch": 0.10625777041027766, + "grad_norm": 0.2900015115737915, + "learning_rate": 4.469125569830087e-06, + "loss": 0.5285, + "step": 1282 + }, + { + "epoch": 0.10634065478657273, + "grad_norm": 0.2631588876247406, + "learning_rate": 4.4687111479486125e-06, + "loss": 0.5154, + "step": 1283 + }, + { + "epoch": 0.1064235391628678, + "grad_norm": 0.2913229167461395, + "learning_rate": 4.468296726067137e-06, + "loss": 0.631, + "step": 1284 + }, + { + "epoch": 0.10650642353916287, + "grad_norm": 0.2864161431789398, + "learning_rate": 4.467882304185661e-06, + "loss": 0.5464, + "step": 1285 + }, + { + "epoch": 0.10658930791545794, + "grad_norm": 0.4167289435863495, + "learning_rate": 4.467467882304186e-06, + "loss": 0.4998, + "step": 1286 + }, + { + "epoch": 0.10667219229175301, + "grad_norm": 0.3192503750324249, + "learning_rate": 4.46705346042271e-06, + "loss": 0.6371, + "step": 1287 + }, + { + "epoch": 0.10675507666804808, + "grad_norm": 0.29683718085289, + "learning_rate": 4.4666390385412355e-06, + "loss": 0.575, + "step": 1288 + }, + { + "epoch": 0.10683796104434314, + "grad_norm": 0.2870124876499176, + "learning_rate": 4.46622461665976e-06, + "loss": 0.552, + "step": 1289 + }, + { + "epoch": 0.1069208454206382, + "grad_norm": 0.25112152099609375, + "learning_rate": 4.465810194778284e-06, + "loss": 0.5255, + "step": 1290 + }, + { + "epoch": 0.10700372979693328, + "grad_norm": 0.29015010595321655, + "learning_rate": 4.46539577289681e-06, + "loss": 0.6046, + "step": 1291 + }, + { + "epoch": 0.10708661417322834, + "grad_norm": 0.3053722083568573, + "learning_rate": 4.464981351015334e-06, + "loss": 0.6005, + "step": 1292 + }, + { + "epoch": 0.10716949854952341, + "grad_norm": 0.2865354120731354, + "learning_rate": 4.4645669291338585e-06, + "loss": 0.6295, + "step": 1293 + }, + { + "epoch": 0.10725238292581848, + "grad_norm": 0.30435192584991455, + "learning_rate": 4.464152507252383e-06, + "loss": 0.5775, + "step": 1294 + }, + { + "epoch": 0.10733526730211355, + "grad_norm": 0.28944721817970276, + "learning_rate": 4.463738085370908e-06, + "loss": 0.5988, + "step": 1295 + }, + { + "epoch": 0.10741815167840862, + "grad_norm": 0.2614794969558716, + "learning_rate": 4.463323663489433e-06, + "loss": 0.579, + "step": 1296 + }, + { + "epoch": 0.1075010360547037, + "grad_norm": 0.28510162234306335, + "learning_rate": 4.462909241607957e-06, + "loss": 0.5759, + "step": 1297 + }, + { + "epoch": 0.10758392043099875, + "grad_norm": 0.2858985662460327, + "learning_rate": 4.4624948197264815e-06, + "loss": 0.5482, + "step": 1298 + }, + { + "epoch": 0.10766680480729382, + "grad_norm": 0.31536534428596497, + "learning_rate": 4.462080397845007e-06, + "loss": 0.6111, + "step": 1299 + }, + { + "epoch": 0.10774968918358889, + "grad_norm": 0.25793635845184326, + "learning_rate": 4.461665975963531e-06, + "loss": 0.5547, + "step": 1300 + }, + { + "epoch": 0.10783257355988396, + "grad_norm": 0.27405041456222534, + "learning_rate": 4.4612515540820565e-06, + "loss": 0.538, + "step": 1301 + }, + { + "epoch": 0.10791545793617903, + "grad_norm": 0.309801310300827, + "learning_rate": 4.460837132200581e-06, + "loss": 0.5831, + "step": 1302 + }, + { + "epoch": 0.1079983423124741, + "grad_norm": 0.2918725311756134, + "learning_rate": 4.460422710319105e-06, + "loss": 0.5844, + "step": 1303 + }, + { + "epoch": 0.10808122668876917, + "grad_norm": 0.3142109215259552, + "learning_rate": 4.46000828843763e-06, + "loss": 0.6179, + "step": 1304 + }, + { + "epoch": 0.10816411106506424, + "grad_norm": 0.28993940353393555, + "learning_rate": 4.459593866556154e-06, + "loss": 0.5419, + "step": 1305 + }, + { + "epoch": 0.10824699544135931, + "grad_norm": 0.2768542170524597, + "learning_rate": 4.4591794446746795e-06, + "loss": 0.5606, + "step": 1306 + }, + { + "epoch": 0.10832987981765438, + "grad_norm": 0.2712918221950531, + "learning_rate": 4.458765022793204e-06, + "loss": 0.5326, + "step": 1307 + }, + { + "epoch": 0.10841276419394943, + "grad_norm": 0.3206329047679901, + "learning_rate": 4.458350600911728e-06, + "loss": 0.6053, + "step": 1308 + }, + { + "epoch": 0.1084956485702445, + "grad_norm": 0.2656137943267822, + "learning_rate": 4.457936179030254e-06, + "loss": 0.54, + "step": 1309 + }, + { + "epoch": 0.10857853294653957, + "grad_norm": 0.2830184996128082, + "learning_rate": 4.457521757148778e-06, + "loss": 0.5701, + "step": 1310 + }, + { + "epoch": 0.10866141732283464, + "grad_norm": 0.2985686957836151, + "learning_rate": 4.4571073352673025e-06, + "loss": 0.5869, + "step": 1311 + }, + { + "epoch": 0.10874430169912971, + "grad_norm": 0.3056812882423401, + "learning_rate": 4.456692913385827e-06, + "loss": 0.6033, + "step": 1312 + }, + { + "epoch": 0.10882718607542478, + "grad_norm": 0.2746906280517578, + "learning_rate": 4.456278491504351e-06, + "loss": 0.5139, + "step": 1313 + }, + { + "epoch": 0.10891007045171985, + "grad_norm": 0.27164915204048157, + "learning_rate": 4.455864069622877e-06, + "loss": 0.5718, + "step": 1314 + }, + { + "epoch": 0.10899295482801492, + "grad_norm": 0.31144529581069946, + "learning_rate": 4.455449647741401e-06, + "loss": 0.5943, + "step": 1315 + }, + { + "epoch": 0.10907583920431, + "grad_norm": 0.2779950797557831, + "learning_rate": 4.4550352258599255e-06, + "loss": 0.577, + "step": 1316 + }, + { + "epoch": 0.10915872358060505, + "grad_norm": 0.3148849606513977, + "learning_rate": 4.454620803978451e-06, + "loss": 0.5747, + "step": 1317 + }, + { + "epoch": 0.10924160795690012, + "grad_norm": 0.29582831263542175, + "learning_rate": 4.454206382096975e-06, + "loss": 0.5648, + "step": 1318 + }, + { + "epoch": 0.10932449233319519, + "grad_norm": 0.2880154252052307, + "learning_rate": 4.4537919602155e-06, + "loss": 0.5698, + "step": 1319 + }, + { + "epoch": 0.10940737670949026, + "grad_norm": 0.2649206221103668, + "learning_rate": 4.453377538334024e-06, + "loss": 0.5223, + "step": 1320 + }, + { + "epoch": 0.10949026108578533, + "grad_norm": 0.2618553936481476, + "learning_rate": 4.4529631164525485e-06, + "loss": 0.5263, + "step": 1321 + }, + { + "epoch": 0.1095731454620804, + "grad_norm": 0.3229990601539612, + "learning_rate": 4.452548694571074e-06, + "loss": 0.5468, + "step": 1322 + }, + { + "epoch": 0.10965602983837547, + "grad_norm": 0.2844531536102295, + "learning_rate": 4.452134272689598e-06, + "loss": 0.5657, + "step": 1323 + }, + { + "epoch": 0.10973891421467054, + "grad_norm": 0.29775670170783997, + "learning_rate": 4.4517198508081235e-06, + "loss": 0.61, + "step": 1324 + }, + { + "epoch": 0.10982179859096561, + "grad_norm": 0.2668749988079071, + "learning_rate": 4.451305428926648e-06, + "loss": 0.5533, + "step": 1325 + }, + { + "epoch": 0.10990468296726068, + "grad_norm": 0.2799801230430603, + "learning_rate": 4.450891007045172e-06, + "loss": 0.5914, + "step": 1326 + }, + { + "epoch": 0.10998756734355573, + "grad_norm": 0.3186360001564026, + "learning_rate": 4.450476585163697e-06, + "loss": 0.6182, + "step": 1327 + }, + { + "epoch": 0.1100704517198508, + "grad_norm": 0.2740856111049652, + "learning_rate": 4.450062163282221e-06, + "loss": 0.5356, + "step": 1328 + }, + { + "epoch": 0.11015333609614587, + "grad_norm": 0.2661397457122803, + "learning_rate": 4.449647741400746e-06, + "loss": 0.5736, + "step": 1329 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 0.28412237763404846, + "learning_rate": 4.449233319519271e-06, + "loss": 0.5726, + "step": 1330 + }, + { + "epoch": 0.11031910484873601, + "grad_norm": 0.3001020848751068, + "learning_rate": 4.448818897637795e-06, + "loss": 0.5333, + "step": 1331 + }, + { + "epoch": 0.11040198922503108, + "grad_norm": 0.273796945810318, + "learning_rate": 4.448404475756321e-06, + "loss": 0.533, + "step": 1332 + }, + { + "epoch": 0.11048487360132615, + "grad_norm": 0.28265902400016785, + "learning_rate": 4.447990053874845e-06, + "loss": 0.6157, + "step": 1333 + }, + { + "epoch": 0.11056775797762122, + "grad_norm": 0.2667519450187683, + "learning_rate": 4.4475756319933695e-06, + "loss": 0.5964, + "step": 1334 + }, + { + "epoch": 0.11065064235391629, + "grad_norm": 0.26582422852516174, + "learning_rate": 4.447161210111894e-06, + "loss": 0.5763, + "step": 1335 + }, + { + "epoch": 0.11073352673021135, + "grad_norm": 0.26707425713539124, + "learning_rate": 4.446746788230419e-06, + "loss": 0.5745, + "step": 1336 + }, + { + "epoch": 0.11081641110650642, + "grad_norm": 0.2821609079837799, + "learning_rate": 4.446332366348944e-06, + "loss": 0.5517, + "step": 1337 + }, + { + "epoch": 0.11089929548280149, + "grad_norm": 0.2834641635417938, + "learning_rate": 4.445917944467468e-06, + "loss": 0.5763, + "step": 1338 + }, + { + "epoch": 0.11098217985909656, + "grad_norm": 0.2795705497264862, + "learning_rate": 4.4455035225859925e-06, + "loss": 0.5606, + "step": 1339 + }, + { + "epoch": 0.11106506423539163, + "grad_norm": 0.3121352195739746, + "learning_rate": 4.445089100704518e-06, + "loss": 0.5384, + "step": 1340 + }, + { + "epoch": 0.1111479486116867, + "grad_norm": 0.28176912665367126, + "learning_rate": 4.444674678823042e-06, + "loss": 0.5479, + "step": 1341 + }, + { + "epoch": 0.11123083298798177, + "grad_norm": 0.29745522141456604, + "learning_rate": 4.4442602569415675e-06, + "loss": 0.5993, + "step": 1342 + }, + { + "epoch": 0.11131371736427684, + "grad_norm": 0.3022138178348541, + "learning_rate": 4.443845835060092e-06, + "loss": 0.6326, + "step": 1343 + }, + { + "epoch": 0.11139660174057191, + "grad_norm": 0.2567228674888611, + "learning_rate": 4.443431413178616e-06, + "loss": 0.5007, + "step": 1344 + }, + { + "epoch": 0.11147948611686698, + "grad_norm": 0.2754369080066681, + "learning_rate": 4.443016991297141e-06, + "loss": 0.5528, + "step": 1345 + }, + { + "epoch": 0.11156237049316203, + "grad_norm": 0.2854289412498474, + "learning_rate": 4.442602569415665e-06, + "loss": 0.5805, + "step": 1346 + }, + { + "epoch": 0.1116452548694571, + "grad_norm": 0.2632337808609009, + "learning_rate": 4.44218814753419e-06, + "loss": 0.5314, + "step": 1347 + }, + { + "epoch": 0.11172813924575217, + "grad_norm": 0.2720893323421478, + "learning_rate": 4.441773725652715e-06, + "loss": 0.5886, + "step": 1348 + }, + { + "epoch": 0.11181102362204724, + "grad_norm": 0.2882750928401947, + "learning_rate": 4.441359303771239e-06, + "loss": 0.5858, + "step": 1349 + }, + { + "epoch": 0.11189390799834231, + "grad_norm": 0.2603878676891327, + "learning_rate": 4.440944881889765e-06, + "loss": 0.5683, + "step": 1350 + }, + { + "epoch": 0.11197679237463738, + "grad_norm": 0.3062310814857483, + "learning_rate": 4.440530460008289e-06, + "loss": 0.6111, + "step": 1351 + }, + { + "epoch": 0.11205967675093245, + "grad_norm": 0.24366699159145355, + "learning_rate": 4.4401160381268135e-06, + "loss": 0.5085, + "step": 1352 + }, + { + "epoch": 0.11214256112722752, + "grad_norm": 0.2771233320236206, + "learning_rate": 4.439701616245338e-06, + "loss": 0.5619, + "step": 1353 + }, + { + "epoch": 0.11222544550352259, + "grad_norm": 0.2871256470680237, + "learning_rate": 4.439287194363862e-06, + "loss": 0.5521, + "step": 1354 + }, + { + "epoch": 0.11230832987981765, + "grad_norm": 0.2748681902885437, + "learning_rate": 4.438872772482388e-06, + "loss": 0.5769, + "step": 1355 + }, + { + "epoch": 0.11239121425611272, + "grad_norm": 1.031336784362793, + "learning_rate": 4.438458350600912e-06, + "loss": 0.5932, + "step": 1356 + }, + { + "epoch": 0.11247409863240779, + "grad_norm": 0.29478520154953003, + "learning_rate": 4.4380439287194365e-06, + "loss": 0.5714, + "step": 1357 + }, + { + "epoch": 0.11255698300870286, + "grad_norm": 0.2796001136302948, + "learning_rate": 4.437629506837962e-06, + "loss": 0.5468, + "step": 1358 + }, + { + "epoch": 0.11263986738499793, + "grad_norm": 0.25888970494270325, + "learning_rate": 4.437215084956486e-06, + "loss": 0.5899, + "step": 1359 + }, + { + "epoch": 0.112722751761293, + "grad_norm": 0.26148784160614014, + "learning_rate": 4.436800663075011e-06, + "loss": 0.5783, + "step": 1360 + }, + { + "epoch": 0.11280563613758807, + "grad_norm": 0.2461990863084793, + "learning_rate": 4.436386241193535e-06, + "loss": 0.5539, + "step": 1361 + }, + { + "epoch": 0.11288852051388314, + "grad_norm": 0.26570430397987366, + "learning_rate": 4.4359718193120596e-06, + "loss": 0.5507, + "step": 1362 + }, + { + "epoch": 0.1129714048901782, + "grad_norm": 0.27796536684036255, + "learning_rate": 4.435557397430585e-06, + "loss": 0.5694, + "step": 1363 + }, + { + "epoch": 0.11305428926647328, + "grad_norm": 0.2918712794780731, + "learning_rate": 4.435142975549109e-06, + "loss": 0.5126, + "step": 1364 + }, + { + "epoch": 0.11313717364276833, + "grad_norm": 0.27244147658348083, + "learning_rate": 4.434728553667634e-06, + "loss": 0.5671, + "step": 1365 + }, + { + "epoch": 0.1132200580190634, + "grad_norm": 0.26977303624153137, + "learning_rate": 4.434314131786159e-06, + "loss": 0.5623, + "step": 1366 + }, + { + "epoch": 0.11330294239535847, + "grad_norm": 0.3012491762638092, + "learning_rate": 4.433899709904683e-06, + "loss": 0.5733, + "step": 1367 + }, + { + "epoch": 0.11338582677165354, + "grad_norm": 0.28007715940475464, + "learning_rate": 4.433485288023208e-06, + "loss": 0.6187, + "step": 1368 + }, + { + "epoch": 0.11346871114794861, + "grad_norm": 0.2780837416648865, + "learning_rate": 4.433070866141732e-06, + "loss": 0.5754, + "step": 1369 + }, + { + "epoch": 0.11355159552424368, + "grad_norm": 0.25297921895980835, + "learning_rate": 4.432656444260257e-06, + "loss": 0.5124, + "step": 1370 + }, + { + "epoch": 0.11363447990053875, + "grad_norm": 0.26524803042411804, + "learning_rate": 4.432242022378782e-06, + "loss": 0.5741, + "step": 1371 + }, + { + "epoch": 0.11371736427683382, + "grad_norm": 0.2801331877708435, + "learning_rate": 4.431827600497306e-06, + "loss": 0.525, + "step": 1372 + }, + { + "epoch": 0.11380024865312889, + "grad_norm": 0.2457887828350067, + "learning_rate": 4.431413178615832e-06, + "loss": 0.498, + "step": 1373 + }, + { + "epoch": 0.11388313302942395, + "grad_norm": 0.2677062451839447, + "learning_rate": 4.430998756734356e-06, + "loss": 0.5826, + "step": 1374 + }, + { + "epoch": 0.11396601740571902, + "grad_norm": 0.260265052318573, + "learning_rate": 4.4305843348528806e-06, + "loss": 0.5432, + "step": 1375 + }, + { + "epoch": 0.11404890178201409, + "grad_norm": 0.28654682636260986, + "learning_rate": 4.430169912971406e-06, + "loss": 0.5589, + "step": 1376 + }, + { + "epoch": 0.11413178615830916, + "grad_norm": 0.3169498145580292, + "learning_rate": 4.42975549108993e-06, + "loss": 0.5355, + "step": 1377 + }, + { + "epoch": 0.11421467053460423, + "grad_norm": 0.262993186712265, + "learning_rate": 4.429341069208455e-06, + "loss": 0.5526, + "step": 1378 + }, + { + "epoch": 0.1142975549108993, + "grad_norm": 0.29232361912727356, + "learning_rate": 4.428926647326979e-06, + "loss": 0.5734, + "step": 1379 + }, + { + "epoch": 0.11438043928719437, + "grad_norm": 0.25822722911834717, + "learning_rate": 4.4285122254455036e-06, + "loss": 0.5468, + "step": 1380 + }, + { + "epoch": 0.11446332366348944, + "grad_norm": 0.2773914933204651, + "learning_rate": 4.428097803564029e-06, + "loss": 0.5822, + "step": 1381 + }, + { + "epoch": 0.1145462080397845, + "grad_norm": 0.2655429244041443, + "learning_rate": 4.427683381682553e-06, + "loss": 0.5463, + "step": 1382 + }, + { + "epoch": 0.11462909241607958, + "grad_norm": 0.2784202992916107, + "learning_rate": 4.427268959801078e-06, + "loss": 0.5635, + "step": 1383 + }, + { + "epoch": 0.11471197679237463, + "grad_norm": 0.35401520133018494, + "learning_rate": 4.426854537919603e-06, + "loss": 0.5704, + "step": 1384 + }, + { + "epoch": 0.1147948611686697, + "grad_norm": 0.2735126316547394, + "learning_rate": 4.426440116038127e-06, + "loss": 0.5325, + "step": 1385 + }, + { + "epoch": 0.11487774554496477, + "grad_norm": 0.2551861107349396, + "learning_rate": 4.426025694156652e-06, + "loss": 0.5075, + "step": 1386 + }, + { + "epoch": 0.11496062992125984, + "grad_norm": 0.2608520984649658, + "learning_rate": 4.425611272275176e-06, + "loss": 0.5635, + "step": 1387 + }, + { + "epoch": 0.11504351429755491, + "grad_norm": 0.32427164912223816, + "learning_rate": 4.425196850393701e-06, + "loss": 0.6306, + "step": 1388 + }, + { + "epoch": 0.11512639867384998, + "grad_norm": 0.2762415111064911, + "learning_rate": 4.424782428512226e-06, + "loss": 0.5617, + "step": 1389 + }, + { + "epoch": 0.11520928305014505, + "grad_norm": 0.29063111543655396, + "learning_rate": 4.42436800663075e-06, + "loss": 0.5872, + "step": 1390 + }, + { + "epoch": 0.11529216742644012, + "grad_norm": 0.2895011007785797, + "learning_rate": 4.423953584749276e-06, + "loss": 0.5493, + "step": 1391 + }, + { + "epoch": 0.11537505180273519, + "grad_norm": 1.4614259004592896, + "learning_rate": 4.4235391628678e-06, + "loss": 0.5646, + "step": 1392 + }, + { + "epoch": 0.11545793617903025, + "grad_norm": 0.2801559269428253, + "learning_rate": 4.4231247409863246e-06, + "loss": 0.56, + "step": 1393 + }, + { + "epoch": 0.11554082055532532, + "grad_norm": 0.314558207988739, + "learning_rate": 4.422710319104849e-06, + "loss": 0.5717, + "step": 1394 + }, + { + "epoch": 0.11562370493162039, + "grad_norm": 0.3097037374973297, + "learning_rate": 4.4222958972233734e-06, + "loss": 0.5964, + "step": 1395 + }, + { + "epoch": 0.11570658930791546, + "grad_norm": 0.29168611764907837, + "learning_rate": 4.421881475341898e-06, + "loss": 0.558, + "step": 1396 + }, + { + "epoch": 0.11578947368421053, + "grad_norm": 0.2703162729740143, + "learning_rate": 4.421467053460423e-06, + "loss": 0.5151, + "step": 1397 + }, + { + "epoch": 0.1158723580605056, + "grad_norm": 0.2855607569217682, + "learning_rate": 4.4210526315789476e-06, + "loss": 0.5552, + "step": 1398 + }, + { + "epoch": 0.11595524243680067, + "grad_norm": 0.2669489085674286, + "learning_rate": 4.420638209697473e-06, + "loss": 0.5262, + "step": 1399 + }, + { + "epoch": 0.11603812681309573, + "grad_norm": 0.27066612243652344, + "learning_rate": 4.420223787815997e-06, + "loss": 0.5937, + "step": 1400 + }, + { + "epoch": 0.1161210111893908, + "grad_norm": 0.29849839210510254, + "learning_rate": 4.419809365934522e-06, + "loss": 0.5803, + "step": 1401 + }, + { + "epoch": 0.11620389556568587, + "grad_norm": 0.2668629288673401, + "learning_rate": 4.419394944053046e-06, + "loss": 0.5303, + "step": 1402 + }, + { + "epoch": 0.11628677994198093, + "grad_norm": 0.2858399450778961, + "learning_rate": 4.4189805221715706e-06, + "loss": 0.5192, + "step": 1403 + }, + { + "epoch": 0.116369664318276, + "grad_norm": 0.28094857931137085, + "learning_rate": 4.418566100290096e-06, + "loss": 0.6412, + "step": 1404 + }, + { + "epoch": 0.11645254869457107, + "grad_norm": 0.2922925353050232, + "learning_rate": 4.41815167840862e-06, + "loss": 0.6057, + "step": 1405 + }, + { + "epoch": 0.11653543307086614, + "grad_norm": 0.2631579339504242, + "learning_rate": 4.417737256527145e-06, + "loss": 0.5554, + "step": 1406 + }, + { + "epoch": 0.11661831744716121, + "grad_norm": 0.2671273648738861, + "learning_rate": 4.41732283464567e-06, + "loss": 0.5575, + "step": 1407 + }, + { + "epoch": 0.11670120182345628, + "grad_norm": 0.28707289695739746, + "learning_rate": 4.4169084127641944e-06, + "loss": 0.5667, + "step": 1408 + }, + { + "epoch": 0.11678408619975135, + "grad_norm": 0.273137629032135, + "learning_rate": 4.416493990882719e-06, + "loss": 0.5718, + "step": 1409 + }, + { + "epoch": 0.11686697057604642, + "grad_norm": 0.279514878988266, + "learning_rate": 4.416079569001243e-06, + "loss": 0.5986, + "step": 1410 + }, + { + "epoch": 0.11694985495234149, + "grad_norm": 0.2856149971485138, + "learning_rate": 4.415665147119768e-06, + "loss": 0.5811, + "step": 1411 + }, + { + "epoch": 0.11703273932863655, + "grad_norm": 0.30906715989112854, + "learning_rate": 4.415250725238293e-06, + "loss": 0.5362, + "step": 1412 + }, + { + "epoch": 0.11711562370493162, + "grad_norm": 0.2850746214389801, + "learning_rate": 4.4148363033568174e-06, + "loss": 0.564, + "step": 1413 + }, + { + "epoch": 0.11719850808122668, + "grad_norm": 0.2715719938278198, + "learning_rate": 4.414421881475342e-06, + "loss": 0.502, + "step": 1414 + }, + { + "epoch": 0.11728139245752175, + "grad_norm": 0.3070889115333557, + "learning_rate": 4.414007459593867e-06, + "loss": 0.5327, + "step": 1415 + }, + { + "epoch": 0.11736427683381682, + "grad_norm": 0.28039684891700745, + "learning_rate": 4.413593037712392e-06, + "loss": 0.5815, + "step": 1416 + }, + { + "epoch": 0.1174471612101119, + "grad_norm": 0.27033740282058716, + "learning_rate": 4.413178615830917e-06, + "loss": 0.5709, + "step": 1417 + }, + { + "epoch": 0.11753004558640696, + "grad_norm": 0.29446959495544434, + "learning_rate": 4.412764193949441e-06, + "loss": 0.5677, + "step": 1418 + }, + { + "epoch": 0.11761292996270203, + "grad_norm": 0.28068095445632935, + "learning_rate": 4.412349772067966e-06, + "loss": 0.5766, + "step": 1419 + }, + { + "epoch": 0.1176958143389971, + "grad_norm": 0.2819221317768097, + "learning_rate": 4.41193535018649e-06, + "loss": 0.5862, + "step": 1420 + }, + { + "epoch": 0.11777869871529217, + "grad_norm": 0.32502666115760803, + "learning_rate": 4.411520928305015e-06, + "loss": 0.5705, + "step": 1421 + }, + { + "epoch": 0.11786158309158723, + "grad_norm": 0.294766366481781, + "learning_rate": 4.41110650642354e-06, + "loss": 0.6078, + "step": 1422 + }, + { + "epoch": 0.1179444674678823, + "grad_norm": 0.2835538685321808, + "learning_rate": 4.410692084542064e-06, + "loss": 0.5142, + "step": 1423 + }, + { + "epoch": 0.11802735184417737, + "grad_norm": 0.29480814933776855, + "learning_rate": 4.410277662660589e-06, + "loss": 0.5579, + "step": 1424 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 0.27884840965270996, + "learning_rate": 4.409863240779114e-06, + "loss": 0.5912, + "step": 1425 + }, + { + "epoch": 0.11819312059676751, + "grad_norm": 0.2864609360694885, + "learning_rate": 4.4094488188976384e-06, + "loss": 0.5908, + "step": 1426 + }, + { + "epoch": 0.11827600497306258, + "grad_norm": 0.2717483937740326, + "learning_rate": 4.409034397016163e-06, + "loss": 0.5784, + "step": 1427 + }, + { + "epoch": 0.11835888934935765, + "grad_norm": 0.2927391827106476, + "learning_rate": 4.408619975134687e-06, + "loss": 0.5684, + "step": 1428 + }, + { + "epoch": 0.11844177372565272, + "grad_norm": 0.2890487313270569, + "learning_rate": 4.408205553253212e-06, + "loss": 0.5901, + "step": 1429 + }, + { + "epoch": 0.11852465810194779, + "grad_norm": 0.27580320835113525, + "learning_rate": 4.407791131371737e-06, + "loss": 0.5552, + "step": 1430 + }, + { + "epoch": 0.11860754247824284, + "grad_norm": 0.2954568862915039, + "learning_rate": 4.4073767094902614e-06, + "loss": 0.5938, + "step": 1431 + }, + { + "epoch": 0.11869042685453791, + "grad_norm": 0.28436821699142456, + "learning_rate": 4.406962287608786e-06, + "loss": 0.624, + "step": 1432 + }, + { + "epoch": 0.11877331123083298, + "grad_norm": 0.2672399580478668, + "learning_rate": 4.406547865727311e-06, + "loss": 0.5042, + "step": 1433 + }, + { + "epoch": 0.11885619560712805, + "grad_norm": 0.26380881667137146, + "learning_rate": 4.406133443845836e-06, + "loss": 0.5002, + "step": 1434 + }, + { + "epoch": 0.11893907998342312, + "grad_norm": 0.29651349782943726, + "learning_rate": 4.40571902196436e-06, + "loss": 0.5825, + "step": 1435 + }, + { + "epoch": 0.1190219643597182, + "grad_norm": 0.2592439651489258, + "learning_rate": 4.4053046000828844e-06, + "loss": 0.5107, + "step": 1436 + }, + { + "epoch": 0.11910484873601326, + "grad_norm": 0.2908269762992859, + "learning_rate": 4.404890178201409e-06, + "loss": 0.5531, + "step": 1437 + }, + { + "epoch": 0.11918773311230833, + "grad_norm": 0.2703227400779724, + "learning_rate": 4.404475756319934e-06, + "loss": 0.5765, + "step": 1438 + }, + { + "epoch": 0.1192706174886034, + "grad_norm": 0.2844199240207672, + "learning_rate": 4.404061334438459e-06, + "loss": 0.5788, + "step": 1439 + }, + { + "epoch": 0.11935350186489847, + "grad_norm": 0.27028200030326843, + "learning_rate": 4.403646912556984e-06, + "loss": 0.5052, + "step": 1440 + }, + { + "epoch": 0.11943638624119353, + "grad_norm": 0.28697702288627625, + "learning_rate": 4.403232490675508e-06, + "loss": 0.5714, + "step": 1441 + }, + { + "epoch": 0.1195192706174886, + "grad_norm": 0.2869735360145569, + "learning_rate": 4.402818068794033e-06, + "loss": 0.5681, + "step": 1442 + }, + { + "epoch": 0.11960215499378367, + "grad_norm": 0.26548799872398376, + "learning_rate": 4.402403646912557e-06, + "loss": 0.5377, + "step": 1443 + }, + { + "epoch": 0.11968503937007874, + "grad_norm": 0.2875244617462158, + "learning_rate": 4.401989225031082e-06, + "loss": 0.5362, + "step": 1444 + }, + { + "epoch": 0.11976792374637381, + "grad_norm": 0.29941526055336, + "learning_rate": 4.401574803149606e-06, + "loss": 0.6011, + "step": 1445 + }, + { + "epoch": 0.11985080812266888, + "grad_norm": 0.29243242740631104, + "learning_rate": 4.401160381268131e-06, + "loss": 0.5363, + "step": 1446 + }, + { + "epoch": 0.11993369249896395, + "grad_norm": 0.23521648347377777, + "learning_rate": 4.400745959386656e-06, + "loss": 0.5075, + "step": 1447 + }, + { + "epoch": 0.12001657687525902, + "grad_norm": 0.2875824570655823, + "learning_rate": 4.400331537505181e-06, + "loss": 0.5667, + "step": 1448 + }, + { + "epoch": 0.12009946125155409, + "grad_norm": 0.24758431315422058, + "learning_rate": 4.3999171156237054e-06, + "loss": 0.5266, + "step": 1449 + }, + { + "epoch": 0.12018234562784914, + "grad_norm": 0.36355748772621155, + "learning_rate": 4.39950269374223e-06, + "loss": 0.5673, + "step": 1450 + }, + { + "epoch": 0.12026523000414421, + "grad_norm": 0.30828428268432617, + "learning_rate": 4.399088271860754e-06, + "loss": 0.6149, + "step": 1451 + }, + { + "epoch": 0.12034811438043928, + "grad_norm": 0.2873872220516205, + "learning_rate": 4.398673849979279e-06, + "loss": 0.573, + "step": 1452 + }, + { + "epoch": 0.12043099875673435, + "grad_norm": 0.2707212269306183, + "learning_rate": 4.398259428097804e-06, + "loss": 0.611, + "step": 1453 + }, + { + "epoch": 0.12051388313302942, + "grad_norm": 0.28814101219177246, + "learning_rate": 4.3978450062163285e-06, + "loss": 0.5863, + "step": 1454 + }, + { + "epoch": 0.12059676750932449, + "grad_norm": 0.27711886167526245, + "learning_rate": 4.397430584334853e-06, + "loss": 0.5944, + "step": 1455 + }, + { + "epoch": 0.12067965188561956, + "grad_norm": 0.2635723352432251, + "learning_rate": 4.397016162453378e-06, + "loss": 0.5512, + "step": 1456 + }, + { + "epoch": 0.12076253626191463, + "grad_norm": 0.2723836302757263, + "learning_rate": 4.396601740571903e-06, + "loss": 0.5152, + "step": 1457 + }, + { + "epoch": 0.1208454206382097, + "grad_norm": 0.2905292212963104, + "learning_rate": 4.396187318690428e-06, + "loss": 0.5666, + "step": 1458 + }, + { + "epoch": 0.12092830501450477, + "grad_norm": 0.2545771300792694, + "learning_rate": 4.395772896808952e-06, + "loss": 0.5565, + "step": 1459 + }, + { + "epoch": 0.12101118939079983, + "grad_norm": 0.28810930252075195, + "learning_rate": 4.395358474927477e-06, + "loss": 0.5132, + "step": 1460 + }, + { + "epoch": 0.1210940737670949, + "grad_norm": 0.26547595858573914, + "learning_rate": 4.394944053046001e-06, + "loss": 0.5556, + "step": 1461 + }, + { + "epoch": 0.12117695814338997, + "grad_norm": 0.25530558824539185, + "learning_rate": 4.394529631164526e-06, + "loss": 0.5416, + "step": 1462 + }, + { + "epoch": 0.12125984251968504, + "grad_norm": 0.32886040210723877, + "learning_rate": 4.39411520928305e-06, + "loss": 0.6209, + "step": 1463 + }, + { + "epoch": 0.12134272689598011, + "grad_norm": 0.2754971385002136, + "learning_rate": 4.393700787401575e-06, + "loss": 0.5479, + "step": 1464 + }, + { + "epoch": 0.12142561127227518, + "grad_norm": 0.2890419065952301, + "learning_rate": 4.3932863655201e-06, + "loss": 0.5392, + "step": 1465 + }, + { + "epoch": 0.12150849564857025, + "grad_norm": 0.27459707856178284, + "learning_rate": 4.392871943638625e-06, + "loss": 0.5271, + "step": 1466 + }, + { + "epoch": 0.12159138002486532, + "grad_norm": 0.2657272219657898, + "learning_rate": 4.3924575217571495e-06, + "loss": 0.5637, + "step": 1467 + }, + { + "epoch": 0.12167426440116039, + "grad_norm": 0.2877826988697052, + "learning_rate": 4.392043099875674e-06, + "loss": 0.574, + "step": 1468 + }, + { + "epoch": 0.12175714877745544, + "grad_norm": 0.2725803554058075, + "learning_rate": 4.391628677994198e-06, + "loss": 0.5605, + "step": 1469 + }, + { + "epoch": 0.12184003315375051, + "grad_norm": 0.27835243940353394, + "learning_rate": 4.391214256112723e-06, + "loss": 0.5587, + "step": 1470 + }, + { + "epoch": 0.12192291753004558, + "grad_norm": 0.28512051701545715, + "learning_rate": 4.390799834231248e-06, + "loss": 0.5606, + "step": 1471 + }, + { + "epoch": 0.12200580190634065, + "grad_norm": 0.2751755118370056, + "learning_rate": 4.3903854123497725e-06, + "loss": 0.5226, + "step": 1472 + }, + { + "epoch": 0.12208868628263572, + "grad_norm": 0.2697247564792633, + "learning_rate": 4.389970990468297e-06, + "loss": 0.5514, + "step": 1473 + }, + { + "epoch": 0.12217157065893079, + "grad_norm": 0.2735845446586609, + "learning_rate": 4.389556568586822e-06, + "loss": 0.5579, + "step": 1474 + }, + { + "epoch": 0.12225445503522586, + "grad_norm": 0.2639630436897278, + "learning_rate": 4.389142146705347e-06, + "loss": 0.5562, + "step": 1475 + }, + { + "epoch": 0.12233733941152093, + "grad_norm": 0.26918190717697144, + "learning_rate": 4.388727724823871e-06, + "loss": 0.5529, + "step": 1476 + }, + { + "epoch": 0.122420223787816, + "grad_norm": 0.2705904245376587, + "learning_rate": 4.3883133029423955e-06, + "loss": 0.5562, + "step": 1477 + }, + { + "epoch": 0.12250310816411107, + "grad_norm": 0.283125102519989, + "learning_rate": 4.38789888106092e-06, + "loss": 0.5798, + "step": 1478 + }, + { + "epoch": 0.12258599254040613, + "grad_norm": 0.27395492792129517, + "learning_rate": 4.387484459179445e-06, + "loss": 0.5526, + "step": 1479 + }, + { + "epoch": 0.1226688769167012, + "grad_norm": 0.26209375262260437, + "learning_rate": 4.38707003729797e-06, + "loss": 0.5464, + "step": 1480 + }, + { + "epoch": 0.12275176129299627, + "grad_norm": 0.26583752036094666, + "learning_rate": 4.386655615416494e-06, + "loss": 0.5558, + "step": 1481 + }, + { + "epoch": 0.12283464566929134, + "grad_norm": 0.3277220129966736, + "learning_rate": 4.386241193535019e-06, + "loss": 0.5716, + "step": 1482 + }, + { + "epoch": 0.1229175300455864, + "grad_norm": 0.2688691020011902, + "learning_rate": 4.385826771653544e-06, + "loss": 0.5706, + "step": 1483 + }, + { + "epoch": 0.12300041442188148, + "grad_norm": 0.2668711245059967, + "learning_rate": 4.385412349772068e-06, + "loss": 0.5544, + "step": 1484 + }, + { + "epoch": 0.12308329879817655, + "grad_norm": 0.2574160695075989, + "learning_rate": 4.384997927890593e-06, + "loss": 0.5267, + "step": 1485 + }, + { + "epoch": 0.12316618317447162, + "grad_norm": 0.2797733247280121, + "learning_rate": 4.384583506009117e-06, + "loss": 0.5594, + "step": 1486 + }, + { + "epoch": 0.12324906755076669, + "grad_norm": 0.2824660837650299, + "learning_rate": 4.384169084127642e-06, + "loss": 0.5648, + "step": 1487 + }, + { + "epoch": 0.12333195192706174, + "grad_norm": 0.38257431983947754, + "learning_rate": 4.383754662246167e-06, + "loss": 0.5493, + "step": 1488 + }, + { + "epoch": 0.12341483630335681, + "grad_norm": 0.2613079249858856, + "learning_rate": 4.383340240364692e-06, + "loss": 0.5724, + "step": 1489 + }, + { + "epoch": 0.12349772067965188, + "grad_norm": 0.2982143461704254, + "learning_rate": 4.3829258184832165e-06, + "loss": 0.5647, + "step": 1490 + }, + { + "epoch": 0.12358060505594695, + "grad_norm": 0.2917243242263794, + "learning_rate": 4.382511396601741e-06, + "loss": 0.5774, + "step": 1491 + }, + { + "epoch": 0.12366348943224202, + "grad_norm": 0.2785344421863556, + "learning_rate": 4.382096974720265e-06, + "loss": 0.5496, + "step": 1492 + }, + { + "epoch": 0.12374637380853709, + "grad_norm": 0.2845265865325928, + "learning_rate": 4.38168255283879e-06, + "loss": 0.5621, + "step": 1493 + }, + { + "epoch": 0.12382925818483216, + "grad_norm": 0.27491500973701477, + "learning_rate": 4.381268130957315e-06, + "loss": 0.5333, + "step": 1494 + }, + { + "epoch": 0.12391214256112723, + "grad_norm": 0.2732330560684204, + "learning_rate": 4.3808537090758395e-06, + "loss": 0.5861, + "step": 1495 + }, + { + "epoch": 0.1239950269374223, + "grad_norm": 0.2471430003643036, + "learning_rate": 4.380439287194364e-06, + "loss": 0.5257, + "step": 1496 + }, + { + "epoch": 0.12407791131371737, + "grad_norm": 0.2740758955478668, + "learning_rate": 4.380024865312889e-06, + "loss": 0.5817, + "step": 1497 + }, + { + "epoch": 0.12416079569001243, + "grad_norm": 0.2849424183368683, + "learning_rate": 4.379610443431414e-06, + "loss": 0.5538, + "step": 1498 + }, + { + "epoch": 0.1242436800663075, + "grad_norm": 0.27986350655555725, + "learning_rate": 4.379196021549938e-06, + "loss": 0.5665, + "step": 1499 + }, + { + "epoch": 0.12432656444260257, + "grad_norm": 0.28577232360839844, + "learning_rate": 4.378781599668463e-06, + "loss": 0.5874, + "step": 1500 + }, + { + "epoch": 0.12440944881889764, + "grad_norm": 0.28587794303894043, + "learning_rate": 4.378367177786988e-06, + "loss": 0.5797, + "step": 1501 + }, + { + "epoch": 0.1244923331951927, + "grad_norm": 0.289269357919693, + "learning_rate": 4.377952755905512e-06, + "loss": 0.5795, + "step": 1502 + }, + { + "epoch": 0.12457521757148778, + "grad_norm": 0.2722938656806946, + "learning_rate": 4.377538334024037e-06, + "loss": 0.5907, + "step": 1503 + }, + { + "epoch": 0.12465810194778285, + "grad_norm": 0.27100393176078796, + "learning_rate": 4.377123912142561e-06, + "loss": 0.576, + "step": 1504 + }, + { + "epoch": 0.12474098632407792, + "grad_norm": 0.2676182985305786, + "learning_rate": 4.376709490261086e-06, + "loss": 0.5638, + "step": 1505 + }, + { + "epoch": 0.12482387070037299, + "grad_norm": 0.2842869758605957, + "learning_rate": 4.376295068379611e-06, + "loss": 0.5867, + "step": 1506 + }, + { + "epoch": 0.12490675507666804, + "grad_norm": 0.2671494483947754, + "learning_rate": 4.375880646498136e-06, + "loss": 0.5462, + "step": 1507 + }, + { + "epoch": 0.12498963945296311, + "grad_norm": 0.2662389576435089, + "learning_rate": 4.3754662246166605e-06, + "loss": 0.5091, + "step": 1508 + }, + { + "epoch": 0.12507252382925818, + "grad_norm": 0.2695591151714325, + "learning_rate": 4.375051802735185e-06, + "loss": 0.5722, + "step": 1509 + }, + { + "epoch": 0.12515540820555326, + "grad_norm": 0.26575204730033875, + "learning_rate": 4.374637380853709e-06, + "loss": 0.5425, + "step": 1510 + }, + { + "epoch": 0.12523829258184832, + "grad_norm": 0.2701515853404999, + "learning_rate": 4.374222958972234e-06, + "loss": 0.5758, + "step": 1511 + }, + { + "epoch": 0.12532117695814338, + "grad_norm": 0.27190062403678894, + "learning_rate": 4.373808537090758e-06, + "loss": 0.6032, + "step": 1512 + }, + { + "epoch": 0.12540406133443846, + "grad_norm": 0.2700665593147278, + "learning_rate": 4.3733941152092835e-06, + "loss": 0.5441, + "step": 1513 + }, + { + "epoch": 0.12548694571073352, + "grad_norm": 0.2587072551250458, + "learning_rate": 4.372979693327808e-06, + "loss": 0.5169, + "step": 1514 + }, + { + "epoch": 0.1255698300870286, + "grad_norm": 0.28307124972343445, + "learning_rate": 4.372565271446333e-06, + "loss": 0.578, + "step": 1515 + }, + { + "epoch": 0.12565271446332366, + "grad_norm": 0.27594107389450073, + "learning_rate": 4.372150849564858e-06, + "loss": 0.5618, + "step": 1516 + }, + { + "epoch": 0.12573559883961874, + "grad_norm": 0.2861180603504181, + "learning_rate": 4.371736427683382e-06, + "loss": 0.5498, + "step": 1517 + }, + { + "epoch": 0.1258184832159138, + "grad_norm": 0.2721790075302124, + "learning_rate": 4.3713220058019065e-06, + "loss": 0.5706, + "step": 1518 + }, + { + "epoch": 0.12590136759220888, + "grad_norm": 0.2664988040924072, + "learning_rate": 4.370907583920431e-06, + "loss": 0.5609, + "step": 1519 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 0.2600027322769165, + "learning_rate": 4.370493162038956e-06, + "loss": 0.5302, + "step": 1520 + }, + { + "epoch": 0.12606713634479902, + "grad_norm": 0.2622773349285126, + "learning_rate": 4.370078740157481e-06, + "loss": 0.5553, + "step": 1521 + }, + { + "epoch": 0.12615002072109407, + "grad_norm": 0.26330605149269104, + "learning_rate": 4.369664318276005e-06, + "loss": 0.5312, + "step": 1522 + }, + { + "epoch": 0.12623290509738913, + "grad_norm": 0.28582873940467834, + "learning_rate": 4.36924989639453e-06, + "loss": 0.5456, + "step": 1523 + }, + { + "epoch": 0.12631578947368421, + "grad_norm": 0.2750333249568939, + "learning_rate": 4.368835474513055e-06, + "loss": 0.5753, + "step": 1524 + }, + { + "epoch": 0.12639867384997927, + "grad_norm": 0.2621963620185852, + "learning_rate": 4.368421052631579e-06, + "loss": 0.5527, + "step": 1525 + }, + { + "epoch": 0.12648155822627435, + "grad_norm": 0.3039098381996155, + "learning_rate": 4.368006630750104e-06, + "loss": 0.5768, + "step": 1526 + }, + { + "epoch": 0.1265644426025694, + "grad_norm": 0.24838294088840485, + "learning_rate": 4.367592208868628e-06, + "loss": 0.5179, + "step": 1527 + }, + { + "epoch": 0.1266473269788645, + "grad_norm": 0.24725095927715302, + "learning_rate": 4.367177786987153e-06, + "loss": 0.5383, + "step": 1528 + }, + { + "epoch": 0.12673021135515955, + "grad_norm": 0.28100404143333435, + "learning_rate": 4.366763365105678e-06, + "loss": 0.5652, + "step": 1529 + }, + { + "epoch": 0.12681309573145463, + "grad_norm": 0.2919383645057678, + "learning_rate": 4.366348943224202e-06, + "loss": 0.5675, + "step": 1530 + }, + { + "epoch": 0.1268959801077497, + "grad_norm": 0.24799682199954987, + "learning_rate": 4.3659345213427275e-06, + "loss": 0.5228, + "step": 1531 + }, + { + "epoch": 0.12697886448404475, + "grad_norm": 0.27015992999076843, + "learning_rate": 4.365520099461252e-06, + "loss": 0.5367, + "step": 1532 + }, + { + "epoch": 0.12706174886033983, + "grad_norm": 0.3140951991081238, + "learning_rate": 4.365105677579776e-06, + "loss": 0.5853, + "step": 1533 + }, + { + "epoch": 0.12714463323663489, + "grad_norm": 0.2638419568538666, + "learning_rate": 4.364691255698302e-06, + "loss": 0.5888, + "step": 1534 + }, + { + "epoch": 0.12722751761292997, + "grad_norm": 0.2520715594291687, + "learning_rate": 4.364276833816826e-06, + "loss": 0.5673, + "step": 1535 + }, + { + "epoch": 0.12731040198922502, + "grad_norm": 0.25373461842536926, + "learning_rate": 4.3638624119353505e-06, + "loss": 0.5243, + "step": 1536 + }, + { + "epoch": 0.1273932863655201, + "grad_norm": 0.25575241446495056, + "learning_rate": 4.363447990053875e-06, + "loss": 0.5686, + "step": 1537 + }, + { + "epoch": 0.12747617074181516, + "grad_norm": 0.22937430441379547, + "learning_rate": 4.3630335681724e-06, + "loss": 0.4778, + "step": 1538 + }, + { + "epoch": 0.12755905511811025, + "grad_norm": 0.2695535123348236, + "learning_rate": 4.362619146290925e-06, + "loss": 0.523, + "step": 1539 + }, + { + "epoch": 0.1276419394944053, + "grad_norm": 0.27817490696907043, + "learning_rate": 4.362204724409449e-06, + "loss": 0.5681, + "step": 1540 + }, + { + "epoch": 0.12772482387070036, + "grad_norm": 0.2824763059616089, + "learning_rate": 4.361790302527974e-06, + "loss": 0.5405, + "step": 1541 + }, + { + "epoch": 0.12780770824699544, + "grad_norm": 0.27647915482521057, + "learning_rate": 4.361375880646499e-06, + "loss": 0.5526, + "step": 1542 + }, + { + "epoch": 0.1278905926232905, + "grad_norm": 0.28012678027153015, + "learning_rate": 4.360961458765023e-06, + "loss": 0.5289, + "step": 1543 + }, + { + "epoch": 0.12797347699958558, + "grad_norm": 0.2873612940311432, + "learning_rate": 4.360547036883548e-06, + "loss": 0.584, + "step": 1544 + }, + { + "epoch": 0.12805636137588064, + "grad_norm": 0.28686320781707764, + "learning_rate": 4.360132615002072e-06, + "loss": 0.6242, + "step": 1545 + }, + { + "epoch": 0.12813924575217572, + "grad_norm": 0.2674858868122101, + "learning_rate": 4.359718193120597e-06, + "loss": 0.5089, + "step": 1546 + }, + { + "epoch": 0.12822213012847078, + "grad_norm": 0.25843581557273865, + "learning_rate": 4.359303771239122e-06, + "loss": 0.4958, + "step": 1547 + }, + { + "epoch": 0.12830501450476586, + "grad_norm": 0.28449106216430664, + "learning_rate": 4.358889349357646e-06, + "loss": 0.5508, + "step": 1548 + }, + { + "epoch": 0.12838789888106092, + "grad_norm": 0.26092904806137085, + "learning_rate": 4.3584749274761715e-06, + "loss": 0.5591, + "step": 1549 + }, + { + "epoch": 0.12847078325735597, + "grad_norm": 0.25312569737434387, + "learning_rate": 4.358060505594696e-06, + "loss": 0.5567, + "step": 1550 + }, + { + "epoch": 0.12855366763365106, + "grad_norm": 0.2939331829547882, + "learning_rate": 4.35764608371322e-06, + "loss": 0.6143, + "step": 1551 + }, + { + "epoch": 0.12863655200994611, + "grad_norm": 0.2866968810558319, + "learning_rate": 4.357231661831745e-06, + "loss": 0.5982, + "step": 1552 + }, + { + "epoch": 0.1287194363862412, + "grad_norm": 0.260982871055603, + "learning_rate": 4.356817239950269e-06, + "loss": 0.548, + "step": 1553 + }, + { + "epoch": 0.12880232076253625, + "grad_norm": 0.2934725284576416, + "learning_rate": 4.3564028180687945e-06, + "loss": 0.5983, + "step": 1554 + }, + { + "epoch": 0.12888520513883134, + "grad_norm": 0.28414833545684814, + "learning_rate": 4.355988396187319e-06, + "loss": 0.5905, + "step": 1555 + }, + { + "epoch": 0.1289680895151264, + "grad_norm": 0.27412810921669006, + "learning_rate": 4.355573974305844e-06, + "loss": 0.5555, + "step": 1556 + }, + { + "epoch": 0.12905097389142148, + "grad_norm": 0.28370746970176697, + "learning_rate": 4.355159552424369e-06, + "loss": 0.5769, + "step": 1557 + }, + { + "epoch": 0.12913385826771653, + "grad_norm": 0.29487597942352295, + "learning_rate": 4.354745130542893e-06, + "loss": 0.5886, + "step": 1558 + }, + { + "epoch": 0.12921674264401162, + "grad_norm": 0.2586494982242584, + "learning_rate": 4.3543307086614175e-06, + "loss": 0.5537, + "step": 1559 + }, + { + "epoch": 0.12929962702030667, + "grad_norm": 0.29433417320251465, + "learning_rate": 4.353916286779942e-06, + "loss": 0.5923, + "step": 1560 + }, + { + "epoch": 0.12938251139660173, + "grad_norm": 0.2759227454662323, + "learning_rate": 4.353501864898467e-06, + "loss": 0.5627, + "step": 1561 + }, + { + "epoch": 0.1294653957728968, + "grad_norm": 0.26656532287597656, + "learning_rate": 4.353087443016992e-06, + "loss": 0.5452, + "step": 1562 + }, + { + "epoch": 0.12954828014919187, + "grad_norm": 0.2669527530670166, + "learning_rate": 4.352673021135516e-06, + "loss": 0.548, + "step": 1563 + }, + { + "epoch": 0.12963116452548695, + "grad_norm": 0.2814754843711853, + "learning_rate": 4.352258599254041e-06, + "loss": 0.5452, + "step": 1564 + }, + { + "epoch": 0.129714048901782, + "grad_norm": 0.25762805342674255, + "learning_rate": 4.351844177372566e-06, + "loss": 0.5793, + "step": 1565 + }, + { + "epoch": 0.1297969332780771, + "grad_norm": 0.2941034734249115, + "learning_rate": 4.35142975549109e-06, + "loss": 0.5724, + "step": 1566 + }, + { + "epoch": 0.12987981765437215, + "grad_norm": 0.2885447144508362, + "learning_rate": 4.351015333609615e-06, + "loss": 0.5752, + "step": 1567 + }, + { + "epoch": 0.12996270203066723, + "grad_norm": 0.26475635170936584, + "learning_rate": 4.350600911728139e-06, + "loss": 0.555, + "step": 1568 + }, + { + "epoch": 0.1300455864069623, + "grad_norm": 0.27669185400009155, + "learning_rate": 4.350186489846664e-06, + "loss": 0.5757, + "step": 1569 + }, + { + "epoch": 0.13012847078325734, + "grad_norm": 0.2811535894870758, + "learning_rate": 4.349772067965189e-06, + "loss": 0.5598, + "step": 1570 + }, + { + "epoch": 0.13021135515955243, + "grad_norm": 0.2826133370399475, + "learning_rate": 4.349357646083713e-06, + "loss": 0.5781, + "step": 1571 + }, + { + "epoch": 0.13029423953584748, + "grad_norm": 0.28698840737342834, + "learning_rate": 4.3489432242022385e-06, + "loss": 0.5887, + "step": 1572 + }, + { + "epoch": 0.13037712391214257, + "grad_norm": 0.2935137450695038, + "learning_rate": 4.348528802320763e-06, + "loss": 0.5714, + "step": 1573 + }, + { + "epoch": 0.13046000828843762, + "grad_norm": 0.26280757784843445, + "learning_rate": 4.348114380439287e-06, + "loss": 0.5384, + "step": 1574 + }, + { + "epoch": 0.1305428926647327, + "grad_norm": 0.29048898816108704, + "learning_rate": 4.347699958557813e-06, + "loss": 0.5826, + "step": 1575 + }, + { + "epoch": 0.13062577704102776, + "grad_norm": 0.26889538764953613, + "learning_rate": 4.347285536676337e-06, + "loss": 0.5226, + "step": 1576 + }, + { + "epoch": 0.13070866141732285, + "grad_norm": 0.28510627150535583, + "learning_rate": 4.3468711147948615e-06, + "loss": 0.5635, + "step": 1577 + }, + { + "epoch": 0.1307915457936179, + "grad_norm": 0.25643566250801086, + "learning_rate": 4.346456692913386e-06, + "loss": 0.5761, + "step": 1578 + }, + { + "epoch": 0.13087443016991296, + "grad_norm": 0.2610361576080322, + "learning_rate": 4.34604227103191e-06, + "loss": 0.6264, + "step": 1579 + }, + { + "epoch": 0.13095731454620804, + "grad_norm": 0.2770853340625763, + "learning_rate": 4.345627849150436e-06, + "loss": 0.5608, + "step": 1580 + }, + { + "epoch": 0.1310401989225031, + "grad_norm": 0.2853512763977051, + "learning_rate": 4.34521342726896e-06, + "loss": 0.5484, + "step": 1581 + }, + { + "epoch": 0.13112308329879818, + "grad_norm": 0.3333970010280609, + "learning_rate": 4.344799005387485e-06, + "loss": 0.5817, + "step": 1582 + }, + { + "epoch": 0.13120596767509324, + "grad_norm": 0.2865373194217682, + "learning_rate": 4.34438458350601e-06, + "loss": 0.571, + "step": 1583 + }, + { + "epoch": 0.13128885205138832, + "grad_norm": 0.24962612986564636, + "learning_rate": 4.343970161624534e-06, + "loss": 0.5276, + "step": 1584 + }, + { + "epoch": 0.13137173642768338, + "grad_norm": 0.2818818688392639, + "learning_rate": 4.343555739743059e-06, + "loss": 0.598, + "step": 1585 + }, + { + "epoch": 0.13145462080397846, + "grad_norm": 0.28253793716430664, + "learning_rate": 4.343141317861583e-06, + "loss": 0.606, + "step": 1586 + }, + { + "epoch": 0.13153750518027352, + "grad_norm": 0.2728898823261261, + "learning_rate": 4.342726895980108e-06, + "loss": 0.5593, + "step": 1587 + }, + { + "epoch": 0.13162038955656857, + "grad_norm": 0.2924867868423462, + "learning_rate": 4.342312474098633e-06, + "loss": 0.6413, + "step": 1588 + }, + { + "epoch": 0.13170327393286366, + "grad_norm": 0.30049893260002136, + "learning_rate": 4.341898052217157e-06, + "loss": 0.5643, + "step": 1589 + }, + { + "epoch": 0.1317861583091587, + "grad_norm": 0.30206209421157837, + "learning_rate": 4.3414836303356825e-06, + "loss": 0.5507, + "step": 1590 + }, + { + "epoch": 0.1318690426854538, + "grad_norm": 0.29439446330070496, + "learning_rate": 4.341069208454207e-06, + "loss": 0.5786, + "step": 1591 + }, + { + "epoch": 0.13195192706174885, + "grad_norm": 0.26484358310699463, + "learning_rate": 4.340654786572731e-06, + "loss": 0.5395, + "step": 1592 + }, + { + "epoch": 0.13203481143804394, + "grad_norm": 0.29492148756980896, + "learning_rate": 4.340240364691256e-06, + "loss": 0.6054, + "step": 1593 + }, + { + "epoch": 0.132117695814339, + "grad_norm": 0.2912951409816742, + "learning_rate": 4.33982594280978e-06, + "loss": 0.5912, + "step": 1594 + }, + { + "epoch": 0.13220058019063408, + "grad_norm": 0.27908119559288025, + "learning_rate": 4.3394115209283055e-06, + "loss": 0.5819, + "step": 1595 + }, + { + "epoch": 0.13228346456692913, + "grad_norm": 0.26995328068733215, + "learning_rate": 4.33899709904683e-06, + "loss": 0.5153, + "step": 1596 + }, + { + "epoch": 0.13236634894322422, + "grad_norm": 0.28916171193122864, + "learning_rate": 4.338582677165354e-06, + "loss": 0.556, + "step": 1597 + }, + { + "epoch": 0.13244923331951927, + "grad_norm": 0.28871873021125793, + "learning_rate": 4.33816825528388e-06, + "loss": 0.5532, + "step": 1598 + }, + { + "epoch": 0.13253211769581433, + "grad_norm": 0.28175821900367737, + "learning_rate": 4.337753833402404e-06, + "loss": 0.5574, + "step": 1599 + }, + { + "epoch": 0.1326150020721094, + "grad_norm": 0.2851613461971283, + "learning_rate": 4.3373394115209285e-06, + "loss": 0.5788, + "step": 1600 + }, + { + "epoch": 0.13269788644840447, + "grad_norm": 0.27367937564849854, + "learning_rate": 4.336924989639453e-06, + "loss": 0.5362, + "step": 1601 + }, + { + "epoch": 0.13278077082469955, + "grad_norm": 0.27447110414505005, + "learning_rate": 4.336510567757977e-06, + "loss": 0.5493, + "step": 1602 + }, + { + "epoch": 0.1328636552009946, + "grad_norm": 0.3154110014438629, + "learning_rate": 4.336096145876503e-06, + "loss": 0.6109, + "step": 1603 + }, + { + "epoch": 0.1329465395772897, + "grad_norm": 0.2500707805156708, + "learning_rate": 4.335681723995027e-06, + "loss": 0.4841, + "step": 1604 + }, + { + "epoch": 0.13302942395358475, + "grad_norm": 0.24213553965091705, + "learning_rate": 4.335267302113552e-06, + "loss": 0.5095, + "step": 1605 + }, + { + "epoch": 0.13311230832987983, + "grad_norm": 0.2790689766407013, + "learning_rate": 4.334852880232077e-06, + "loss": 0.5076, + "step": 1606 + }, + { + "epoch": 0.1331951927061749, + "grad_norm": 0.2800616919994354, + "learning_rate": 4.334438458350601e-06, + "loss": 0.5526, + "step": 1607 + }, + { + "epoch": 0.13327807708246994, + "grad_norm": 0.28267550468444824, + "learning_rate": 4.334024036469126e-06, + "loss": 0.5762, + "step": 1608 + }, + { + "epoch": 0.13336096145876503, + "grad_norm": 0.27068421244621277, + "learning_rate": 4.33360961458765e-06, + "loss": 0.597, + "step": 1609 + }, + { + "epoch": 0.13344384583506008, + "grad_norm": 0.25286030769348145, + "learning_rate": 4.333195192706175e-06, + "loss": 0.5141, + "step": 1610 + }, + { + "epoch": 0.13352673021135517, + "grad_norm": 0.29003140330314636, + "learning_rate": 4.3327807708247e-06, + "loss": 0.5454, + "step": 1611 + }, + { + "epoch": 0.13360961458765022, + "grad_norm": 0.2772515118122101, + "learning_rate": 4.332366348943224e-06, + "loss": 0.5611, + "step": 1612 + }, + { + "epoch": 0.1336924989639453, + "grad_norm": 0.2840704023838043, + "learning_rate": 4.3319519270617495e-06, + "loss": 0.5847, + "step": 1613 + }, + { + "epoch": 0.13377538334024036, + "grad_norm": 0.320759654045105, + "learning_rate": 4.331537505180274e-06, + "loss": 0.6001, + "step": 1614 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 0.2999081611633301, + "learning_rate": 4.331123083298798e-06, + "loss": 0.5624, + "step": 1615 + }, + { + "epoch": 0.1339411520928305, + "grad_norm": 0.2634432315826416, + "learning_rate": 4.330708661417324e-06, + "loss": 0.5498, + "step": 1616 + }, + { + "epoch": 0.13402403646912556, + "grad_norm": 0.2796984612941742, + "learning_rate": 4.330294239535848e-06, + "loss": 0.5473, + "step": 1617 + }, + { + "epoch": 0.13410692084542064, + "grad_norm": 0.2730984091758728, + "learning_rate": 4.3298798176543725e-06, + "loss": 0.5979, + "step": 1618 + }, + { + "epoch": 0.1341898052217157, + "grad_norm": 0.27126243710517883, + "learning_rate": 4.329465395772897e-06, + "loss": 0.5085, + "step": 1619 + }, + { + "epoch": 0.13427268959801078, + "grad_norm": 0.28295519948005676, + "learning_rate": 4.329050973891421e-06, + "loss": 0.595, + "step": 1620 + }, + { + "epoch": 0.13435557397430584, + "grad_norm": 0.28713035583496094, + "learning_rate": 4.328636552009947e-06, + "loss": 0.5983, + "step": 1621 + }, + { + "epoch": 0.13443845835060092, + "grad_norm": 0.2748029828071594, + "learning_rate": 4.328222130128471e-06, + "loss": 0.519, + "step": 1622 + }, + { + "epoch": 0.13452134272689598, + "grad_norm": 0.2860753834247589, + "learning_rate": 4.327807708246996e-06, + "loss": 0.5829, + "step": 1623 + }, + { + "epoch": 0.13460422710319106, + "grad_norm": 0.25690898299217224, + "learning_rate": 4.327393286365521e-06, + "loss": 0.5438, + "step": 1624 + }, + { + "epoch": 0.13468711147948612, + "grad_norm": 0.28068679571151733, + "learning_rate": 4.326978864484045e-06, + "loss": 0.5465, + "step": 1625 + }, + { + "epoch": 0.13476999585578117, + "grad_norm": 0.2899668216705322, + "learning_rate": 4.32656444260257e-06, + "loss": 0.61, + "step": 1626 + }, + { + "epoch": 0.13485288023207626, + "grad_norm": 0.2559133470058441, + "learning_rate": 4.326150020721094e-06, + "loss": 0.5547, + "step": 1627 + }, + { + "epoch": 0.1349357646083713, + "grad_norm": 0.25401178002357483, + "learning_rate": 4.325735598839619e-06, + "loss": 0.5248, + "step": 1628 + }, + { + "epoch": 0.1350186489846664, + "grad_norm": 0.27748844027519226, + "learning_rate": 4.325321176958144e-06, + "loss": 0.5518, + "step": 1629 + }, + { + "epoch": 0.13510153336096145, + "grad_norm": 0.2918599545955658, + "learning_rate": 4.324906755076668e-06, + "loss": 0.5751, + "step": 1630 + }, + { + "epoch": 0.13518441773725653, + "grad_norm": 0.27956855297088623, + "learning_rate": 4.3244923331951935e-06, + "loss": 0.5647, + "step": 1631 + }, + { + "epoch": 0.1352673021135516, + "grad_norm": 0.31983640789985657, + "learning_rate": 4.324077911313718e-06, + "loss": 0.5201, + "step": 1632 + }, + { + "epoch": 0.13535018648984667, + "grad_norm": 0.24679087102413177, + "learning_rate": 4.323663489432242e-06, + "loss": 0.5306, + "step": 1633 + }, + { + "epoch": 0.13543307086614173, + "grad_norm": 0.2706209123134613, + "learning_rate": 4.323249067550767e-06, + "loss": 0.4897, + "step": 1634 + }, + { + "epoch": 0.13551595524243681, + "grad_norm": 0.28974393010139465, + "learning_rate": 4.322834645669291e-06, + "loss": 0.548, + "step": 1635 + }, + { + "epoch": 0.13559883961873187, + "grad_norm": 0.29759132862091064, + "learning_rate": 4.3224202237878165e-06, + "loss": 0.5881, + "step": 1636 + }, + { + "epoch": 0.13568172399502693, + "grad_norm": 0.2966158092021942, + "learning_rate": 4.322005801906341e-06, + "loss": 0.5472, + "step": 1637 + }, + { + "epoch": 0.135764608371322, + "grad_norm": 0.24994872510433197, + "learning_rate": 4.321591380024865e-06, + "loss": 0.5182, + "step": 1638 + }, + { + "epoch": 0.13584749274761707, + "grad_norm": 0.25828081369400024, + "learning_rate": 4.321176958143391e-06, + "loss": 0.5099, + "step": 1639 + }, + { + "epoch": 0.13593037712391215, + "grad_norm": 0.28497645258903503, + "learning_rate": 4.320762536261915e-06, + "loss": 0.5318, + "step": 1640 + }, + { + "epoch": 0.1360132615002072, + "grad_norm": 0.3114713132381439, + "learning_rate": 4.3203481143804396e-06, + "loss": 0.6038, + "step": 1641 + }, + { + "epoch": 0.1360961458765023, + "grad_norm": 0.3079416751861572, + "learning_rate": 4.319933692498964e-06, + "loss": 0.5639, + "step": 1642 + }, + { + "epoch": 0.13617903025279735, + "grad_norm": 0.3184870183467865, + "learning_rate": 4.319519270617488e-06, + "loss": 0.5768, + "step": 1643 + }, + { + "epoch": 0.13626191462909243, + "grad_norm": 0.29088908433914185, + "learning_rate": 4.319104848736014e-06, + "loss": 0.5751, + "step": 1644 + }, + { + "epoch": 0.13634479900538748, + "grad_norm": 0.29711803793907166, + "learning_rate": 4.318690426854538e-06, + "loss": 0.6003, + "step": 1645 + }, + { + "epoch": 0.13642768338168254, + "grad_norm": 0.3042736053466797, + "learning_rate": 4.318276004973063e-06, + "loss": 0.5591, + "step": 1646 + }, + { + "epoch": 0.13651056775797762, + "grad_norm": 0.269216924905777, + "learning_rate": 4.317861583091588e-06, + "loss": 0.5564, + "step": 1647 + }, + { + "epoch": 0.13659345213427268, + "grad_norm": 0.25594064593315125, + "learning_rate": 4.317447161210112e-06, + "loss": 0.5139, + "step": 1648 + }, + { + "epoch": 0.13667633651056776, + "grad_norm": 0.27803829312324524, + "learning_rate": 4.317032739328637e-06, + "loss": 0.5513, + "step": 1649 + }, + { + "epoch": 0.13675922088686282, + "grad_norm": 0.24999171495437622, + "learning_rate": 4.316618317447161e-06, + "loss": 0.4981, + "step": 1650 + }, + { + "epoch": 0.1368421052631579, + "grad_norm": 0.28480181097984314, + "learning_rate": 4.3162038955656856e-06, + "loss": 0.5656, + "step": 1651 + }, + { + "epoch": 0.13692498963945296, + "grad_norm": 0.2805059552192688, + "learning_rate": 4.315789473684211e-06, + "loss": 0.5422, + "step": 1652 + }, + { + "epoch": 0.13700787401574804, + "grad_norm": 0.28125107288360596, + "learning_rate": 4.315375051802735e-06, + "loss": 0.5789, + "step": 1653 + }, + { + "epoch": 0.1370907583920431, + "grad_norm": 0.28004276752471924, + "learning_rate": 4.3149606299212606e-06, + "loss": 0.583, + "step": 1654 + }, + { + "epoch": 0.13717364276833816, + "grad_norm": 0.2893523871898651, + "learning_rate": 4.314546208039785e-06, + "loss": 0.5873, + "step": 1655 + }, + { + "epoch": 0.13725652714463324, + "grad_norm": 0.27600812911987305, + "learning_rate": 4.314131786158309e-06, + "loss": 0.5923, + "step": 1656 + }, + { + "epoch": 0.1373394115209283, + "grad_norm": 0.2782376706600189, + "learning_rate": 4.313717364276835e-06, + "loss": 0.5374, + "step": 1657 + }, + { + "epoch": 0.13742229589722338, + "grad_norm": 0.2627936899662018, + "learning_rate": 4.313302942395359e-06, + "loss": 0.6052, + "step": 1658 + }, + { + "epoch": 0.13750518027351843, + "grad_norm": 0.28474947810173035, + "learning_rate": 4.3128885205138836e-06, + "loss": 0.6062, + "step": 1659 + }, + { + "epoch": 0.13758806464981352, + "grad_norm": 0.28253453969955444, + "learning_rate": 4.312474098632408e-06, + "loss": 0.6071, + "step": 1660 + }, + { + "epoch": 0.13767094902610857, + "grad_norm": 0.27660882472991943, + "learning_rate": 4.3120596767509324e-06, + "loss": 0.5662, + "step": 1661 + }, + { + "epoch": 0.13775383340240366, + "grad_norm": 0.2607196569442749, + "learning_rate": 4.311645254869458e-06, + "loss": 0.5574, + "step": 1662 + }, + { + "epoch": 0.13783671777869871, + "grad_norm": 0.2711961567401886, + "learning_rate": 4.311230832987982e-06, + "loss": 0.5639, + "step": 1663 + }, + { + "epoch": 0.13791960215499377, + "grad_norm": 0.3175261914730072, + "learning_rate": 4.3108164111065066e-06, + "loss": 0.5717, + "step": 1664 + }, + { + "epoch": 0.13800248653128885, + "grad_norm": 0.259816974401474, + "learning_rate": 4.310401989225032e-06, + "loss": 0.5436, + "step": 1665 + }, + { + "epoch": 0.1380853709075839, + "grad_norm": 0.2540152370929718, + "learning_rate": 4.309987567343556e-06, + "loss": 0.5492, + "step": 1666 + }, + { + "epoch": 0.138168255283879, + "grad_norm": 0.2858999967575073, + "learning_rate": 4.309573145462081e-06, + "loss": 0.5711, + "step": 1667 + }, + { + "epoch": 0.13825113966017405, + "grad_norm": 0.27569910883903503, + "learning_rate": 4.309158723580605e-06, + "loss": 0.6034, + "step": 1668 + }, + { + "epoch": 0.13833402403646913, + "grad_norm": 0.2673574984073639, + "learning_rate": 4.3087443016991296e-06, + "loss": 0.584, + "step": 1669 + }, + { + "epoch": 0.1384169084127642, + "grad_norm": 0.2628247141838074, + "learning_rate": 4.308329879817655e-06, + "loss": 0.5497, + "step": 1670 + }, + { + "epoch": 0.13849979278905927, + "grad_norm": 0.2781273424625397, + "learning_rate": 4.307915457936179e-06, + "loss": 0.5348, + "step": 1671 + }, + { + "epoch": 0.13858267716535433, + "grad_norm": 0.25737789273262024, + "learning_rate": 4.3075010360547046e-06, + "loss": 0.5594, + "step": 1672 + }, + { + "epoch": 0.1386655615416494, + "grad_norm": 0.28385135531425476, + "learning_rate": 4.307086614173229e-06, + "loss": 0.5689, + "step": 1673 + }, + { + "epoch": 0.13874844591794447, + "grad_norm": 0.2999393939971924, + "learning_rate": 4.3066721922917534e-06, + "loss": 0.5825, + "step": 1674 + }, + { + "epoch": 0.13883133029423952, + "grad_norm": 0.2696375250816345, + "learning_rate": 4.306257770410278e-06, + "loss": 0.5464, + "step": 1675 + }, + { + "epoch": 0.1389142146705346, + "grad_norm": 0.2701771557331085, + "learning_rate": 4.305843348528802e-06, + "loss": 0.5618, + "step": 1676 + }, + { + "epoch": 0.13899709904682966, + "grad_norm": 0.2710022032260895, + "learning_rate": 4.3054289266473276e-06, + "loss": 0.5968, + "step": 1677 + }, + { + "epoch": 0.13907998342312475, + "grad_norm": 0.26860296726226807, + "learning_rate": 4.305014504765852e-06, + "loss": 0.552, + "step": 1678 + }, + { + "epoch": 0.1391628677994198, + "grad_norm": 0.2998672425746918, + "learning_rate": 4.3046000828843764e-06, + "loss": 0.559, + "step": 1679 + }, + { + "epoch": 0.1392457521757149, + "grad_norm": 0.2893415093421936, + "learning_rate": 4.304185661002902e-06, + "loss": 0.5255, + "step": 1680 + }, + { + "epoch": 0.13932863655200994, + "grad_norm": 0.2670835554599762, + "learning_rate": 4.303771239121426e-06, + "loss": 0.4932, + "step": 1681 + }, + { + "epoch": 0.13941152092830503, + "grad_norm": 0.2748625874519348, + "learning_rate": 4.3033568172399506e-06, + "loss": 0.5544, + "step": 1682 + }, + { + "epoch": 0.13949440530460008, + "grad_norm": 0.2648703157901764, + "learning_rate": 4.302942395358475e-06, + "loss": 0.5251, + "step": 1683 + }, + { + "epoch": 0.13957728968089514, + "grad_norm": 0.25089898705482483, + "learning_rate": 4.3025279734769994e-06, + "loss": 0.5355, + "step": 1684 + }, + { + "epoch": 0.13966017405719022, + "grad_norm": 0.26564204692840576, + "learning_rate": 4.302113551595525e-06, + "loss": 0.5561, + "step": 1685 + }, + { + "epoch": 0.13974305843348528, + "grad_norm": 0.2737155258655548, + "learning_rate": 4.301699129714049e-06, + "loss": 0.5162, + "step": 1686 + }, + { + "epoch": 0.13982594280978036, + "grad_norm": 0.3173932135105133, + "learning_rate": 4.301284707832574e-06, + "loss": 0.5745, + "step": 1687 + }, + { + "epoch": 0.13990882718607542, + "grad_norm": 0.2807474434375763, + "learning_rate": 4.300870285951099e-06, + "loss": 0.5101, + "step": 1688 + }, + { + "epoch": 0.1399917115623705, + "grad_norm": 0.3005138337612152, + "learning_rate": 4.300455864069623e-06, + "loss": 0.5908, + "step": 1689 + }, + { + "epoch": 0.14007459593866556, + "grad_norm": 0.2669462561607361, + "learning_rate": 4.300041442188148e-06, + "loss": 0.5398, + "step": 1690 + }, + { + "epoch": 0.14015748031496064, + "grad_norm": 0.27503737807273865, + "learning_rate": 4.299627020306672e-06, + "loss": 0.5673, + "step": 1691 + }, + { + "epoch": 0.1402403646912557, + "grad_norm": 0.25698792934417725, + "learning_rate": 4.299212598425197e-06, + "loss": 0.576, + "step": 1692 + }, + { + "epoch": 0.14032324906755075, + "grad_norm": 0.27407822012901306, + "learning_rate": 4.298798176543722e-06, + "loss": 0.54, + "step": 1693 + }, + { + "epoch": 0.14040613344384584, + "grad_norm": 0.2971380352973938, + "learning_rate": 4.298383754662246e-06, + "loss": 0.5442, + "step": 1694 + }, + { + "epoch": 0.1404890178201409, + "grad_norm": 0.280580997467041, + "learning_rate": 4.2979693327807716e-06, + "loss": 0.558, + "step": 1695 + }, + { + "epoch": 0.14057190219643598, + "grad_norm": 0.2961485683917999, + "learning_rate": 4.297554910899296e-06, + "loss": 0.5573, + "step": 1696 + }, + { + "epoch": 0.14065478657273103, + "grad_norm": 0.2985275089740753, + "learning_rate": 4.2971404890178204e-06, + "loss": 0.5589, + "step": 1697 + }, + { + "epoch": 0.14073767094902612, + "grad_norm": 0.26914310455322266, + "learning_rate": 4.296726067136346e-06, + "loss": 0.5228, + "step": 1698 + }, + { + "epoch": 0.14082055532532117, + "grad_norm": 0.28270745277404785, + "learning_rate": 4.29631164525487e-06, + "loss": 0.5414, + "step": 1699 + }, + { + "epoch": 0.14090343970161626, + "grad_norm": 0.2685942053794861, + "learning_rate": 4.295897223373395e-06, + "loss": 0.5331, + "step": 1700 + }, + { + "epoch": 0.1409863240779113, + "grad_norm": 0.27768346667289734, + "learning_rate": 4.295482801491919e-06, + "loss": 0.5327, + "step": 1701 + }, + { + "epoch": 0.14106920845420637, + "grad_norm": 0.254181444644928, + "learning_rate": 4.2950683796104434e-06, + "loss": 0.5383, + "step": 1702 + }, + { + "epoch": 0.14115209283050145, + "grad_norm": 0.2765803635120392, + "learning_rate": 4.294653957728969e-06, + "loss": 0.5205, + "step": 1703 + }, + { + "epoch": 0.1412349772067965, + "grad_norm": 0.2429630309343338, + "learning_rate": 4.294239535847493e-06, + "loss": 0.5168, + "step": 1704 + }, + { + "epoch": 0.1413178615830916, + "grad_norm": 0.27459123730659485, + "learning_rate": 4.293825113966018e-06, + "loss": 0.5762, + "step": 1705 + }, + { + "epoch": 0.14140074595938665, + "grad_norm": 0.28701385855674744, + "learning_rate": 4.293410692084543e-06, + "loss": 0.5755, + "step": 1706 + }, + { + "epoch": 0.14148363033568173, + "grad_norm": 0.2768159806728363, + "learning_rate": 4.292996270203067e-06, + "loss": 0.5425, + "step": 1707 + }, + { + "epoch": 0.1415665147119768, + "grad_norm": 0.26406174898147583, + "learning_rate": 4.292581848321592e-06, + "loss": 0.5475, + "step": 1708 + }, + { + "epoch": 0.14164939908827187, + "grad_norm": 0.26000452041625977, + "learning_rate": 4.292167426440116e-06, + "loss": 0.5222, + "step": 1709 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 0.28582435846328735, + "learning_rate": 4.291753004558641e-06, + "loss": 0.6094, + "step": 1710 + }, + { + "epoch": 0.141815167840862, + "grad_norm": 0.2855716049671173, + "learning_rate": 4.291338582677166e-06, + "loss": 0.5373, + "step": 1711 + }, + { + "epoch": 0.14189805221715707, + "grad_norm": 0.26058393716812134, + "learning_rate": 4.29092416079569e-06, + "loss": 0.5297, + "step": 1712 + }, + { + "epoch": 0.14198093659345212, + "grad_norm": 0.2707999050617218, + "learning_rate": 4.290509738914216e-06, + "loss": 0.5613, + "step": 1713 + }, + { + "epoch": 0.1420638209697472, + "grad_norm": 0.28181377053260803, + "learning_rate": 4.29009531703274e-06, + "loss": 0.6052, + "step": 1714 + }, + { + "epoch": 0.14214670534604226, + "grad_norm": 0.272036612033844, + "learning_rate": 4.2896808951512644e-06, + "loss": 0.6036, + "step": 1715 + }, + { + "epoch": 0.14222958972233735, + "grad_norm": 0.26384884119033813, + "learning_rate": 4.289266473269789e-06, + "loss": 0.5555, + "step": 1716 + }, + { + "epoch": 0.1423124740986324, + "grad_norm": 0.2750566005706787, + "learning_rate": 4.288852051388313e-06, + "loss": 0.5231, + "step": 1717 + }, + { + "epoch": 0.14239535847492749, + "grad_norm": 0.25701603293418884, + "learning_rate": 4.288437629506838e-06, + "loss": 0.5244, + "step": 1718 + }, + { + "epoch": 0.14247824285122254, + "grad_norm": 0.2818688452243805, + "learning_rate": 4.288023207625363e-06, + "loss": 0.5603, + "step": 1719 + }, + { + "epoch": 0.14256112722751763, + "grad_norm": 0.2662424147129059, + "learning_rate": 4.2876087857438875e-06, + "loss": 0.5715, + "step": 1720 + }, + { + "epoch": 0.14264401160381268, + "grad_norm": 0.25738129019737244, + "learning_rate": 4.287194363862413e-06, + "loss": 0.5378, + "step": 1721 + }, + { + "epoch": 0.14272689598010774, + "grad_norm": 0.26282891631126404, + "learning_rate": 4.286779941980937e-06, + "loss": 0.5901, + "step": 1722 + }, + { + "epoch": 0.14280978035640282, + "grad_norm": 0.27161285281181335, + "learning_rate": 4.286365520099462e-06, + "loss": 0.5632, + "step": 1723 + }, + { + "epoch": 0.14289266473269788, + "grad_norm": 0.2927175760269165, + "learning_rate": 4.285951098217986e-06, + "loss": 0.5796, + "step": 1724 + }, + { + "epoch": 0.14297554910899296, + "grad_norm": 0.2850145399570465, + "learning_rate": 4.2855366763365105e-06, + "loss": 0.5741, + "step": 1725 + }, + { + "epoch": 0.14305843348528802, + "grad_norm": 0.26218917965888977, + "learning_rate": 4.285122254455036e-06, + "loss": 0.5362, + "step": 1726 + }, + { + "epoch": 0.1431413178615831, + "grad_norm": 0.2738637328147888, + "learning_rate": 4.28470783257356e-06, + "loss": 0.5437, + "step": 1727 + }, + { + "epoch": 0.14322420223787816, + "grad_norm": 0.2684585750102997, + "learning_rate": 4.284293410692085e-06, + "loss": 0.5226, + "step": 1728 + }, + { + "epoch": 0.14330708661417324, + "grad_norm": 0.2716994285583496, + "learning_rate": 4.28387898881061e-06, + "loss": 0.5779, + "step": 1729 + }, + { + "epoch": 0.1433899709904683, + "grad_norm": 0.2545342743396759, + "learning_rate": 4.283464566929134e-06, + "loss": 0.4997, + "step": 1730 + }, + { + "epoch": 0.14347285536676335, + "grad_norm": 0.2845572829246521, + "learning_rate": 4.283050145047659e-06, + "loss": 0.5618, + "step": 1731 + }, + { + "epoch": 0.14355573974305844, + "grad_norm": 0.2708759605884552, + "learning_rate": 4.282635723166183e-06, + "loss": 0.5686, + "step": 1732 + }, + { + "epoch": 0.1436386241193535, + "grad_norm": 0.2682381868362427, + "learning_rate": 4.2822213012847085e-06, + "loss": 0.5587, + "step": 1733 + }, + { + "epoch": 0.14372150849564858, + "grad_norm": 0.31862473487854004, + "learning_rate": 4.281806879403233e-06, + "loss": 0.5427, + "step": 1734 + }, + { + "epoch": 0.14380439287194363, + "grad_norm": 0.2795774042606354, + "learning_rate": 4.281392457521757e-06, + "loss": 0.5599, + "step": 1735 + }, + { + "epoch": 0.14388727724823872, + "grad_norm": 0.28843915462493896, + "learning_rate": 4.280978035640282e-06, + "loss": 0.5415, + "step": 1736 + }, + { + "epoch": 0.14397016162453377, + "grad_norm": 0.2657806873321533, + "learning_rate": 4.280563613758807e-06, + "loss": 0.5423, + "step": 1737 + }, + { + "epoch": 0.14405304600082885, + "grad_norm": 0.257146418094635, + "learning_rate": 4.2801491918773315e-06, + "loss": 0.5908, + "step": 1738 + }, + { + "epoch": 0.1441359303771239, + "grad_norm": 0.2664475739002228, + "learning_rate": 4.279734769995857e-06, + "loss": 0.5341, + "step": 1739 + }, + { + "epoch": 0.14421881475341897, + "grad_norm": 0.27432116866111755, + "learning_rate": 4.279320348114381e-06, + "loss": 0.5835, + "step": 1740 + }, + { + "epoch": 0.14430169912971405, + "grad_norm": 0.2707568407058716, + "learning_rate": 4.278905926232906e-06, + "loss": 0.5996, + "step": 1741 + }, + { + "epoch": 0.1443845835060091, + "grad_norm": 0.28005385398864746, + "learning_rate": 4.27849150435143e-06, + "loss": 0.5883, + "step": 1742 + }, + { + "epoch": 0.1444674678823042, + "grad_norm": 0.28041183948516846, + "learning_rate": 4.2780770824699545e-06, + "loss": 0.5293, + "step": 1743 + }, + { + "epoch": 0.14455035225859925, + "grad_norm": 0.2564071714878082, + "learning_rate": 4.27766266058848e-06, + "loss": 0.5145, + "step": 1744 + }, + { + "epoch": 0.14463323663489433, + "grad_norm": 0.2908639907836914, + "learning_rate": 4.277248238707004e-06, + "loss": 0.6141, + "step": 1745 + }, + { + "epoch": 0.14471612101118939, + "grad_norm": 0.2563062012195587, + "learning_rate": 4.276833816825529e-06, + "loss": 0.4982, + "step": 1746 + }, + { + "epoch": 0.14479900538748447, + "grad_norm": 0.2731834650039673, + "learning_rate": 4.276419394944054e-06, + "loss": 0.572, + "step": 1747 + }, + { + "epoch": 0.14488188976377953, + "grad_norm": 0.2645590603351593, + "learning_rate": 4.276004973062578e-06, + "loss": 0.4986, + "step": 1748 + }, + { + "epoch": 0.1449647741400746, + "grad_norm": 0.28606581687927246, + "learning_rate": 4.275590551181103e-06, + "loss": 0.5991, + "step": 1749 + }, + { + "epoch": 0.14504765851636967, + "grad_norm": 0.271640807390213, + "learning_rate": 4.275176129299627e-06, + "loss": 0.5623, + "step": 1750 + }, + { + "epoch": 0.14513054289266472, + "grad_norm": 0.2561786472797394, + "learning_rate": 4.274761707418152e-06, + "loss": 0.5397, + "step": 1751 + }, + { + "epoch": 0.1452134272689598, + "grad_norm": 0.27662578225135803, + "learning_rate": 4.274347285536677e-06, + "loss": 0.6084, + "step": 1752 + }, + { + "epoch": 0.14529631164525486, + "grad_norm": 0.2667611241340637, + "learning_rate": 4.273932863655201e-06, + "loss": 0.5666, + "step": 1753 + }, + { + "epoch": 0.14537919602154994, + "grad_norm": 0.27839094400405884, + "learning_rate": 4.273518441773726e-06, + "loss": 0.5645, + "step": 1754 + }, + { + "epoch": 0.145462080397845, + "grad_norm": 0.26895421743392944, + "learning_rate": 4.273104019892251e-06, + "loss": 0.5743, + "step": 1755 + }, + { + "epoch": 0.14554496477414008, + "grad_norm": 0.24678248167037964, + "learning_rate": 4.2726895980107755e-06, + "loss": 0.5059, + "step": 1756 + }, + { + "epoch": 0.14562784915043514, + "grad_norm": 0.284420371055603, + "learning_rate": 4.2722751761293e-06, + "loss": 0.6121, + "step": 1757 + }, + { + "epoch": 0.14571073352673022, + "grad_norm": 0.26515284180641174, + "learning_rate": 4.271860754247824e-06, + "loss": 0.5347, + "step": 1758 + }, + { + "epoch": 0.14579361790302528, + "grad_norm": 0.255928635597229, + "learning_rate": 4.271446332366349e-06, + "loss": 0.5226, + "step": 1759 + }, + { + "epoch": 0.14587650227932034, + "grad_norm": 0.2705543637275696, + "learning_rate": 4.271031910484874e-06, + "loss": 0.5523, + "step": 1760 + }, + { + "epoch": 0.14595938665561542, + "grad_norm": 0.2641673684120178, + "learning_rate": 4.2706174886033985e-06, + "loss": 0.5444, + "step": 1761 + }, + { + "epoch": 0.14604227103191048, + "grad_norm": 0.28166326880455017, + "learning_rate": 4.270203066721924e-06, + "loss": 0.585, + "step": 1762 + }, + { + "epoch": 0.14612515540820556, + "grad_norm": 0.25896739959716797, + "learning_rate": 4.269788644840448e-06, + "loss": 0.5353, + "step": 1763 + }, + { + "epoch": 0.14620803978450062, + "grad_norm": 0.2902170419692993, + "learning_rate": 4.269374222958973e-06, + "loss": 0.5589, + "step": 1764 + }, + { + "epoch": 0.1462909241607957, + "grad_norm": 0.30642521381378174, + "learning_rate": 4.268959801077497e-06, + "loss": 0.5463, + "step": 1765 + }, + { + "epoch": 0.14637380853709075, + "grad_norm": 0.2809872627258301, + "learning_rate": 4.2685453791960215e-06, + "loss": 0.6007, + "step": 1766 + }, + { + "epoch": 0.14645669291338584, + "grad_norm": 0.29176244139671326, + "learning_rate": 4.268130957314546e-06, + "loss": 0.5952, + "step": 1767 + }, + { + "epoch": 0.1465395772896809, + "grad_norm": 0.26443272829055786, + "learning_rate": 4.267716535433071e-06, + "loss": 0.5343, + "step": 1768 + }, + { + "epoch": 0.14662246166597595, + "grad_norm": 0.29710614681243896, + "learning_rate": 4.267302113551596e-06, + "loss": 0.5347, + "step": 1769 + }, + { + "epoch": 0.14670534604227103, + "grad_norm": 0.28707796335220337, + "learning_rate": 4.266887691670121e-06, + "loss": 0.5848, + "step": 1770 + }, + { + "epoch": 0.1467882304185661, + "grad_norm": 0.27212873101234436, + "learning_rate": 4.266473269788645e-06, + "loss": 0.5759, + "step": 1771 + }, + { + "epoch": 0.14687111479486117, + "grad_norm": 0.2709381878376007, + "learning_rate": 4.26605884790717e-06, + "loss": 0.5803, + "step": 1772 + }, + { + "epoch": 0.14695399917115623, + "grad_norm": 0.3104915916919708, + "learning_rate": 4.265644426025694e-06, + "loss": 0.5747, + "step": 1773 + }, + { + "epoch": 0.1470368835474513, + "grad_norm": 0.2698417007923126, + "learning_rate": 4.2652300041442195e-06, + "loss": 0.526, + "step": 1774 + }, + { + "epoch": 0.14711976792374637, + "grad_norm": 0.2719137668609619, + "learning_rate": 4.264815582262744e-06, + "loss": 0.5442, + "step": 1775 + }, + { + "epoch": 0.14720265230004145, + "grad_norm": 0.27978095412254333, + "learning_rate": 4.264401160381268e-06, + "loss": 0.5561, + "step": 1776 + }, + { + "epoch": 0.1472855366763365, + "grad_norm": 0.2645707428455353, + "learning_rate": 4.263986738499793e-06, + "loss": 0.5657, + "step": 1777 + }, + { + "epoch": 0.14736842105263157, + "grad_norm": 0.27960363030433655, + "learning_rate": 4.263572316618318e-06, + "loss": 0.5479, + "step": 1778 + }, + { + "epoch": 0.14745130542892665, + "grad_norm": 0.28003883361816406, + "learning_rate": 4.2631578947368425e-06, + "loss": 0.5742, + "step": 1779 + }, + { + "epoch": 0.1475341898052217, + "grad_norm": 0.30412012338638306, + "learning_rate": 4.262743472855368e-06, + "loss": 0.5724, + "step": 1780 + }, + { + "epoch": 0.1476170741815168, + "grad_norm": 0.2649060785770416, + "learning_rate": 4.262329050973892e-06, + "loss": 0.5095, + "step": 1781 + }, + { + "epoch": 0.14769995855781184, + "grad_norm": 0.25302785634994507, + "learning_rate": 4.261914629092417e-06, + "loss": 0.5299, + "step": 1782 + }, + { + "epoch": 0.14778284293410693, + "grad_norm": 0.2846055328845978, + "learning_rate": 4.261500207210941e-06, + "loss": 0.6034, + "step": 1783 + }, + { + "epoch": 0.14786572731040198, + "grad_norm": 0.2680841088294983, + "learning_rate": 4.2610857853294655e-06, + "loss": 0.5401, + "step": 1784 + }, + { + "epoch": 0.14794861168669707, + "grad_norm": 0.25780385732650757, + "learning_rate": 4.26067136344799e-06, + "loss": 0.5508, + "step": 1785 + }, + { + "epoch": 0.14803149606299212, + "grad_norm": 0.2594532072544098, + "learning_rate": 4.260256941566515e-06, + "loss": 0.5367, + "step": 1786 + }, + { + "epoch": 0.1481143804392872, + "grad_norm": 0.2615540027618408, + "learning_rate": 4.25984251968504e-06, + "loss": 0.5305, + "step": 1787 + }, + { + "epoch": 0.14819726481558226, + "grad_norm": 0.28790387511253357, + "learning_rate": 4.259428097803565e-06, + "loss": 0.5328, + "step": 1788 + }, + { + "epoch": 0.14828014919187732, + "grad_norm": 0.29342079162597656, + "learning_rate": 4.259013675922089e-06, + "loss": 0.5768, + "step": 1789 + }, + { + "epoch": 0.1483630335681724, + "grad_norm": 0.2570093870162964, + "learning_rate": 4.258599254040614e-06, + "loss": 0.5488, + "step": 1790 + }, + { + "epoch": 0.14844591794446746, + "grad_norm": 0.2769079804420471, + "learning_rate": 4.258184832159138e-06, + "loss": 0.564, + "step": 1791 + }, + { + "epoch": 0.14852880232076254, + "grad_norm": 0.2697581350803375, + "learning_rate": 4.257770410277663e-06, + "loss": 0.5654, + "step": 1792 + }, + { + "epoch": 0.1486116866970576, + "grad_norm": 0.27945297956466675, + "learning_rate": 4.257355988396188e-06, + "loss": 0.5503, + "step": 1793 + }, + { + "epoch": 0.14869457107335268, + "grad_norm": 0.24802587926387787, + "learning_rate": 4.256941566514712e-06, + "loss": 0.5527, + "step": 1794 + }, + { + "epoch": 0.14877745544964774, + "grad_norm": 0.2701224684715271, + "learning_rate": 4.256527144633237e-06, + "loss": 0.5151, + "step": 1795 + }, + { + "epoch": 0.14886033982594282, + "grad_norm": 0.2826712131500244, + "learning_rate": 4.256112722751762e-06, + "loss": 0.5883, + "step": 1796 + }, + { + "epoch": 0.14894322420223788, + "grad_norm": 0.2625892460346222, + "learning_rate": 4.2556983008702865e-06, + "loss": 0.5527, + "step": 1797 + }, + { + "epoch": 0.14902610857853293, + "grad_norm": 0.26786547899246216, + "learning_rate": 4.255283878988811e-06, + "loss": 0.5729, + "step": 1798 + }, + { + "epoch": 0.14910899295482802, + "grad_norm": 0.2695467472076416, + "learning_rate": 4.254869457107335e-06, + "loss": 0.5774, + "step": 1799 + }, + { + "epoch": 0.14919187733112307, + "grad_norm": 0.27063116431236267, + "learning_rate": 4.25445503522586e-06, + "loss": 0.5139, + "step": 1800 + }, + { + "epoch": 0.14927476170741816, + "grad_norm": 0.25972333550453186, + "learning_rate": 4.254040613344385e-06, + "loss": 0.5117, + "step": 1801 + }, + { + "epoch": 0.1493576460837132, + "grad_norm": 0.26561757922172546, + "learning_rate": 4.2536261914629095e-06, + "loss": 0.5923, + "step": 1802 + }, + { + "epoch": 0.1494405304600083, + "grad_norm": 0.2632332444190979, + "learning_rate": 4.253211769581434e-06, + "loss": 0.5959, + "step": 1803 + }, + { + "epoch": 0.14952341483630335, + "grad_norm": 0.28568193316459656, + "learning_rate": 4.252797347699959e-06, + "loss": 0.5649, + "step": 1804 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 0.2797972857952118, + "learning_rate": 4.252382925818484e-06, + "loss": 0.5853, + "step": 1805 + }, + { + "epoch": 0.1496891835888935, + "grad_norm": 0.2702978253364563, + "learning_rate": 4.251968503937008e-06, + "loss": 0.5112, + "step": 1806 + }, + { + "epoch": 0.14977206796518855, + "grad_norm": 0.2605855166912079, + "learning_rate": 4.2515540820555325e-06, + "loss": 0.555, + "step": 1807 + }, + { + "epoch": 0.14985495234148363, + "grad_norm": 0.25785329937934875, + "learning_rate": 4.251139660174057e-06, + "loss": 0.5275, + "step": 1808 + }, + { + "epoch": 0.1499378367177787, + "grad_norm": 0.2762613892555237, + "learning_rate": 4.250725238292582e-06, + "loss": 0.5865, + "step": 1809 + }, + { + "epoch": 0.15002072109407377, + "grad_norm": 0.26050519943237305, + "learning_rate": 4.250310816411107e-06, + "loss": 0.5525, + "step": 1810 + }, + { + "epoch": 0.15010360547036883, + "grad_norm": 0.25268733501434326, + "learning_rate": 4.249896394529632e-06, + "loss": 0.5225, + "step": 1811 + }, + { + "epoch": 0.1501864898466639, + "grad_norm": 0.25539523363113403, + "learning_rate": 4.249481972648156e-06, + "loss": 0.5122, + "step": 1812 + }, + { + "epoch": 0.15026937422295897, + "grad_norm": 0.2728694975376129, + "learning_rate": 4.249067550766681e-06, + "loss": 0.5511, + "step": 1813 + }, + { + "epoch": 0.15035225859925405, + "grad_norm": 0.29335442185401917, + "learning_rate": 4.248653128885206e-06, + "loss": 0.562, + "step": 1814 + }, + { + "epoch": 0.1504351429755491, + "grad_norm": 0.2425563782453537, + "learning_rate": 4.2482387070037305e-06, + "loss": 0.5205, + "step": 1815 + }, + { + "epoch": 0.15051802735184416, + "grad_norm": 0.26460596919059753, + "learning_rate": 4.247824285122255e-06, + "loss": 0.4978, + "step": 1816 + }, + { + "epoch": 0.15060091172813925, + "grad_norm": 0.2673284411430359, + "learning_rate": 4.247409863240779e-06, + "loss": 0.5641, + "step": 1817 + }, + { + "epoch": 0.1506837961044343, + "grad_norm": 0.2396930456161499, + "learning_rate": 4.246995441359304e-06, + "loss": 0.5011, + "step": 1818 + }, + { + "epoch": 0.1507666804807294, + "grad_norm": 0.3084789514541626, + "learning_rate": 4.246581019477829e-06, + "loss": 0.5789, + "step": 1819 + }, + { + "epoch": 0.15084956485702444, + "grad_norm": 0.29116374254226685, + "learning_rate": 4.2461665975963535e-06, + "loss": 0.5618, + "step": 1820 + }, + { + "epoch": 0.15093244923331953, + "grad_norm": 0.25358980894088745, + "learning_rate": 4.245752175714878e-06, + "loss": 0.5418, + "step": 1821 + }, + { + "epoch": 0.15101533360961458, + "grad_norm": 0.30477315187454224, + "learning_rate": 4.245337753833403e-06, + "loss": 0.5798, + "step": 1822 + }, + { + "epoch": 0.15109821798590967, + "grad_norm": 0.25209420919418335, + "learning_rate": 4.244923331951928e-06, + "loss": 0.5294, + "step": 1823 + }, + { + "epoch": 0.15118110236220472, + "grad_norm": 0.2870826721191406, + "learning_rate": 4.244508910070452e-06, + "loss": 0.6018, + "step": 1824 + }, + { + "epoch": 0.1512639867384998, + "grad_norm": 0.26396510004997253, + "learning_rate": 4.2440944881889765e-06, + "loss": 0.5393, + "step": 1825 + }, + { + "epoch": 0.15134687111479486, + "grad_norm": 0.2524496614933014, + "learning_rate": 4.243680066307501e-06, + "loss": 0.5593, + "step": 1826 + }, + { + "epoch": 0.15142975549108992, + "grad_norm": 0.2928551137447357, + "learning_rate": 4.243265644426026e-06, + "loss": 0.6014, + "step": 1827 + }, + { + "epoch": 0.151512639867385, + "grad_norm": 0.27711066603660583, + "learning_rate": 4.242851222544551e-06, + "loss": 0.5742, + "step": 1828 + }, + { + "epoch": 0.15159552424368006, + "grad_norm": 0.2958548963069916, + "learning_rate": 4.242436800663076e-06, + "loss": 0.5164, + "step": 1829 + }, + { + "epoch": 0.15167840861997514, + "grad_norm": 0.27645784616470337, + "learning_rate": 4.2420223787816e-06, + "loss": 0.5679, + "step": 1830 + }, + { + "epoch": 0.1517612929962702, + "grad_norm": 0.37783828377723694, + "learning_rate": 4.241607956900125e-06, + "loss": 0.5433, + "step": 1831 + }, + { + "epoch": 0.15184417737256528, + "grad_norm": 0.2580871880054474, + "learning_rate": 4.241193535018649e-06, + "loss": 0.5103, + "step": 1832 + }, + { + "epoch": 0.15192706174886034, + "grad_norm": 0.29404810070991516, + "learning_rate": 4.240779113137174e-06, + "loss": 0.5312, + "step": 1833 + }, + { + "epoch": 0.15200994612515542, + "grad_norm": 0.2734946310520172, + "learning_rate": 4.240364691255698e-06, + "loss": 0.5557, + "step": 1834 + }, + { + "epoch": 0.15209283050145048, + "grad_norm": 0.2792963683605194, + "learning_rate": 4.239950269374223e-06, + "loss": 0.5963, + "step": 1835 + }, + { + "epoch": 0.15217571487774553, + "grad_norm": 0.27902570366859436, + "learning_rate": 4.239535847492748e-06, + "loss": 0.5785, + "step": 1836 + }, + { + "epoch": 0.15225859925404062, + "grad_norm": 0.2616313695907593, + "learning_rate": 4.239121425611273e-06, + "loss": 0.5184, + "step": 1837 + }, + { + "epoch": 0.15234148363033567, + "grad_norm": 0.27760109305381775, + "learning_rate": 4.2387070037297975e-06, + "loss": 0.5951, + "step": 1838 + }, + { + "epoch": 0.15242436800663076, + "grad_norm": 0.2674952745437622, + "learning_rate": 4.238292581848322e-06, + "loss": 0.5565, + "step": 1839 + }, + { + "epoch": 0.1525072523829258, + "grad_norm": 0.269868940114975, + "learning_rate": 4.237878159966846e-06, + "loss": 0.5153, + "step": 1840 + }, + { + "epoch": 0.1525901367592209, + "grad_norm": 0.24599729478359222, + "learning_rate": 4.237463738085371e-06, + "loss": 0.5306, + "step": 1841 + }, + { + "epoch": 0.15267302113551595, + "grad_norm": 0.6079660654067993, + "learning_rate": 4.237049316203896e-06, + "loss": 0.5304, + "step": 1842 + }, + { + "epoch": 0.15275590551181104, + "grad_norm": 0.2864278256893158, + "learning_rate": 4.2366348943224205e-06, + "loss": 0.5728, + "step": 1843 + }, + { + "epoch": 0.1528387898881061, + "grad_norm": 0.27228835225105286, + "learning_rate": 4.236220472440945e-06, + "loss": 0.6068, + "step": 1844 + }, + { + "epoch": 0.15292167426440115, + "grad_norm": 0.2958449721336365, + "learning_rate": 4.23580605055947e-06, + "loss": 0.5587, + "step": 1845 + }, + { + "epoch": 0.15300455864069623, + "grad_norm": 0.27066853642463684, + "learning_rate": 4.235391628677995e-06, + "loss": 0.5608, + "step": 1846 + }, + { + "epoch": 0.1530874430169913, + "grad_norm": 0.2649061381816864, + "learning_rate": 4.234977206796519e-06, + "loss": 0.5092, + "step": 1847 + }, + { + "epoch": 0.15317032739328637, + "grad_norm": 0.2782347500324249, + "learning_rate": 4.2345627849150435e-06, + "loss": 0.5918, + "step": 1848 + }, + { + "epoch": 0.15325321176958143, + "grad_norm": 0.29520606994628906, + "learning_rate": 4.234148363033568e-06, + "loss": 0.5531, + "step": 1849 + }, + { + "epoch": 0.1533360961458765, + "grad_norm": 0.28510841727256775, + "learning_rate": 4.233733941152093e-06, + "loss": 0.5861, + "step": 1850 + }, + { + "epoch": 0.15341898052217157, + "grad_norm": 0.2871469259262085, + "learning_rate": 4.233319519270618e-06, + "loss": 0.5588, + "step": 1851 + }, + { + "epoch": 0.15350186489846665, + "grad_norm": 0.2667398452758789, + "learning_rate": 4.232905097389142e-06, + "loss": 0.5562, + "step": 1852 + }, + { + "epoch": 0.1535847492747617, + "grad_norm": 0.30248406529426575, + "learning_rate": 4.232490675507667e-06, + "loss": 0.6002, + "step": 1853 + }, + { + "epoch": 0.15366763365105676, + "grad_norm": 0.2782074809074402, + "learning_rate": 4.232076253626192e-06, + "loss": 0.5252, + "step": 1854 + }, + { + "epoch": 0.15375051802735185, + "grad_norm": 0.2766932249069214, + "learning_rate": 4.231661831744717e-06, + "loss": 0.5696, + "step": 1855 + }, + { + "epoch": 0.1538334024036469, + "grad_norm": 0.29575294256210327, + "learning_rate": 4.2312474098632415e-06, + "loss": 0.503, + "step": 1856 + }, + { + "epoch": 0.15391628677994199, + "grad_norm": 0.2765498459339142, + "learning_rate": 4.230832987981766e-06, + "loss": 0.5789, + "step": 1857 + }, + { + "epoch": 0.15399917115623704, + "grad_norm": 0.2657349705696106, + "learning_rate": 4.23041856610029e-06, + "loss": 0.5598, + "step": 1858 + }, + { + "epoch": 0.15408205553253213, + "grad_norm": 0.29461055994033813, + "learning_rate": 4.230004144218815e-06, + "loss": 0.5505, + "step": 1859 + }, + { + "epoch": 0.15416493990882718, + "grad_norm": 0.2798333466053009, + "learning_rate": 4.22958972233734e-06, + "loss": 0.5954, + "step": 1860 + }, + { + "epoch": 0.15424782428512226, + "grad_norm": 0.28567254543304443, + "learning_rate": 4.2291753004558645e-06, + "loss": 0.5831, + "step": 1861 + }, + { + "epoch": 0.15433070866141732, + "grad_norm": 0.2687372863292694, + "learning_rate": 4.228760878574389e-06, + "loss": 0.5574, + "step": 1862 + }, + { + "epoch": 0.1544135930377124, + "grad_norm": 0.2678685784339905, + "learning_rate": 4.228346456692914e-06, + "loss": 0.5368, + "step": 1863 + }, + { + "epoch": 0.15449647741400746, + "grad_norm": 0.27275487780570984, + "learning_rate": 4.227932034811439e-06, + "loss": 0.5665, + "step": 1864 + }, + { + "epoch": 0.15457936179030252, + "grad_norm": 0.2543792128562927, + "learning_rate": 4.227517612929963e-06, + "loss": 0.5348, + "step": 1865 + }, + { + "epoch": 0.1546622461665976, + "grad_norm": 0.28110694885253906, + "learning_rate": 4.2271031910484875e-06, + "loss": 0.5888, + "step": 1866 + }, + { + "epoch": 0.15474513054289266, + "grad_norm": 0.24935384094715118, + "learning_rate": 4.226688769167012e-06, + "loss": 0.5079, + "step": 1867 + }, + { + "epoch": 0.15482801491918774, + "grad_norm": 0.27127379179000854, + "learning_rate": 4.226274347285537e-06, + "loss": 0.5693, + "step": 1868 + }, + { + "epoch": 0.1549108992954828, + "grad_norm": 0.2588350772857666, + "learning_rate": 4.225859925404062e-06, + "loss": 0.5271, + "step": 1869 + }, + { + "epoch": 0.15499378367177788, + "grad_norm": 0.25457122921943665, + "learning_rate": 4.225445503522586e-06, + "loss": 0.5533, + "step": 1870 + }, + { + "epoch": 0.15507666804807294, + "grad_norm": 0.28135257959365845, + "learning_rate": 4.225031081641111e-06, + "loss": 0.5828, + "step": 1871 + }, + { + "epoch": 0.15515955242436802, + "grad_norm": 0.2434971183538437, + "learning_rate": 4.224616659759636e-06, + "loss": 0.5281, + "step": 1872 + }, + { + "epoch": 0.15524243680066308, + "grad_norm": 0.2769758403301239, + "learning_rate": 4.22420223787816e-06, + "loss": 0.5175, + "step": 1873 + }, + { + "epoch": 0.15532532117695813, + "grad_norm": 0.25638318061828613, + "learning_rate": 4.223787815996685e-06, + "loss": 0.5416, + "step": 1874 + }, + { + "epoch": 0.15540820555325321, + "grad_norm": 0.25841161608695984, + "learning_rate": 4.223373394115209e-06, + "loss": 0.5887, + "step": 1875 + }, + { + "epoch": 0.15549108992954827, + "grad_norm": 0.2742508053779602, + "learning_rate": 4.222958972233734e-06, + "loss": 0.5236, + "step": 1876 + }, + { + "epoch": 0.15557397430584335, + "grad_norm": 0.2738596796989441, + "learning_rate": 4.222544550352259e-06, + "loss": 0.5867, + "step": 1877 + }, + { + "epoch": 0.1556568586821384, + "grad_norm": 0.2650969624519348, + "learning_rate": 4.222130128470784e-06, + "loss": 0.5724, + "step": 1878 + }, + { + "epoch": 0.1557397430584335, + "grad_norm": 0.26592594385147095, + "learning_rate": 4.2217157065893085e-06, + "loss": 0.5524, + "step": 1879 + }, + { + "epoch": 0.15582262743472855, + "grad_norm": 0.275267630815506, + "learning_rate": 4.221301284707833e-06, + "loss": 0.5566, + "step": 1880 + }, + { + "epoch": 0.15590551181102363, + "grad_norm": 0.2667222023010254, + "learning_rate": 4.220886862826357e-06, + "loss": 0.5398, + "step": 1881 + }, + { + "epoch": 0.1559883961873187, + "grad_norm": 0.2477063238620758, + "learning_rate": 4.220472440944882e-06, + "loss": 0.5051, + "step": 1882 + }, + { + "epoch": 0.15607128056361375, + "grad_norm": 0.2829664647579193, + "learning_rate": 4.220058019063406e-06, + "loss": 0.5557, + "step": 1883 + }, + { + "epoch": 0.15615416493990883, + "grad_norm": 0.28973567485809326, + "learning_rate": 4.2196435971819315e-06, + "loss": 0.5613, + "step": 1884 + }, + { + "epoch": 0.15623704931620389, + "grad_norm": 0.29911860823631287, + "learning_rate": 4.219229175300456e-06, + "loss": 0.5676, + "step": 1885 + }, + { + "epoch": 0.15631993369249897, + "grad_norm": 0.26925766468048096, + "learning_rate": 4.218814753418981e-06, + "loss": 0.5433, + "step": 1886 + }, + { + "epoch": 0.15640281806879403, + "grad_norm": 0.26011285185813904, + "learning_rate": 4.218400331537506e-06, + "loss": 0.5375, + "step": 1887 + }, + { + "epoch": 0.1564857024450891, + "grad_norm": 0.27961525321006775, + "learning_rate": 4.21798590965603e-06, + "loss": 0.5533, + "step": 1888 + }, + { + "epoch": 0.15656858682138416, + "grad_norm": 0.29485729336738586, + "learning_rate": 4.2175714877745545e-06, + "loss": 0.5861, + "step": 1889 + }, + { + "epoch": 0.15665147119767925, + "grad_norm": 0.2844937741756439, + "learning_rate": 4.217157065893079e-06, + "loss": 0.5306, + "step": 1890 + }, + { + "epoch": 0.1567343555739743, + "grad_norm": 0.29659023880958557, + "learning_rate": 4.216742644011604e-06, + "loss": 0.5016, + "step": 1891 + }, + { + "epoch": 0.15681723995026936, + "grad_norm": 0.28529301285743713, + "learning_rate": 4.216328222130129e-06, + "loss": 0.5818, + "step": 1892 + }, + { + "epoch": 0.15690012432656444, + "grad_norm": 0.27522245049476624, + "learning_rate": 4.215913800248653e-06, + "loss": 0.5907, + "step": 1893 + }, + { + "epoch": 0.1569830087028595, + "grad_norm": 0.26932865381240845, + "learning_rate": 4.215499378367178e-06, + "loss": 0.5309, + "step": 1894 + }, + { + "epoch": 0.15706589307915458, + "grad_norm": 0.32195374369621277, + "learning_rate": 4.215084956485703e-06, + "loss": 0.5944, + "step": 1895 + }, + { + "epoch": 0.15714877745544964, + "grad_norm": 0.31900128722190857, + "learning_rate": 4.214670534604228e-06, + "loss": 0.5629, + "step": 1896 + }, + { + "epoch": 0.15723166183174472, + "grad_norm": 0.29316234588623047, + "learning_rate": 4.2142561127227525e-06, + "loss": 0.5301, + "step": 1897 + }, + { + "epoch": 0.15731454620803978, + "grad_norm": 0.2621859014034271, + "learning_rate": 4.213841690841277e-06, + "loss": 0.5321, + "step": 1898 + }, + { + "epoch": 0.15739743058433486, + "grad_norm": 0.274506151676178, + "learning_rate": 4.213427268959801e-06, + "loss": 0.5326, + "step": 1899 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 0.2841678261756897, + "learning_rate": 4.213012847078326e-06, + "loss": 0.5595, + "step": 1900 + }, + { + "epoch": 0.157563199336925, + "grad_norm": 0.2788396179676056, + "learning_rate": 4.21259842519685e-06, + "loss": 0.5014, + "step": 1901 + }, + { + "epoch": 0.15764608371322006, + "grad_norm": 0.2711344063282013, + "learning_rate": 4.2121840033153755e-06, + "loss": 0.5369, + "step": 1902 + }, + { + "epoch": 0.15772896808951511, + "grad_norm": 0.282368004322052, + "learning_rate": 4.2117695814339e-06, + "loss": 0.5256, + "step": 1903 + }, + { + "epoch": 0.1578118524658102, + "grad_norm": 0.2754930555820465, + "learning_rate": 4.211355159552425e-06, + "loss": 0.5544, + "step": 1904 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 0.28224897384643555, + "learning_rate": 4.21094073767095e-06, + "loss": 0.5682, + "step": 1905 + }, + { + "epoch": 0.15797762121840034, + "grad_norm": 0.2902466356754303, + "learning_rate": 4.210526315789474e-06, + "loss": 0.5873, + "step": 1906 + }, + { + "epoch": 0.1580605055946954, + "grad_norm": 0.25453484058380127, + "learning_rate": 4.2101118939079985e-06, + "loss": 0.5101, + "step": 1907 + }, + { + "epoch": 0.15814338997099048, + "grad_norm": 0.27739229798316956, + "learning_rate": 4.209697472026523e-06, + "loss": 0.5787, + "step": 1908 + }, + { + "epoch": 0.15822627434728553, + "grad_norm": 0.2887730002403259, + "learning_rate": 4.209283050145048e-06, + "loss": 0.5421, + "step": 1909 + }, + { + "epoch": 0.15830915872358062, + "grad_norm": 0.27504026889801025, + "learning_rate": 4.208868628263573e-06, + "loss": 0.5518, + "step": 1910 + }, + { + "epoch": 0.15839204309987567, + "grad_norm": 0.290152907371521, + "learning_rate": 4.208454206382097e-06, + "loss": 0.5403, + "step": 1911 + }, + { + "epoch": 0.15847492747617073, + "grad_norm": 0.2566439211368561, + "learning_rate": 4.208039784500622e-06, + "loss": 0.5275, + "step": 1912 + }, + { + "epoch": 0.1585578118524658, + "grad_norm": 0.26218381524086, + "learning_rate": 4.207625362619147e-06, + "loss": 0.5408, + "step": 1913 + }, + { + "epoch": 0.15864069622876087, + "grad_norm": 0.2873484790325165, + "learning_rate": 4.207210940737671e-06, + "loss": 0.59, + "step": 1914 + }, + { + "epoch": 0.15872358060505595, + "grad_norm": 0.2732827365398407, + "learning_rate": 4.206796518856196e-06, + "loss": 0.5472, + "step": 1915 + }, + { + "epoch": 0.158806464981351, + "grad_norm": 0.28533899784088135, + "learning_rate": 4.20638209697472e-06, + "loss": 0.6183, + "step": 1916 + }, + { + "epoch": 0.1588893493576461, + "grad_norm": 0.28466692566871643, + "learning_rate": 4.205967675093245e-06, + "loss": 0.5859, + "step": 1917 + }, + { + "epoch": 0.15897223373394115, + "grad_norm": 0.27268388867378235, + "learning_rate": 4.20555325321177e-06, + "loss": 0.5218, + "step": 1918 + }, + { + "epoch": 0.15905511811023623, + "grad_norm": 0.32592371106147766, + "learning_rate": 4.205138831330294e-06, + "loss": 0.5542, + "step": 1919 + }, + { + "epoch": 0.1591380024865313, + "grad_norm": 0.2834593653678894, + "learning_rate": 4.2047244094488196e-06, + "loss": 0.579, + "step": 1920 + }, + { + "epoch": 0.15922088686282634, + "grad_norm": 0.27913281321525574, + "learning_rate": 4.204309987567344e-06, + "loss": 0.5611, + "step": 1921 + }, + { + "epoch": 0.15930377123912143, + "grad_norm": 0.2914995849132538, + "learning_rate": 4.203895565685868e-06, + "loss": 0.5778, + "step": 1922 + }, + { + "epoch": 0.15938665561541648, + "grad_norm": 0.28368231654167175, + "learning_rate": 4.203481143804393e-06, + "loss": 0.5873, + "step": 1923 + }, + { + "epoch": 0.15946953999171157, + "grad_norm": 0.2768825590610504, + "learning_rate": 4.203066721922917e-06, + "loss": 0.4913, + "step": 1924 + }, + { + "epoch": 0.15955242436800662, + "grad_norm": 0.2846238613128662, + "learning_rate": 4.2026523000414426e-06, + "loss": 0.5831, + "step": 1925 + }, + { + "epoch": 0.1596353087443017, + "grad_norm": 0.28491559624671936, + "learning_rate": 4.202237878159967e-06, + "loss": 0.5682, + "step": 1926 + }, + { + "epoch": 0.15971819312059676, + "grad_norm": 0.27987778186798096, + "learning_rate": 4.201823456278492e-06, + "loss": 0.5945, + "step": 1927 + }, + { + "epoch": 0.15980107749689185, + "grad_norm": 0.2560672461986542, + "learning_rate": 4.201409034397017e-06, + "loss": 0.5532, + "step": 1928 + }, + { + "epoch": 0.1598839618731869, + "grad_norm": 0.31468021869659424, + "learning_rate": 4.200994612515541e-06, + "loss": 0.6115, + "step": 1929 + }, + { + "epoch": 0.159966846249482, + "grad_norm": 0.2669489085674286, + "learning_rate": 4.2005801906340656e-06, + "loss": 0.5048, + "step": 1930 + }, + { + "epoch": 0.16004973062577704, + "grad_norm": 0.2547210454940796, + "learning_rate": 4.20016576875259e-06, + "loss": 0.5433, + "step": 1931 + }, + { + "epoch": 0.1601326150020721, + "grad_norm": 0.2799730598926544, + "learning_rate": 4.199751346871115e-06, + "loss": 0.5828, + "step": 1932 + }, + { + "epoch": 0.16021549937836718, + "grad_norm": 0.2640237510204315, + "learning_rate": 4.19933692498964e-06, + "loss": 0.5654, + "step": 1933 + }, + { + "epoch": 0.16029838375466224, + "grad_norm": 0.26056697964668274, + "learning_rate": 4.198922503108164e-06, + "loss": 0.5692, + "step": 1934 + }, + { + "epoch": 0.16038126813095732, + "grad_norm": 0.26559358835220337, + "learning_rate": 4.198508081226689e-06, + "loss": 0.5755, + "step": 1935 + }, + { + "epoch": 0.16046415250725238, + "grad_norm": 0.26949816942214966, + "learning_rate": 4.198093659345214e-06, + "loss": 0.545, + "step": 1936 + }, + { + "epoch": 0.16054703688354746, + "grad_norm": 0.26607683300971985, + "learning_rate": 4.197679237463738e-06, + "loss": 0.5563, + "step": 1937 + }, + { + "epoch": 0.16062992125984252, + "grad_norm": 0.2661038041114807, + "learning_rate": 4.1972648155822636e-06, + "loss": 0.5492, + "step": 1938 + }, + { + "epoch": 0.1607128056361376, + "grad_norm": 0.38006165623664856, + "learning_rate": 4.196850393700788e-06, + "loss": 0.5223, + "step": 1939 + }, + { + "epoch": 0.16079569001243266, + "grad_norm": 0.26673129200935364, + "learning_rate": 4.196435971819312e-06, + "loss": 0.5154, + "step": 1940 + }, + { + "epoch": 0.1608785743887277, + "grad_norm": 0.29528021812438965, + "learning_rate": 4.196021549937837e-06, + "loss": 0.6031, + "step": 1941 + }, + { + "epoch": 0.1609614587650228, + "grad_norm": 0.26482680439949036, + "learning_rate": 4.195607128056361e-06, + "loss": 0.5449, + "step": 1942 + }, + { + "epoch": 0.16104434314131785, + "grad_norm": 0.281461238861084, + "learning_rate": 4.1951927061748866e-06, + "loss": 0.5778, + "step": 1943 + }, + { + "epoch": 0.16112722751761294, + "grad_norm": 0.27524223923683167, + "learning_rate": 4.194778284293411e-06, + "loss": 0.5449, + "step": 1944 + }, + { + "epoch": 0.161210111893908, + "grad_norm": 0.26290664076805115, + "learning_rate": 4.194363862411936e-06, + "loss": 0.5282, + "step": 1945 + }, + { + "epoch": 0.16129299627020308, + "grad_norm": 0.272779643535614, + "learning_rate": 4.193949440530461e-06, + "loss": 0.59, + "step": 1946 + }, + { + "epoch": 0.16137588064649813, + "grad_norm": 0.29584088921546936, + "learning_rate": 4.193535018648985e-06, + "loss": 0.5807, + "step": 1947 + }, + { + "epoch": 0.16145876502279322, + "grad_norm": 0.26675668358802795, + "learning_rate": 4.1931205967675096e-06, + "loss": 0.5721, + "step": 1948 + }, + { + "epoch": 0.16154164939908827, + "grad_norm": 0.2727329432964325, + "learning_rate": 4.192706174886034e-06, + "loss": 0.5725, + "step": 1949 + }, + { + "epoch": 0.16162453377538333, + "grad_norm": 0.3151703476905823, + "learning_rate": 4.1922917530045584e-06, + "loss": 0.5731, + "step": 1950 + }, + { + "epoch": 0.1617074181516784, + "grad_norm": 0.28236123919487, + "learning_rate": 4.191877331123084e-06, + "loss": 0.5491, + "step": 1951 + }, + { + "epoch": 0.16179030252797347, + "grad_norm": 0.2813953459262848, + "learning_rate": 4.191462909241608e-06, + "loss": 0.5474, + "step": 1952 + }, + { + "epoch": 0.16187318690426855, + "grad_norm": 0.24914345145225525, + "learning_rate": 4.1910484873601334e-06, + "loss": 0.5043, + "step": 1953 + }, + { + "epoch": 0.1619560712805636, + "grad_norm": 0.28688064217567444, + "learning_rate": 4.190634065478658e-06, + "loss": 0.5304, + "step": 1954 + }, + { + "epoch": 0.1620389556568587, + "grad_norm": 0.2956068515777588, + "learning_rate": 4.190219643597182e-06, + "loss": 0.5493, + "step": 1955 + }, + { + "epoch": 0.16212184003315375, + "grad_norm": 0.28021684288978577, + "learning_rate": 4.189805221715707e-06, + "loss": 0.5993, + "step": 1956 + }, + { + "epoch": 0.16220472440944883, + "grad_norm": 0.29198047518730164, + "learning_rate": 4.189390799834231e-06, + "loss": 0.5724, + "step": 1957 + }, + { + "epoch": 0.1622876087857439, + "grad_norm": 0.25886765122413635, + "learning_rate": 4.1889763779527564e-06, + "loss": 0.5284, + "step": 1958 + }, + { + "epoch": 0.16237049316203894, + "grad_norm": 0.2754187285900116, + "learning_rate": 4.188561956071281e-06, + "loss": 0.5566, + "step": 1959 + }, + { + "epoch": 0.16245337753833403, + "grad_norm": 0.25957244634628296, + "learning_rate": 4.188147534189805e-06, + "loss": 0.5545, + "step": 1960 + }, + { + "epoch": 0.16253626191462908, + "grad_norm": 0.26637139916419983, + "learning_rate": 4.1877331123083306e-06, + "loss": 0.5147, + "step": 1961 + }, + { + "epoch": 0.16261914629092417, + "grad_norm": 0.2973175048828125, + "learning_rate": 4.187318690426855e-06, + "loss": 0.5872, + "step": 1962 + }, + { + "epoch": 0.16270203066721922, + "grad_norm": 0.26458829641342163, + "learning_rate": 4.1869042685453794e-06, + "loss": 0.5358, + "step": 1963 + }, + { + "epoch": 0.1627849150435143, + "grad_norm": 0.2749732732772827, + "learning_rate": 4.186489846663904e-06, + "loss": 0.5461, + "step": 1964 + }, + { + "epoch": 0.16286779941980936, + "grad_norm": 0.25149208307266235, + "learning_rate": 4.186075424782428e-06, + "loss": 0.5198, + "step": 1965 + }, + { + "epoch": 0.16295068379610445, + "grad_norm": 0.2595387101173401, + "learning_rate": 4.185661002900954e-06, + "loss": 0.5717, + "step": 1966 + }, + { + "epoch": 0.1630335681723995, + "grad_norm": 0.2802722454071045, + "learning_rate": 4.185246581019478e-06, + "loss": 0.5489, + "step": 1967 + }, + { + "epoch": 0.16311645254869458, + "grad_norm": 0.2591615915298462, + "learning_rate": 4.1848321591380024e-06, + "loss": 0.5582, + "step": 1968 + }, + { + "epoch": 0.16319933692498964, + "grad_norm": 0.2729601562023163, + "learning_rate": 4.184417737256528e-06, + "loss": 0.5834, + "step": 1969 + }, + { + "epoch": 0.1632822213012847, + "grad_norm": 0.2672160565853119, + "learning_rate": 4.184003315375052e-06, + "loss": 0.5795, + "step": 1970 + }, + { + "epoch": 0.16336510567757978, + "grad_norm": 0.2539224326610565, + "learning_rate": 4.183588893493577e-06, + "loss": 0.5152, + "step": 1971 + }, + { + "epoch": 0.16344799005387484, + "grad_norm": 0.27869316935539246, + "learning_rate": 4.183174471612102e-06, + "loss": 0.5454, + "step": 1972 + }, + { + "epoch": 0.16353087443016992, + "grad_norm": 0.28222018480300903, + "learning_rate": 4.182760049730626e-06, + "loss": 0.5627, + "step": 1973 + }, + { + "epoch": 0.16361375880646498, + "grad_norm": 0.2610744833946228, + "learning_rate": 4.182345627849151e-06, + "loss": 0.533, + "step": 1974 + }, + { + "epoch": 0.16369664318276006, + "grad_norm": 0.24774634838104248, + "learning_rate": 4.181931205967675e-06, + "loss": 0.512, + "step": 1975 + }, + { + "epoch": 0.16377952755905512, + "grad_norm": 0.2638547420501709, + "learning_rate": 4.1815167840862004e-06, + "loss": 0.5283, + "step": 1976 + }, + { + "epoch": 0.1638624119353502, + "grad_norm": 0.2737526297569275, + "learning_rate": 4.181102362204725e-06, + "loss": 0.5342, + "step": 1977 + }, + { + "epoch": 0.16394529631164526, + "grad_norm": 0.2617841958999634, + "learning_rate": 4.180687940323249e-06, + "loss": 0.5303, + "step": 1978 + }, + { + "epoch": 0.1640281806879403, + "grad_norm": 0.29324397444725037, + "learning_rate": 4.180273518441775e-06, + "loss": 0.5435, + "step": 1979 + }, + { + "epoch": 0.1641110650642354, + "grad_norm": 0.3061814606189728, + "learning_rate": 4.179859096560299e-06, + "loss": 0.6302, + "step": 1980 + }, + { + "epoch": 0.16419394944053045, + "grad_norm": 0.2713508605957031, + "learning_rate": 4.1794446746788234e-06, + "loss": 0.582, + "step": 1981 + }, + { + "epoch": 0.16427683381682553, + "grad_norm": 0.2633136808872223, + "learning_rate": 4.179030252797348e-06, + "loss": 0.5775, + "step": 1982 + }, + { + "epoch": 0.1643597181931206, + "grad_norm": 0.2627008557319641, + "learning_rate": 4.178615830915872e-06, + "loss": 0.5673, + "step": 1983 + }, + { + "epoch": 0.16444260256941567, + "grad_norm": 0.24696899950504303, + "learning_rate": 4.178201409034398e-06, + "loss": 0.5216, + "step": 1984 + }, + { + "epoch": 0.16452548694571073, + "grad_norm": 0.2700790762901306, + "learning_rate": 4.177786987152922e-06, + "loss": 0.5188, + "step": 1985 + }, + { + "epoch": 0.16460837132200581, + "grad_norm": 0.2636084258556366, + "learning_rate": 4.1773725652714464e-06, + "loss": 0.566, + "step": 1986 + }, + { + "epoch": 0.16469125569830087, + "grad_norm": 0.2739083468914032, + "learning_rate": 4.176958143389972e-06, + "loss": 0.5184, + "step": 1987 + }, + { + "epoch": 0.16477414007459593, + "grad_norm": 0.2592485845088959, + "learning_rate": 4.176543721508496e-06, + "loss": 0.5164, + "step": 1988 + }, + { + "epoch": 0.164857024450891, + "grad_norm": 0.2962060868740082, + "learning_rate": 4.176129299627021e-06, + "loss": 0.595, + "step": 1989 + }, + { + "epoch": 0.16493990882718607, + "grad_norm": 0.2901153266429901, + "learning_rate": 4.175714877745545e-06, + "loss": 0.5573, + "step": 1990 + }, + { + "epoch": 0.16502279320348115, + "grad_norm": 0.26649436354637146, + "learning_rate": 4.1753004558640695e-06, + "loss": 0.5872, + "step": 1991 + }, + { + "epoch": 0.1651056775797762, + "grad_norm": 0.3866247236728668, + "learning_rate": 4.174886033982595e-06, + "loss": 0.5929, + "step": 1992 + }, + { + "epoch": 0.1651885619560713, + "grad_norm": 0.2563925087451935, + "learning_rate": 4.174471612101119e-06, + "loss": 0.5397, + "step": 1993 + }, + { + "epoch": 0.16527144633236635, + "grad_norm": 0.27170687913894653, + "learning_rate": 4.1740571902196444e-06, + "loss": 0.5267, + "step": 1994 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 0.27032408118247986, + "learning_rate": 4.173642768338169e-06, + "loss": 0.5433, + "step": 1995 + }, + { + "epoch": 0.16543721508495648, + "grad_norm": 0.27610039710998535, + "learning_rate": 4.173228346456693e-06, + "loss": 0.544, + "step": 1996 + }, + { + "epoch": 0.16552009946125154, + "grad_norm": 0.2911313772201538, + "learning_rate": 4.172813924575218e-06, + "loss": 0.5447, + "step": 1997 + }, + { + "epoch": 0.16560298383754662, + "grad_norm": 0.27439117431640625, + "learning_rate": 4.172399502693742e-06, + "loss": 0.5762, + "step": 1998 + }, + { + "epoch": 0.16568586821384168, + "grad_norm": 0.26667729020118713, + "learning_rate": 4.1719850808122675e-06, + "loss": 0.5019, + "step": 1999 + }, + { + "epoch": 0.16576875259013676, + "grad_norm": 0.2646658718585968, + "learning_rate": 4.171570658930792e-06, + "loss": 0.5309, + "step": 2000 + }, + { + "epoch": 0.16585163696643182, + "grad_norm": 0.2706732749938965, + "learning_rate": 4.171156237049316e-06, + "loss": 0.513, + "step": 2001 + }, + { + "epoch": 0.1659345213427269, + "grad_norm": 0.287985622882843, + "learning_rate": 4.170741815167842e-06, + "loss": 0.5965, + "step": 2002 + }, + { + "epoch": 0.16601740571902196, + "grad_norm": 0.2561929225921631, + "learning_rate": 4.170327393286366e-06, + "loss": 0.5255, + "step": 2003 + }, + { + "epoch": 0.16610029009531704, + "grad_norm": 0.28128287196159363, + "learning_rate": 4.1699129714048905e-06, + "loss": 0.5616, + "step": 2004 + }, + { + "epoch": 0.1661831744716121, + "grad_norm": 0.2639615535736084, + "learning_rate": 4.169498549523415e-06, + "loss": 0.563, + "step": 2005 + }, + { + "epoch": 0.16626605884790718, + "grad_norm": 0.25811880826950073, + "learning_rate": 4.169084127641939e-06, + "loss": 0.5276, + "step": 2006 + }, + { + "epoch": 0.16634894322420224, + "grad_norm": 0.23341991007328033, + "learning_rate": 4.168669705760465e-06, + "loss": 0.4812, + "step": 2007 + }, + { + "epoch": 0.1664318276004973, + "grad_norm": 0.27402937412261963, + "learning_rate": 4.168255283878989e-06, + "loss": 0.5443, + "step": 2008 + }, + { + "epoch": 0.16651471197679238, + "grad_norm": 0.3141966164112091, + "learning_rate": 4.1678408619975135e-06, + "loss": 0.6327, + "step": 2009 + }, + { + "epoch": 0.16659759635308743, + "grad_norm": 0.25224700570106506, + "learning_rate": 4.167426440116039e-06, + "loss": 0.5213, + "step": 2010 + }, + { + "epoch": 0.16668048072938252, + "grad_norm": 0.25797897577285767, + "learning_rate": 4.167012018234563e-06, + "loss": 0.5256, + "step": 2011 + }, + { + "epoch": 0.16676336510567757, + "grad_norm": 0.28318285942077637, + "learning_rate": 4.166597596353088e-06, + "loss": 0.5107, + "step": 2012 + }, + { + "epoch": 0.16684624948197266, + "grad_norm": 0.2970010042190552, + "learning_rate": 4.166183174471613e-06, + "loss": 0.5887, + "step": 2013 + }, + { + "epoch": 0.16692913385826771, + "grad_norm": 0.2610986530780792, + "learning_rate": 4.165768752590137e-06, + "loss": 0.5408, + "step": 2014 + }, + { + "epoch": 0.1670120182345628, + "grad_norm": 0.26555612683296204, + "learning_rate": 4.165354330708662e-06, + "loss": 0.5869, + "step": 2015 + }, + { + "epoch": 0.16709490261085785, + "grad_norm": 0.27892255783081055, + "learning_rate": 4.164939908827186e-06, + "loss": 0.5672, + "step": 2016 + }, + { + "epoch": 0.1671777869871529, + "grad_norm": 0.2851126790046692, + "learning_rate": 4.1645254869457115e-06, + "loss": 0.5936, + "step": 2017 + }, + { + "epoch": 0.167260671363448, + "grad_norm": 0.2855212390422821, + "learning_rate": 4.164111065064236e-06, + "loss": 0.5238, + "step": 2018 + }, + { + "epoch": 0.16734355573974305, + "grad_norm": 0.261750727891922, + "learning_rate": 4.16369664318276e-06, + "loss": 0.5256, + "step": 2019 + }, + { + "epoch": 0.16742644011603813, + "grad_norm": 0.25606632232666016, + "learning_rate": 4.163282221301286e-06, + "loss": 0.5391, + "step": 2020 + }, + { + "epoch": 0.1675093244923332, + "grad_norm": 0.26972681283950806, + "learning_rate": 4.16286779941981e-06, + "loss": 0.5956, + "step": 2021 + }, + { + "epoch": 0.16759220886862827, + "grad_norm": 0.27646562457084656, + "learning_rate": 4.1624533775383345e-06, + "loss": 0.4913, + "step": 2022 + }, + { + "epoch": 0.16767509324492333, + "grad_norm": 0.30021342635154724, + "learning_rate": 4.162038955656859e-06, + "loss": 0.5616, + "step": 2023 + }, + { + "epoch": 0.1677579776212184, + "grad_norm": 0.28466561436653137, + "learning_rate": 4.161624533775383e-06, + "loss": 0.5944, + "step": 2024 + }, + { + "epoch": 0.16784086199751347, + "grad_norm": 0.2597118318080902, + "learning_rate": 4.161210111893909e-06, + "loss": 0.5305, + "step": 2025 + }, + { + "epoch": 0.16792374637380852, + "grad_norm": 0.2827736735343933, + "learning_rate": 4.160795690012433e-06, + "loss": 0.5463, + "step": 2026 + }, + { + "epoch": 0.1680066307501036, + "grad_norm": 0.26398012042045593, + "learning_rate": 4.1603812681309575e-06, + "loss": 0.4966, + "step": 2027 + }, + { + "epoch": 0.16808951512639866, + "grad_norm": 0.2701379954814911, + "learning_rate": 4.159966846249483e-06, + "loss": 0.5551, + "step": 2028 + }, + { + "epoch": 0.16817239950269375, + "grad_norm": 0.2713738977909088, + "learning_rate": 4.159552424368007e-06, + "loss": 0.5659, + "step": 2029 + }, + { + "epoch": 0.1682552838789888, + "grad_norm": 0.30023106932640076, + "learning_rate": 4.159138002486532e-06, + "loss": 0.6233, + "step": 2030 + }, + { + "epoch": 0.1683381682552839, + "grad_norm": 0.2818370759487152, + "learning_rate": 4.158723580605056e-06, + "loss": 0.5396, + "step": 2031 + }, + { + "epoch": 0.16842105263157894, + "grad_norm": 0.2761830985546112, + "learning_rate": 4.1583091587235805e-06, + "loss": 0.5336, + "step": 2032 + }, + { + "epoch": 0.16850393700787403, + "grad_norm": 0.27792617678642273, + "learning_rate": 4.157894736842106e-06, + "loss": 0.5306, + "step": 2033 + }, + { + "epoch": 0.16858682138416908, + "grad_norm": 0.2841368019580841, + "learning_rate": 4.15748031496063e-06, + "loss": 0.5378, + "step": 2034 + }, + { + "epoch": 0.16866970576046414, + "grad_norm": 0.27338504791259766, + "learning_rate": 4.157065893079155e-06, + "loss": 0.5337, + "step": 2035 + }, + { + "epoch": 0.16875259013675922, + "grad_norm": 0.2565787434577942, + "learning_rate": 4.15665147119768e-06, + "loss": 0.5228, + "step": 2036 + }, + { + "epoch": 0.16883547451305428, + "grad_norm": 0.25613701343536377, + "learning_rate": 4.156237049316204e-06, + "loss": 0.5479, + "step": 2037 + }, + { + "epoch": 0.16891835888934936, + "grad_norm": 0.2483033686876297, + "learning_rate": 4.155822627434729e-06, + "loss": 0.5234, + "step": 2038 + }, + { + "epoch": 0.16900124326564442, + "grad_norm": 0.2922164499759674, + "learning_rate": 4.155408205553253e-06, + "loss": 0.5789, + "step": 2039 + }, + { + "epoch": 0.1690841276419395, + "grad_norm": 0.26827818155288696, + "learning_rate": 4.154993783671778e-06, + "loss": 0.5707, + "step": 2040 + }, + { + "epoch": 0.16916701201823456, + "grad_norm": 0.2582467198371887, + "learning_rate": 4.154579361790303e-06, + "loss": 0.5359, + "step": 2041 + }, + { + "epoch": 0.16924989639452964, + "grad_norm": 0.2667381763458252, + "learning_rate": 4.154164939908827e-06, + "loss": 0.5346, + "step": 2042 + }, + { + "epoch": 0.1693327807708247, + "grad_norm": 0.2630237638950348, + "learning_rate": 4.153750518027353e-06, + "loss": 0.5132, + "step": 2043 + }, + { + "epoch": 0.16941566514711978, + "grad_norm": 0.25836125016212463, + "learning_rate": 4.153336096145877e-06, + "loss": 0.5123, + "step": 2044 + }, + { + "epoch": 0.16949854952341484, + "grad_norm": 0.25621500611305237, + "learning_rate": 4.1529216742644015e-06, + "loss": 0.5529, + "step": 2045 + }, + { + "epoch": 0.1695814338997099, + "grad_norm": 0.26006752252578735, + "learning_rate": 4.152507252382926e-06, + "loss": 0.5185, + "step": 2046 + }, + { + "epoch": 0.16966431827600498, + "grad_norm": 0.24170981347560883, + "learning_rate": 4.15209283050145e-06, + "loss": 0.5018, + "step": 2047 + }, + { + "epoch": 0.16974720265230003, + "grad_norm": 0.2577148675918579, + "learning_rate": 4.151678408619976e-06, + "loss": 0.5269, + "step": 2048 + }, + { + "epoch": 0.16983008702859512, + "grad_norm": 0.27741295099258423, + "learning_rate": 4.1512639867385e-06, + "loss": 0.5568, + "step": 2049 + }, + { + "epoch": 0.16991297140489017, + "grad_norm": 0.25974008440971375, + "learning_rate": 4.1508495648570245e-06, + "loss": 0.5233, + "step": 2050 + }, + { + "epoch": 0.16999585578118526, + "grad_norm": 0.244883194565773, + "learning_rate": 4.15043514297555e-06, + "loss": 0.4319, + "step": 2051 + }, + { + "epoch": 0.1700787401574803, + "grad_norm": 0.2566312849521637, + "learning_rate": 4.150020721094074e-06, + "loss": 0.504, + "step": 2052 + }, + { + "epoch": 0.1701616245337754, + "grad_norm": 0.2895858883857727, + "learning_rate": 4.149606299212599e-06, + "loss": 0.577, + "step": 2053 + }, + { + "epoch": 0.17024450891007045, + "grad_norm": 0.27848902344703674, + "learning_rate": 4.149191877331124e-06, + "loss": 0.5744, + "step": 2054 + }, + { + "epoch": 0.1703273932863655, + "grad_norm": 0.2912577688694, + "learning_rate": 4.148777455449648e-06, + "loss": 0.5446, + "step": 2055 + }, + { + "epoch": 0.1704102776626606, + "grad_norm": 0.29133203625679016, + "learning_rate": 4.148363033568173e-06, + "loss": 0.5812, + "step": 2056 + }, + { + "epoch": 0.17049316203895565, + "grad_norm": 0.287628173828125, + "learning_rate": 4.147948611686697e-06, + "loss": 0.5241, + "step": 2057 + }, + { + "epoch": 0.17057604641525073, + "grad_norm": 0.2580616772174835, + "learning_rate": 4.147534189805222e-06, + "loss": 0.5413, + "step": 2058 + }, + { + "epoch": 0.1706589307915458, + "grad_norm": 0.27896368503570557, + "learning_rate": 4.147119767923747e-06, + "loss": 0.5516, + "step": 2059 + }, + { + "epoch": 0.17074181516784087, + "grad_norm": 0.2671978175640106, + "learning_rate": 4.146705346042271e-06, + "loss": 0.5819, + "step": 2060 + }, + { + "epoch": 0.17082469954413593, + "grad_norm": 0.26327651739120483, + "learning_rate": 4.146290924160797e-06, + "loss": 0.5833, + "step": 2061 + }, + { + "epoch": 0.170907583920431, + "grad_norm": 0.27495628595352173, + "learning_rate": 4.145876502279321e-06, + "loss": 0.5396, + "step": 2062 + }, + { + "epoch": 0.17099046829672607, + "grad_norm": 0.25468021631240845, + "learning_rate": 4.1454620803978455e-06, + "loss": 0.5532, + "step": 2063 + }, + { + "epoch": 0.17107335267302112, + "grad_norm": 0.2607797384262085, + "learning_rate": 4.14504765851637e-06, + "loss": 0.5515, + "step": 2064 + }, + { + "epoch": 0.1711562370493162, + "grad_norm": 0.2768995463848114, + "learning_rate": 4.144633236634894e-06, + "loss": 0.5735, + "step": 2065 + }, + { + "epoch": 0.17123912142561126, + "grad_norm": 0.2774958312511444, + "learning_rate": 4.14421881475342e-06, + "loss": 0.5708, + "step": 2066 + }, + { + "epoch": 0.17132200580190635, + "grad_norm": 0.2658000588417053, + "learning_rate": 4.143804392871944e-06, + "loss": 0.5533, + "step": 2067 + }, + { + "epoch": 0.1714048901782014, + "grad_norm": 0.25881674885749817, + "learning_rate": 4.1433899709904685e-06, + "loss": 0.5665, + "step": 2068 + }, + { + "epoch": 0.17148777455449649, + "grad_norm": 0.29636815190315247, + "learning_rate": 4.142975549108994e-06, + "loss": 0.6224, + "step": 2069 + }, + { + "epoch": 0.17157065893079154, + "grad_norm": 0.2652575373649597, + "learning_rate": 4.142561127227518e-06, + "loss": 0.5333, + "step": 2070 + }, + { + "epoch": 0.17165354330708663, + "grad_norm": 0.2575199604034424, + "learning_rate": 4.142146705346043e-06, + "loss": 0.5651, + "step": 2071 + }, + { + "epoch": 0.17173642768338168, + "grad_norm": 0.2405645251274109, + "learning_rate": 4.141732283464567e-06, + "loss": 0.5037, + "step": 2072 + }, + { + "epoch": 0.17181931205967674, + "grad_norm": 0.25403115153312683, + "learning_rate": 4.1413178615830915e-06, + "loss": 0.5363, + "step": 2073 + }, + { + "epoch": 0.17190219643597182, + "grad_norm": 0.289071261882782, + "learning_rate": 4.140903439701617e-06, + "loss": 0.516, + "step": 2074 + }, + { + "epoch": 0.17198508081226688, + "grad_norm": 0.28808948397636414, + "learning_rate": 4.140489017820141e-06, + "loss": 0.5945, + "step": 2075 + }, + { + "epoch": 0.17206796518856196, + "grad_norm": 0.27880623936653137, + "learning_rate": 4.140074595938666e-06, + "loss": 0.4766, + "step": 2076 + }, + { + "epoch": 0.17215084956485702, + "grad_norm": 0.29038065671920776, + "learning_rate": 4.139660174057191e-06, + "loss": 0.5793, + "step": 2077 + }, + { + "epoch": 0.1722337339411521, + "grad_norm": 0.2861940264701843, + "learning_rate": 4.139245752175715e-06, + "loss": 0.5411, + "step": 2078 + }, + { + "epoch": 0.17231661831744716, + "grad_norm": 0.2686365246772766, + "learning_rate": 4.13883133029424e-06, + "loss": 0.5408, + "step": 2079 + }, + { + "epoch": 0.17239950269374224, + "grad_norm": 0.28542107343673706, + "learning_rate": 4.138416908412764e-06, + "loss": 0.5753, + "step": 2080 + }, + { + "epoch": 0.1724823870700373, + "grad_norm": 0.2637152075767517, + "learning_rate": 4.138002486531289e-06, + "loss": 0.4889, + "step": 2081 + }, + { + "epoch": 0.17256527144633238, + "grad_norm": 0.27426877617836, + "learning_rate": 4.137588064649814e-06, + "loss": 0.5435, + "step": 2082 + }, + { + "epoch": 0.17264815582262744, + "grad_norm": 0.256077378988266, + "learning_rate": 4.137173642768338e-06, + "loss": 0.5468, + "step": 2083 + }, + { + "epoch": 0.1727310401989225, + "grad_norm": 0.27761921286582947, + "learning_rate": 4.136759220886864e-06, + "loss": 0.5876, + "step": 2084 + }, + { + "epoch": 0.17281392457521758, + "grad_norm": 0.27004480361938477, + "learning_rate": 4.136344799005388e-06, + "loss": 0.5739, + "step": 2085 + }, + { + "epoch": 0.17289680895151263, + "grad_norm": 0.2693946659564972, + "learning_rate": 4.1359303771239125e-06, + "loss": 0.5313, + "step": 2086 + }, + { + "epoch": 0.17297969332780772, + "grad_norm": 0.2871758043766022, + "learning_rate": 4.135515955242437e-06, + "loss": 0.6268, + "step": 2087 + }, + { + "epoch": 0.17306257770410277, + "grad_norm": 0.255222886800766, + "learning_rate": 4.135101533360961e-06, + "loss": 0.5755, + "step": 2088 + }, + { + "epoch": 0.17314546208039786, + "grad_norm": 0.2716340124607086, + "learning_rate": 4.134687111479486e-06, + "loss": 0.5097, + "step": 2089 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 0.2631749212741852, + "learning_rate": 4.134272689598011e-06, + "loss": 0.538, + "step": 2090 + }, + { + "epoch": 0.173311230832988, + "grad_norm": 0.27473685145378113, + "learning_rate": 4.1338582677165355e-06, + "loss": 0.5376, + "step": 2091 + }, + { + "epoch": 0.17339411520928305, + "grad_norm": 0.2618618905544281, + "learning_rate": 4.133443845835061e-06, + "loss": 0.5509, + "step": 2092 + }, + { + "epoch": 0.1734769995855781, + "grad_norm": 0.27917131781578064, + "learning_rate": 4.133029423953585e-06, + "loss": 0.5337, + "step": 2093 + }, + { + "epoch": 0.1735598839618732, + "grad_norm": 0.2767510712146759, + "learning_rate": 4.13261500207211e-06, + "loss": 0.5233, + "step": 2094 + }, + { + "epoch": 0.17364276833816825, + "grad_norm": 0.2460285723209381, + "learning_rate": 4.132200580190635e-06, + "loss": 0.5148, + "step": 2095 + }, + { + "epoch": 0.17372565271446333, + "grad_norm": 0.26370516419410706, + "learning_rate": 4.131786158309159e-06, + "loss": 0.5278, + "step": 2096 + }, + { + "epoch": 0.17380853709075839, + "grad_norm": 0.27561622858047485, + "learning_rate": 4.131371736427684e-06, + "loss": 0.6045, + "step": 2097 + }, + { + "epoch": 0.17389142146705347, + "grad_norm": 0.28658270835876465, + "learning_rate": 4.130957314546208e-06, + "loss": 0.5492, + "step": 2098 + }, + { + "epoch": 0.17397430584334853, + "grad_norm": 0.26457318663597107, + "learning_rate": 4.130542892664733e-06, + "loss": 0.5393, + "step": 2099 + }, + { + "epoch": 0.1740571902196436, + "grad_norm": 0.31130775809288025, + "learning_rate": 4.130128470783258e-06, + "loss": 0.6139, + "step": 2100 + }, + { + "epoch": 0.17414007459593867, + "grad_norm": 0.2562922239303589, + "learning_rate": 4.129714048901782e-06, + "loss": 0.5402, + "step": 2101 + }, + { + "epoch": 0.17422295897223372, + "grad_norm": 0.27774399518966675, + "learning_rate": 4.129299627020307e-06, + "loss": 0.5522, + "step": 2102 + }, + { + "epoch": 0.1743058433485288, + "grad_norm": 0.2579796016216278, + "learning_rate": 4.128885205138832e-06, + "loss": 0.5315, + "step": 2103 + }, + { + "epoch": 0.17438872772482386, + "grad_norm": 0.2795619070529938, + "learning_rate": 4.1284707832573565e-06, + "loss": 0.5313, + "step": 2104 + }, + { + "epoch": 0.17447161210111894, + "grad_norm": 0.2606847882270813, + "learning_rate": 4.128056361375881e-06, + "loss": 0.5487, + "step": 2105 + }, + { + "epoch": 0.174554496477414, + "grad_norm": 0.29977601766586304, + "learning_rate": 4.127641939494405e-06, + "loss": 0.5962, + "step": 2106 + }, + { + "epoch": 0.17463738085370908, + "grad_norm": 0.26048222184181213, + "learning_rate": 4.12722751761293e-06, + "loss": 0.5325, + "step": 2107 + }, + { + "epoch": 0.17472026523000414, + "grad_norm": 0.2763049006462097, + "learning_rate": 4.126813095731455e-06, + "loss": 0.5842, + "step": 2108 + }, + { + "epoch": 0.17480314960629922, + "grad_norm": 0.2841283977031708, + "learning_rate": 4.1263986738499795e-06, + "loss": 0.5613, + "step": 2109 + }, + { + "epoch": 0.17488603398259428, + "grad_norm": 0.27078044414520264, + "learning_rate": 4.125984251968505e-06, + "loss": 0.5637, + "step": 2110 + }, + { + "epoch": 0.17496891835888934, + "grad_norm": 0.2822072505950928, + "learning_rate": 4.125569830087029e-06, + "loss": 0.5712, + "step": 2111 + }, + { + "epoch": 0.17505180273518442, + "grad_norm": 0.2621695101261139, + "learning_rate": 4.125155408205554e-06, + "loss": 0.5319, + "step": 2112 + }, + { + "epoch": 0.17513468711147948, + "grad_norm": 0.2753240466117859, + "learning_rate": 4.124740986324078e-06, + "loss": 0.5933, + "step": 2113 + }, + { + "epoch": 0.17521757148777456, + "grad_norm": 0.2776031792163849, + "learning_rate": 4.1243265644426025e-06, + "loss": 0.5778, + "step": 2114 + }, + { + "epoch": 0.17530045586406962, + "grad_norm": 0.2796436548233032, + "learning_rate": 4.123912142561128e-06, + "loss": 0.5498, + "step": 2115 + }, + { + "epoch": 0.1753833402403647, + "grad_norm": 0.28615719079971313, + "learning_rate": 4.123497720679652e-06, + "loss": 0.5723, + "step": 2116 + }, + { + "epoch": 0.17546622461665976, + "grad_norm": 0.263494074344635, + "learning_rate": 4.123083298798177e-06, + "loss": 0.5585, + "step": 2117 + }, + { + "epoch": 0.17554910899295484, + "grad_norm": 0.2780320644378662, + "learning_rate": 4.122668876916702e-06, + "loss": 0.5653, + "step": 2118 + }, + { + "epoch": 0.1756319933692499, + "grad_norm": 0.2642548978328705, + "learning_rate": 4.122254455035226e-06, + "loss": 0.5533, + "step": 2119 + }, + { + "epoch": 0.17571487774554498, + "grad_norm": 0.3108097016811371, + "learning_rate": 4.121840033153751e-06, + "loss": 0.6006, + "step": 2120 + }, + { + "epoch": 0.17579776212184003, + "grad_norm": 0.2683505117893219, + "learning_rate": 4.121425611272275e-06, + "loss": 0.5222, + "step": 2121 + }, + { + "epoch": 0.1758806464981351, + "grad_norm": 0.27472734451293945, + "learning_rate": 4.1210111893908e-06, + "loss": 0.5554, + "step": 2122 + }, + { + "epoch": 0.17596353087443017, + "grad_norm": 0.27349042892456055, + "learning_rate": 4.120596767509325e-06, + "loss": 0.5861, + "step": 2123 + }, + { + "epoch": 0.17604641525072523, + "grad_norm": 0.2763645648956299, + "learning_rate": 4.120182345627849e-06, + "loss": 0.5462, + "step": 2124 + }, + { + "epoch": 0.1761292996270203, + "grad_norm": 0.27528637647628784, + "learning_rate": 4.119767923746374e-06, + "loss": 0.5799, + "step": 2125 + }, + { + "epoch": 0.17621218400331537, + "grad_norm": 0.24261344969272614, + "learning_rate": 4.119353501864899e-06, + "loss": 0.5121, + "step": 2126 + }, + { + "epoch": 0.17629506837961045, + "grad_norm": 0.2595287263393402, + "learning_rate": 4.1189390799834235e-06, + "loss": 0.5108, + "step": 2127 + }, + { + "epoch": 0.1763779527559055, + "grad_norm": 0.2718043327331543, + "learning_rate": 4.118524658101948e-06, + "loss": 0.5536, + "step": 2128 + }, + { + "epoch": 0.1764608371322006, + "grad_norm": 0.25715652108192444, + "learning_rate": 4.118110236220472e-06, + "loss": 0.5154, + "step": 2129 + }, + { + "epoch": 0.17654372150849565, + "grad_norm": 0.257711797952652, + "learning_rate": 4.117695814338997e-06, + "loss": 0.5492, + "step": 2130 + }, + { + "epoch": 0.1766266058847907, + "grad_norm": 0.2701655924320221, + "learning_rate": 4.117281392457522e-06, + "loss": 0.569, + "step": 2131 + }, + { + "epoch": 0.1767094902610858, + "grad_norm": 0.2610805332660675, + "learning_rate": 4.1168669705760465e-06, + "loss": 0.5651, + "step": 2132 + }, + { + "epoch": 0.17679237463738084, + "grad_norm": 0.34755533933639526, + "learning_rate": 4.116452548694572e-06, + "loss": 0.4984, + "step": 2133 + }, + { + "epoch": 0.17687525901367593, + "grad_norm": 0.27667728066444397, + "learning_rate": 4.116038126813096e-06, + "loss": 0.5796, + "step": 2134 + }, + { + "epoch": 0.17695814338997098, + "grad_norm": 0.3115779161453247, + "learning_rate": 4.115623704931621e-06, + "loss": 0.6004, + "step": 2135 + }, + { + "epoch": 0.17704102776626607, + "grad_norm": 0.2547604441642761, + "learning_rate": 4.115209283050146e-06, + "loss": 0.5323, + "step": 2136 + }, + { + "epoch": 0.17712391214256112, + "grad_norm": 0.2667185068130493, + "learning_rate": 4.11479486116867e-06, + "loss": 0.577, + "step": 2137 + }, + { + "epoch": 0.1772067965188562, + "grad_norm": 0.26088082790374756, + "learning_rate": 4.114380439287195e-06, + "loss": 0.538, + "step": 2138 + }, + { + "epoch": 0.17728968089515126, + "grad_norm": 0.2772800326347351, + "learning_rate": 4.113966017405719e-06, + "loss": 0.499, + "step": 2139 + }, + { + "epoch": 0.17737256527144632, + "grad_norm": 0.27790775895118713, + "learning_rate": 4.113551595524244e-06, + "loss": 0.5523, + "step": 2140 + }, + { + "epoch": 0.1774554496477414, + "grad_norm": 0.2982091009616852, + "learning_rate": 4.113137173642769e-06, + "loss": 0.5908, + "step": 2141 + }, + { + "epoch": 0.17753833402403646, + "grad_norm": 0.27961692214012146, + "learning_rate": 4.112722751761293e-06, + "loss": 0.5818, + "step": 2142 + }, + { + "epoch": 0.17762121840033154, + "grad_norm": 0.2596482038497925, + "learning_rate": 4.112308329879818e-06, + "loss": 0.5573, + "step": 2143 + }, + { + "epoch": 0.1777041027766266, + "grad_norm": 0.2611553370952606, + "learning_rate": 4.111893907998343e-06, + "loss": 0.4959, + "step": 2144 + }, + { + "epoch": 0.17778698715292168, + "grad_norm": 0.2772889733314514, + "learning_rate": 4.1114794861168675e-06, + "loss": 0.5628, + "step": 2145 + }, + { + "epoch": 0.17786987152921674, + "grad_norm": 0.26048675179481506, + "learning_rate": 4.111065064235392e-06, + "loss": 0.5213, + "step": 2146 + }, + { + "epoch": 0.17795275590551182, + "grad_norm": 0.2899695634841919, + "learning_rate": 4.110650642353916e-06, + "loss": 0.539, + "step": 2147 + }, + { + "epoch": 0.17803564028180688, + "grad_norm": 0.2832890450954437, + "learning_rate": 4.110236220472441e-06, + "loss": 0.5554, + "step": 2148 + }, + { + "epoch": 0.17811852465810193, + "grad_norm": 0.29113590717315674, + "learning_rate": 4.109821798590966e-06, + "loss": 0.5748, + "step": 2149 + }, + { + "epoch": 0.17820140903439702, + "grad_norm": 0.2547609508037567, + "learning_rate": 4.1094073767094905e-06, + "loss": 0.5589, + "step": 2150 + }, + { + "epoch": 0.17828429341069207, + "grad_norm": 0.2580583691596985, + "learning_rate": 4.108992954828016e-06, + "loss": 0.5318, + "step": 2151 + }, + { + "epoch": 0.17836717778698716, + "grad_norm": 0.2695213258266449, + "learning_rate": 4.10857853294654e-06, + "loss": 0.5704, + "step": 2152 + }, + { + "epoch": 0.1784500621632822, + "grad_norm": 0.2627929151058197, + "learning_rate": 4.108164111065065e-06, + "loss": 0.5038, + "step": 2153 + }, + { + "epoch": 0.1785329465395773, + "grad_norm": 0.2896193861961365, + "learning_rate": 4.107749689183589e-06, + "loss": 0.5901, + "step": 2154 + }, + { + "epoch": 0.17861583091587235, + "grad_norm": 0.27312129735946655, + "learning_rate": 4.1073352673021135e-06, + "loss": 0.5319, + "step": 2155 + }, + { + "epoch": 0.17869871529216744, + "grad_norm": 0.27774879336357117, + "learning_rate": 4.106920845420638e-06, + "loss": 0.552, + "step": 2156 + }, + { + "epoch": 0.1787815996684625, + "grad_norm": 0.254697322845459, + "learning_rate": 4.106506423539163e-06, + "loss": 0.5322, + "step": 2157 + }, + { + "epoch": 0.17886448404475758, + "grad_norm": 0.2580520808696747, + "learning_rate": 4.106092001657688e-06, + "loss": 0.5367, + "step": 2158 + }, + { + "epoch": 0.17894736842105263, + "grad_norm": 0.27399492263793945, + "learning_rate": 4.105677579776213e-06, + "loss": 0.5432, + "step": 2159 + }, + { + "epoch": 0.1790302527973477, + "grad_norm": 0.27739670872688293, + "learning_rate": 4.105263157894737e-06, + "loss": 0.5735, + "step": 2160 + }, + { + "epoch": 0.17911313717364277, + "grad_norm": 0.2774333953857422, + "learning_rate": 4.104848736013262e-06, + "loss": 0.5729, + "step": 2161 + }, + { + "epoch": 0.17919602154993783, + "grad_norm": 0.25231924653053284, + "learning_rate": 4.104434314131786e-06, + "loss": 0.5001, + "step": 2162 + }, + { + "epoch": 0.1792789059262329, + "grad_norm": 0.28005942702293396, + "learning_rate": 4.104019892250311e-06, + "loss": 0.5558, + "step": 2163 + }, + { + "epoch": 0.17936179030252797, + "grad_norm": 0.2781507670879364, + "learning_rate": 4.103605470368836e-06, + "loss": 0.5273, + "step": 2164 + }, + { + "epoch": 0.17944467467882305, + "grad_norm": 0.28107398748397827, + "learning_rate": 4.10319104848736e-06, + "loss": 0.5503, + "step": 2165 + }, + { + "epoch": 0.1795275590551181, + "grad_norm": 0.26558297872543335, + "learning_rate": 4.102776626605885e-06, + "loss": 0.5446, + "step": 2166 + }, + { + "epoch": 0.1796104434314132, + "grad_norm": 0.259957879781723, + "learning_rate": 4.10236220472441e-06, + "loss": 0.5889, + "step": 2167 + }, + { + "epoch": 0.17969332780770825, + "grad_norm": 0.23630711436271667, + "learning_rate": 4.1019477828429345e-06, + "loss": 0.4979, + "step": 2168 + }, + { + "epoch": 0.1797762121840033, + "grad_norm": 0.25615254044532776, + "learning_rate": 4.101533360961459e-06, + "loss": 0.5456, + "step": 2169 + }, + { + "epoch": 0.1798590965602984, + "grad_norm": 0.28263112902641296, + "learning_rate": 4.101118939079983e-06, + "loss": 0.5484, + "step": 2170 + }, + { + "epoch": 0.17994198093659344, + "grad_norm": 0.25831347703933716, + "learning_rate": 4.100704517198509e-06, + "loss": 0.5478, + "step": 2171 + }, + { + "epoch": 0.18002486531288853, + "grad_norm": 0.28031378984451294, + "learning_rate": 4.100290095317033e-06, + "loss": 0.5825, + "step": 2172 + }, + { + "epoch": 0.18010774968918358, + "grad_norm": 0.2687190771102905, + "learning_rate": 4.0998756734355575e-06, + "loss": 0.5729, + "step": 2173 + }, + { + "epoch": 0.18019063406547867, + "grad_norm": 0.27235525846481323, + "learning_rate": 4.099461251554082e-06, + "loss": 0.5215, + "step": 2174 + }, + { + "epoch": 0.18027351844177372, + "grad_norm": 0.26538413763046265, + "learning_rate": 4.099046829672607e-06, + "loss": 0.5311, + "step": 2175 + }, + { + "epoch": 0.1803564028180688, + "grad_norm": 0.27148905396461487, + "learning_rate": 4.098632407791132e-06, + "loss": 0.543, + "step": 2176 + }, + { + "epoch": 0.18043928719436386, + "grad_norm": 0.27654844522476196, + "learning_rate": 4.098217985909657e-06, + "loss": 0.5453, + "step": 2177 + }, + { + "epoch": 0.18052217157065892, + "grad_norm": 0.2879367768764496, + "learning_rate": 4.097803564028181e-06, + "loss": 0.5721, + "step": 2178 + }, + { + "epoch": 0.180605055946954, + "grad_norm": 0.28972557187080383, + "learning_rate": 4.097389142146706e-06, + "loss": 0.5688, + "step": 2179 + }, + { + "epoch": 0.18068794032324906, + "grad_norm": 0.27814891934394836, + "learning_rate": 4.09697472026523e-06, + "loss": 0.5297, + "step": 2180 + }, + { + "epoch": 0.18077082469954414, + "grad_norm": 0.27545663714408875, + "learning_rate": 4.096560298383755e-06, + "loss": 0.5526, + "step": 2181 + }, + { + "epoch": 0.1808537090758392, + "grad_norm": 0.26810920238494873, + "learning_rate": 4.09614587650228e-06, + "loss": 0.5388, + "step": 2182 + }, + { + "epoch": 0.18093659345213428, + "grad_norm": 0.25870051980018616, + "learning_rate": 4.095731454620804e-06, + "loss": 0.5109, + "step": 2183 + }, + { + "epoch": 0.18101947782842934, + "grad_norm": 0.2692229747772217, + "learning_rate": 4.095317032739329e-06, + "loss": 0.5517, + "step": 2184 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 0.25549760460853577, + "learning_rate": 4.094902610857854e-06, + "loss": 0.5566, + "step": 2185 + }, + { + "epoch": 0.18118524658101948, + "grad_norm": 0.2634415924549103, + "learning_rate": 4.0944881889763785e-06, + "loss": 0.5406, + "step": 2186 + }, + { + "epoch": 0.18126813095731453, + "grad_norm": 0.2421761304140091, + "learning_rate": 4.094073767094903e-06, + "loss": 0.5164, + "step": 2187 + }, + { + "epoch": 0.18135101533360962, + "grad_norm": 0.26543185114860535, + "learning_rate": 4.093659345213427e-06, + "loss": 0.5303, + "step": 2188 + }, + { + "epoch": 0.18143389970990467, + "grad_norm": 0.269571989774704, + "learning_rate": 4.093244923331952e-06, + "loss": 0.5431, + "step": 2189 + }, + { + "epoch": 0.18151678408619976, + "grad_norm": 0.25305432081222534, + "learning_rate": 4.092830501450477e-06, + "loss": 0.5063, + "step": 2190 + }, + { + "epoch": 0.1815996684624948, + "grad_norm": 0.28362029790878296, + "learning_rate": 4.0924160795690016e-06, + "loss": 0.5604, + "step": 2191 + }, + { + "epoch": 0.1816825528387899, + "grad_norm": 0.2708631157875061, + "learning_rate": 4.092001657687526e-06, + "loss": 0.555, + "step": 2192 + }, + { + "epoch": 0.18176543721508495, + "grad_norm": 0.27427443861961365, + "learning_rate": 4.091587235806051e-06, + "loss": 0.6011, + "step": 2193 + }, + { + "epoch": 0.18184832159138004, + "grad_norm": 0.3002326488494873, + "learning_rate": 4.091172813924576e-06, + "loss": 0.5869, + "step": 2194 + }, + { + "epoch": 0.1819312059676751, + "grad_norm": 0.265473872423172, + "learning_rate": 4.0907583920431e-06, + "loss": 0.5184, + "step": 2195 + }, + { + "epoch": 0.18201409034397018, + "grad_norm": 0.2841622531414032, + "learning_rate": 4.0903439701616246e-06, + "loss": 0.5577, + "step": 2196 + }, + { + "epoch": 0.18209697472026523, + "grad_norm": 0.28869691491127014, + "learning_rate": 4.089929548280149e-06, + "loss": 0.5692, + "step": 2197 + }, + { + "epoch": 0.1821798590965603, + "grad_norm": 0.28688114881515503, + "learning_rate": 4.089515126398674e-06, + "loss": 0.556, + "step": 2198 + }, + { + "epoch": 0.18226274347285537, + "grad_norm": 0.27100831270217896, + "learning_rate": 4.089100704517199e-06, + "loss": 0.5525, + "step": 2199 + }, + { + "epoch": 0.18234562784915043, + "grad_norm": 0.27683451771736145, + "learning_rate": 4.088686282635724e-06, + "loss": 0.5406, + "step": 2200 + }, + { + "epoch": 0.1824285122254455, + "grad_norm": 0.27016952633857727, + "learning_rate": 4.088271860754248e-06, + "loss": 0.5976, + "step": 2201 + }, + { + "epoch": 0.18251139660174057, + "grad_norm": 0.273169606924057, + "learning_rate": 4.087857438872773e-06, + "loss": 0.5472, + "step": 2202 + }, + { + "epoch": 0.18259428097803565, + "grad_norm": 0.2897804081439972, + "learning_rate": 4.087443016991297e-06, + "loss": 0.561, + "step": 2203 + }, + { + "epoch": 0.1826771653543307, + "grad_norm": 0.29060930013656616, + "learning_rate": 4.087028595109822e-06, + "loss": 0.6098, + "step": 2204 + }, + { + "epoch": 0.1827600497306258, + "grad_norm": 0.28506648540496826, + "learning_rate": 4.086614173228346e-06, + "loss": 0.6107, + "step": 2205 + }, + { + "epoch": 0.18284293410692085, + "grad_norm": 0.2679462432861328, + "learning_rate": 4.086199751346871e-06, + "loss": 0.552, + "step": 2206 + }, + { + "epoch": 0.1829258184832159, + "grad_norm": 0.25959378480911255, + "learning_rate": 4.085785329465396e-06, + "loss": 0.5417, + "step": 2207 + }, + { + "epoch": 0.18300870285951099, + "grad_norm": 0.29177919030189514, + "learning_rate": 4.085370907583921e-06, + "loss": 0.534, + "step": 2208 + }, + { + "epoch": 0.18309158723580604, + "grad_norm": 0.3079068958759308, + "learning_rate": 4.0849564857024456e-06, + "loss": 0.5635, + "step": 2209 + }, + { + "epoch": 0.18317447161210113, + "grad_norm": 0.2770945727825165, + "learning_rate": 4.08454206382097e-06, + "loss": 0.5342, + "step": 2210 + }, + { + "epoch": 0.18325735598839618, + "grad_norm": 0.26037099957466125, + "learning_rate": 4.0841276419394944e-06, + "loss": 0.5382, + "step": 2211 + }, + { + "epoch": 0.18334024036469126, + "grad_norm": 0.3163505494594574, + "learning_rate": 4.08371322005802e-06, + "loss": 0.621, + "step": 2212 + }, + { + "epoch": 0.18342312474098632, + "grad_norm": 0.250284343957901, + "learning_rate": 4.083298798176544e-06, + "loss": 0.5154, + "step": 2213 + }, + { + "epoch": 0.1835060091172814, + "grad_norm": 0.30347418785095215, + "learning_rate": 4.0828843762950686e-06, + "loss": 0.6345, + "step": 2214 + }, + { + "epoch": 0.18358889349357646, + "grad_norm": 0.2694132626056671, + "learning_rate": 4.082469954413593e-06, + "loss": 0.5541, + "step": 2215 + }, + { + "epoch": 0.18367177786987152, + "grad_norm": 0.2756105363368988, + "learning_rate": 4.082055532532118e-06, + "loss": 0.5704, + "step": 2216 + }, + { + "epoch": 0.1837546622461666, + "grad_norm": 0.26654288172721863, + "learning_rate": 4.081641110650643e-06, + "loss": 0.5602, + "step": 2217 + }, + { + "epoch": 0.18383754662246166, + "grad_norm": 0.28528207540512085, + "learning_rate": 4.081226688769168e-06, + "loss": 0.532, + "step": 2218 + }, + { + "epoch": 0.18392043099875674, + "grad_norm": 0.29187774658203125, + "learning_rate": 4.080812266887692e-06, + "loss": 0.5424, + "step": 2219 + }, + { + "epoch": 0.1840033153750518, + "grad_norm": 0.27899590134620667, + "learning_rate": 4.080397845006217e-06, + "loss": 0.5355, + "step": 2220 + }, + { + "epoch": 0.18408619975134688, + "grad_norm": 0.2719401717185974, + "learning_rate": 4.079983423124741e-06, + "loss": 0.5447, + "step": 2221 + }, + { + "epoch": 0.18416908412764194, + "grad_norm": 0.2626775801181793, + "learning_rate": 4.079569001243266e-06, + "loss": 0.5646, + "step": 2222 + }, + { + "epoch": 0.18425196850393702, + "grad_norm": 0.31802812218666077, + "learning_rate": 4.07915457936179e-06, + "loss": 0.562, + "step": 2223 + }, + { + "epoch": 0.18433485288023208, + "grad_norm": 0.24221940338611603, + "learning_rate": 4.0787401574803154e-06, + "loss": 0.538, + "step": 2224 + }, + { + "epoch": 0.18441773725652713, + "grad_norm": 0.2976449131965637, + "learning_rate": 4.07832573559884e-06, + "loss": 0.5457, + "step": 2225 + }, + { + "epoch": 0.18450062163282221, + "grad_norm": 0.2721804678440094, + "learning_rate": 4.077911313717365e-06, + "loss": 0.5583, + "step": 2226 + }, + { + "epoch": 0.18458350600911727, + "grad_norm": 0.25466489791870117, + "learning_rate": 4.0774968918358896e-06, + "loss": 0.5242, + "step": 2227 + }, + { + "epoch": 0.18466639038541235, + "grad_norm": 0.29417937994003296, + "learning_rate": 4.077082469954414e-06, + "loss": 0.5726, + "step": 2228 + }, + { + "epoch": 0.1847492747617074, + "grad_norm": 0.24328860640525818, + "learning_rate": 4.0766680480729384e-06, + "loss": 0.544, + "step": 2229 + }, + { + "epoch": 0.1848321591380025, + "grad_norm": 0.2635261118412018, + "learning_rate": 4.076253626191463e-06, + "loss": 0.5518, + "step": 2230 + }, + { + "epoch": 0.18491504351429755, + "grad_norm": 0.27303656935691833, + "learning_rate": 4.075839204309988e-06, + "loss": 0.5072, + "step": 2231 + }, + { + "epoch": 0.18499792789059263, + "grad_norm": 0.2513248920440674, + "learning_rate": 4.0754247824285126e-06, + "loss": 0.5298, + "step": 2232 + }, + { + "epoch": 0.1850808122668877, + "grad_norm": 0.2624649703502655, + "learning_rate": 4.075010360547037e-06, + "loss": 0.5611, + "step": 2233 + }, + { + "epoch": 0.18516369664318277, + "grad_norm": 0.2610453963279724, + "learning_rate": 4.074595938665562e-06, + "loss": 0.5131, + "step": 2234 + }, + { + "epoch": 0.18524658101947783, + "grad_norm": 0.2756372392177582, + "learning_rate": 4.074181516784087e-06, + "loss": 0.55, + "step": 2235 + }, + { + "epoch": 0.18532946539577289, + "grad_norm": 0.26098373532295227, + "learning_rate": 4.073767094902611e-06, + "loss": 0.5565, + "step": 2236 + }, + { + "epoch": 0.18541234977206797, + "grad_norm": 0.2741055488586426, + "learning_rate": 4.073352673021136e-06, + "loss": 0.5652, + "step": 2237 + }, + { + "epoch": 0.18549523414836303, + "grad_norm": 0.28589749336242676, + "learning_rate": 4.07293825113966e-06, + "loss": 0.541, + "step": 2238 + }, + { + "epoch": 0.1855781185246581, + "grad_norm": 0.6147609949111938, + "learning_rate": 4.072523829258185e-06, + "loss": 0.5618, + "step": 2239 + }, + { + "epoch": 0.18566100290095316, + "grad_norm": 0.2747034728527069, + "learning_rate": 4.07210940737671e-06, + "loss": 0.5562, + "step": 2240 + }, + { + "epoch": 0.18574388727724825, + "grad_norm": 0.26667550206184387, + "learning_rate": 4.071694985495234e-06, + "loss": 0.5492, + "step": 2241 + }, + { + "epoch": 0.1858267716535433, + "grad_norm": 0.26056209206581116, + "learning_rate": 4.0712805636137594e-06, + "loss": 0.5421, + "step": 2242 + }, + { + "epoch": 0.1859096560298384, + "grad_norm": 0.27092719078063965, + "learning_rate": 4.070866141732284e-06, + "loss": 0.5403, + "step": 2243 + }, + { + "epoch": 0.18599254040613344, + "grad_norm": 0.26334983110427856, + "learning_rate": 4.070451719850808e-06, + "loss": 0.5152, + "step": 2244 + }, + { + "epoch": 0.1860754247824285, + "grad_norm": 0.28946277499198914, + "learning_rate": 4.070037297969333e-06, + "loss": 0.5859, + "step": 2245 + }, + { + "epoch": 0.18615830915872358, + "grad_norm": 0.29279983043670654, + "learning_rate": 4.069622876087857e-06, + "loss": 0.5834, + "step": 2246 + }, + { + "epoch": 0.18624119353501864, + "grad_norm": 0.25063827633857727, + "learning_rate": 4.0692084542063824e-06, + "loss": 0.5644, + "step": 2247 + }, + { + "epoch": 0.18632407791131372, + "grad_norm": 0.26608169078826904, + "learning_rate": 4.068794032324907e-06, + "loss": 0.5724, + "step": 2248 + }, + { + "epoch": 0.18640696228760878, + "grad_norm": 0.26576173305511475, + "learning_rate": 4.068379610443432e-06, + "loss": 0.5728, + "step": 2249 + }, + { + "epoch": 0.18648984666390386, + "grad_norm": 0.2720705568790436, + "learning_rate": 4.067965188561957e-06, + "loss": 0.5674, + "step": 2250 + }, + { + "epoch": 0.18657273104019892, + "grad_norm": 0.26050689816474915, + "learning_rate": 4.067550766680481e-06, + "loss": 0.5316, + "step": 2251 + }, + { + "epoch": 0.186655615416494, + "grad_norm": 0.2644106447696686, + "learning_rate": 4.067136344799006e-06, + "loss": 0.5886, + "step": 2252 + }, + { + "epoch": 0.18673849979278906, + "grad_norm": 0.25253725051879883, + "learning_rate": 4.066721922917531e-06, + "loss": 0.5372, + "step": 2253 + }, + { + "epoch": 0.18682138416908411, + "grad_norm": 0.2670529782772064, + "learning_rate": 4.066307501036055e-06, + "loss": 0.531, + "step": 2254 + }, + { + "epoch": 0.1869042685453792, + "grad_norm": 0.25000160932540894, + "learning_rate": 4.06589307915458e-06, + "loss": 0.5405, + "step": 2255 + }, + { + "epoch": 0.18698715292167425, + "grad_norm": 0.2665954828262329, + "learning_rate": 4.065478657273104e-06, + "loss": 0.5591, + "step": 2256 + }, + { + "epoch": 0.18707003729796934, + "grad_norm": 0.26462697982788086, + "learning_rate": 4.065064235391629e-06, + "loss": 0.5807, + "step": 2257 + }, + { + "epoch": 0.1871529216742644, + "grad_norm": 0.26314330101013184, + "learning_rate": 4.064649813510154e-06, + "loss": 0.5289, + "step": 2258 + }, + { + "epoch": 0.18723580605055948, + "grad_norm": 0.27945247292518616, + "learning_rate": 4.064235391628678e-06, + "loss": 0.5699, + "step": 2259 + }, + { + "epoch": 0.18731869042685453, + "grad_norm": 0.2667839527130127, + "learning_rate": 4.0638209697472034e-06, + "loss": 0.5214, + "step": 2260 + }, + { + "epoch": 0.18740157480314962, + "grad_norm": 0.25985005497932434, + "learning_rate": 4.063406547865728e-06, + "loss": 0.5333, + "step": 2261 + }, + { + "epoch": 0.18748445917944467, + "grad_norm": 0.29666680097579956, + "learning_rate": 4.062992125984252e-06, + "loss": 0.5964, + "step": 2262 + }, + { + "epoch": 0.18756734355573973, + "grad_norm": 0.28239476680755615, + "learning_rate": 4.062577704102777e-06, + "loss": 0.5695, + "step": 2263 + }, + { + "epoch": 0.1876502279320348, + "grad_norm": 0.26193368434906006, + "learning_rate": 4.062163282221301e-06, + "loss": 0.5262, + "step": 2264 + }, + { + "epoch": 0.18773311230832987, + "grad_norm": 0.27485114336013794, + "learning_rate": 4.0617488603398264e-06, + "loss": 0.5602, + "step": 2265 + }, + { + "epoch": 0.18781599668462495, + "grad_norm": 0.26167067885398865, + "learning_rate": 4.061334438458351e-06, + "loss": 0.5122, + "step": 2266 + }, + { + "epoch": 0.18789888106092, + "grad_norm": 0.27411457896232605, + "learning_rate": 4.060920016576876e-06, + "loss": 0.5175, + "step": 2267 + }, + { + "epoch": 0.1879817654372151, + "grad_norm": 0.25722000002861023, + "learning_rate": 4.060505594695401e-06, + "loss": 0.4974, + "step": 2268 + }, + { + "epoch": 0.18806464981351015, + "grad_norm": 0.26605045795440674, + "learning_rate": 4.060091172813925e-06, + "loss": 0.5713, + "step": 2269 + }, + { + "epoch": 0.18814753418980523, + "grad_norm": 0.27133867144584656, + "learning_rate": 4.0596767509324495e-06, + "loss": 0.5528, + "step": 2270 + }, + { + "epoch": 0.1882304185661003, + "grad_norm": 0.28278085589408875, + "learning_rate": 4.059262329050974e-06, + "loss": 0.5615, + "step": 2271 + }, + { + "epoch": 0.18831330294239537, + "grad_norm": 0.2700541615486145, + "learning_rate": 4.058847907169498e-06, + "loss": 0.5545, + "step": 2272 + }, + { + "epoch": 0.18839618731869043, + "grad_norm": 0.26167821884155273, + "learning_rate": 4.058433485288024e-06, + "loss": 0.5355, + "step": 2273 + }, + { + "epoch": 0.18847907169498548, + "grad_norm": 0.27506908774375916, + "learning_rate": 4.058019063406548e-06, + "loss": 0.5422, + "step": 2274 + }, + { + "epoch": 0.18856195607128057, + "grad_norm": 0.2647694945335388, + "learning_rate": 4.057604641525073e-06, + "loss": 0.549, + "step": 2275 + }, + { + "epoch": 0.18864484044757562, + "grad_norm": 0.26303529739379883, + "learning_rate": 4.057190219643598e-06, + "loss": 0.5045, + "step": 2276 + }, + { + "epoch": 0.1887277248238707, + "grad_norm": 0.27338048815727234, + "learning_rate": 4.056775797762122e-06, + "loss": 0.5163, + "step": 2277 + }, + { + "epoch": 0.18881060920016576, + "grad_norm": 0.2745503783226013, + "learning_rate": 4.056361375880647e-06, + "loss": 0.546, + "step": 2278 + }, + { + "epoch": 0.18889349357646085, + "grad_norm": 0.2510717511177063, + "learning_rate": 4.055946953999171e-06, + "loss": 0.5194, + "step": 2279 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 0.24865059554576874, + "learning_rate": 4.055532532117696e-06, + "loss": 0.5421, + "step": 2280 + }, + { + "epoch": 0.189059262329051, + "grad_norm": 0.2797532081604004, + "learning_rate": 4.055118110236221e-06, + "loss": 0.5578, + "step": 2281 + }, + { + "epoch": 0.18914214670534604, + "grad_norm": 0.2846514582633972, + "learning_rate": 4.054703688354745e-06, + "loss": 0.5361, + "step": 2282 + }, + { + "epoch": 0.1892250310816411, + "grad_norm": 0.29234302043914795, + "learning_rate": 4.0542892664732705e-06, + "loss": 0.5493, + "step": 2283 + }, + { + "epoch": 0.18930791545793618, + "grad_norm": 0.26689597964286804, + "learning_rate": 4.053874844591795e-06, + "loss": 0.5637, + "step": 2284 + }, + { + "epoch": 0.18939079983423124, + "grad_norm": 0.2632163465023041, + "learning_rate": 4.053460422710319e-06, + "loss": 0.5756, + "step": 2285 + }, + { + "epoch": 0.18947368421052632, + "grad_norm": 0.25191187858581543, + "learning_rate": 4.053046000828844e-06, + "loss": 0.5628, + "step": 2286 + }, + { + "epoch": 0.18955656858682138, + "grad_norm": 0.26570284366607666, + "learning_rate": 4.052631578947368e-06, + "loss": 0.5089, + "step": 2287 + }, + { + "epoch": 0.18963945296311646, + "grad_norm": 0.28901341557502747, + "learning_rate": 4.0522171570658935e-06, + "loss": 0.5333, + "step": 2288 + }, + { + "epoch": 0.18972233733941152, + "grad_norm": 0.26453861594200134, + "learning_rate": 4.051802735184418e-06, + "loss": 0.5696, + "step": 2289 + }, + { + "epoch": 0.1898052217157066, + "grad_norm": 0.2689468264579773, + "learning_rate": 4.051388313302942e-06, + "loss": 0.5643, + "step": 2290 + }, + { + "epoch": 0.18988810609200166, + "grad_norm": 0.27002766728401184, + "learning_rate": 4.050973891421468e-06, + "loss": 0.5502, + "step": 2291 + }, + { + "epoch": 0.1899709904682967, + "grad_norm": 0.27927565574645996, + "learning_rate": 4.050559469539992e-06, + "loss": 0.5133, + "step": 2292 + }, + { + "epoch": 0.1900538748445918, + "grad_norm": 0.2642829120159149, + "learning_rate": 4.050145047658517e-06, + "loss": 0.5526, + "step": 2293 + }, + { + "epoch": 0.19013675922088685, + "grad_norm": 0.30148187279701233, + "learning_rate": 4.049730625777042e-06, + "loss": 0.5542, + "step": 2294 + }, + { + "epoch": 0.19021964359718194, + "grad_norm": 0.2608180642127991, + "learning_rate": 4.049316203895566e-06, + "loss": 0.5616, + "step": 2295 + }, + { + "epoch": 0.190302527973477, + "grad_norm": 0.2528807520866394, + "learning_rate": 4.048901782014091e-06, + "loss": 0.5159, + "step": 2296 + }, + { + "epoch": 0.19038541234977208, + "grad_norm": 0.25838831067085266, + "learning_rate": 4.048487360132615e-06, + "loss": 0.5141, + "step": 2297 + }, + { + "epoch": 0.19046829672606713, + "grad_norm": 0.27882540225982666, + "learning_rate": 4.04807293825114e-06, + "loss": 0.5724, + "step": 2298 + }, + { + "epoch": 0.19055118110236222, + "grad_norm": 0.2510994076728821, + "learning_rate": 4.047658516369665e-06, + "loss": 0.5616, + "step": 2299 + }, + { + "epoch": 0.19063406547865727, + "grad_norm": 0.28078368306159973, + "learning_rate": 4.047244094488189e-06, + "loss": 0.5502, + "step": 2300 + }, + { + "epoch": 0.19071694985495233, + "grad_norm": 0.2439626008272171, + "learning_rate": 4.0468296726067145e-06, + "loss": 0.5374, + "step": 2301 + }, + { + "epoch": 0.1907998342312474, + "grad_norm": 0.2626286447048187, + "learning_rate": 4.046415250725239e-06, + "loss": 0.545, + "step": 2302 + }, + { + "epoch": 0.19088271860754247, + "grad_norm": 0.29036855697631836, + "learning_rate": 4.046000828843763e-06, + "loss": 0.5257, + "step": 2303 + }, + { + "epoch": 0.19096560298383755, + "grad_norm": 0.2643589973449707, + "learning_rate": 4.045586406962288e-06, + "loss": 0.5618, + "step": 2304 + }, + { + "epoch": 0.1910484873601326, + "grad_norm": 0.2666000425815582, + "learning_rate": 4.045171985080812e-06, + "loss": 0.5432, + "step": 2305 + }, + { + "epoch": 0.1911313717364277, + "grad_norm": 0.27714964747428894, + "learning_rate": 4.0447575631993375e-06, + "loss": 0.54, + "step": 2306 + }, + { + "epoch": 0.19121425611272275, + "grad_norm": 0.3020853102207184, + "learning_rate": 4.044343141317862e-06, + "loss": 0.5748, + "step": 2307 + }, + { + "epoch": 0.19129714048901783, + "grad_norm": 0.24792137742042542, + "learning_rate": 4.043928719436386e-06, + "loss": 0.5523, + "step": 2308 + }, + { + "epoch": 0.1913800248653129, + "grad_norm": 0.2546408474445343, + "learning_rate": 4.043514297554912e-06, + "loss": 0.5363, + "step": 2309 + }, + { + "epoch": 0.19146290924160797, + "grad_norm": 0.3034515082836151, + "learning_rate": 4.043099875673436e-06, + "loss": 0.5449, + "step": 2310 + }, + { + "epoch": 0.19154579361790303, + "grad_norm": 0.2710953950881958, + "learning_rate": 4.0426854537919605e-06, + "loss": 0.5117, + "step": 2311 + }, + { + "epoch": 0.19162867799419808, + "grad_norm": 0.26484861969947815, + "learning_rate": 4.042271031910485e-06, + "loss": 0.52, + "step": 2312 + }, + { + "epoch": 0.19171156237049317, + "grad_norm": 0.274967759847641, + "learning_rate": 4.041856610029009e-06, + "loss": 0.5626, + "step": 2313 + }, + { + "epoch": 0.19179444674678822, + "grad_norm": 0.2616547644138336, + "learning_rate": 4.041442188147535e-06, + "loss": 0.5713, + "step": 2314 + }, + { + "epoch": 0.1918773311230833, + "grad_norm": 0.2724941074848175, + "learning_rate": 4.041027766266059e-06, + "loss": 0.5341, + "step": 2315 + }, + { + "epoch": 0.19196021549937836, + "grad_norm": 0.26077380776405334, + "learning_rate": 4.040613344384584e-06, + "loss": 0.5341, + "step": 2316 + }, + { + "epoch": 0.19204309987567345, + "grad_norm": 0.26929205656051636, + "learning_rate": 4.040198922503109e-06, + "loss": 0.5477, + "step": 2317 + }, + { + "epoch": 0.1921259842519685, + "grad_norm": 0.26607075333595276, + "learning_rate": 4.039784500621633e-06, + "loss": 0.5488, + "step": 2318 + }, + { + "epoch": 0.19220886862826358, + "grad_norm": 0.2683679759502411, + "learning_rate": 4.039370078740158e-06, + "loss": 0.53, + "step": 2319 + }, + { + "epoch": 0.19229175300455864, + "grad_norm": 0.2973255515098572, + "learning_rate": 4.038955656858682e-06, + "loss": 0.5681, + "step": 2320 + }, + { + "epoch": 0.1923746373808537, + "grad_norm": 0.27342161536216736, + "learning_rate": 4.038541234977207e-06, + "loss": 0.5876, + "step": 2321 + }, + { + "epoch": 0.19245752175714878, + "grad_norm": 0.25479137897491455, + "learning_rate": 4.038126813095732e-06, + "loss": 0.5724, + "step": 2322 + }, + { + "epoch": 0.19254040613344384, + "grad_norm": 0.27936598658561707, + "learning_rate": 4.037712391214256e-06, + "loss": 0.5626, + "step": 2323 + }, + { + "epoch": 0.19262329050973892, + "grad_norm": 0.2911500334739685, + "learning_rate": 4.0372979693327815e-06, + "loss": 0.5789, + "step": 2324 + }, + { + "epoch": 0.19270617488603398, + "grad_norm": 0.28159719705581665, + "learning_rate": 4.036883547451306e-06, + "loss": 0.5349, + "step": 2325 + }, + { + "epoch": 0.19278905926232906, + "grad_norm": 0.32901349663734436, + "learning_rate": 4.03646912556983e-06, + "loss": 0.5187, + "step": 2326 + }, + { + "epoch": 0.19287194363862412, + "grad_norm": 0.27703285217285156, + "learning_rate": 4.036054703688355e-06, + "loss": 0.5438, + "step": 2327 + }, + { + "epoch": 0.1929548280149192, + "grad_norm": 0.2976789176464081, + "learning_rate": 4.035640281806879e-06, + "loss": 0.5765, + "step": 2328 + }, + { + "epoch": 0.19303771239121426, + "grad_norm": 0.2562313675880432, + "learning_rate": 4.0352258599254045e-06, + "loss": 0.5377, + "step": 2329 + }, + { + "epoch": 0.1931205967675093, + "grad_norm": 0.2726582884788513, + "learning_rate": 4.034811438043929e-06, + "loss": 0.5527, + "step": 2330 + }, + { + "epoch": 0.1932034811438044, + "grad_norm": 0.28239166736602783, + "learning_rate": 4.034397016162453e-06, + "loss": 0.5123, + "step": 2331 + }, + { + "epoch": 0.19328636552009945, + "grad_norm": 0.26762309670448303, + "learning_rate": 4.033982594280979e-06, + "loss": 0.5467, + "step": 2332 + }, + { + "epoch": 0.19336924989639453, + "grad_norm": 0.2759624421596527, + "learning_rate": 4.033568172399503e-06, + "loss": 0.5651, + "step": 2333 + }, + { + "epoch": 0.1934521342726896, + "grad_norm": 0.2743963301181793, + "learning_rate": 4.033153750518028e-06, + "loss": 0.5745, + "step": 2334 + }, + { + "epoch": 0.19353501864898467, + "grad_norm": 0.25294139981269836, + "learning_rate": 4.032739328636553e-06, + "loss": 0.5403, + "step": 2335 + }, + { + "epoch": 0.19361790302527973, + "grad_norm": 0.28267377614974976, + "learning_rate": 4.032324906755077e-06, + "loss": 0.5892, + "step": 2336 + }, + { + "epoch": 0.19370078740157481, + "grad_norm": 0.2650120258331299, + "learning_rate": 4.031910484873602e-06, + "loss": 0.5984, + "step": 2337 + }, + { + "epoch": 0.19378367177786987, + "grad_norm": 0.26959946751594543, + "learning_rate": 4.031496062992126e-06, + "loss": 0.541, + "step": 2338 + }, + { + "epoch": 0.19386655615416493, + "grad_norm": 0.27294597029685974, + "learning_rate": 4.0310816411106505e-06, + "loss": 0.5079, + "step": 2339 + }, + { + "epoch": 0.19394944053046, + "grad_norm": 0.27900630235671997, + "learning_rate": 4.030667219229176e-06, + "loss": 0.5731, + "step": 2340 + }, + { + "epoch": 0.19403232490675507, + "grad_norm": 0.2748972177505493, + "learning_rate": 4.0302527973477e-06, + "loss": 0.5527, + "step": 2341 + }, + { + "epoch": 0.19411520928305015, + "grad_norm": 0.2801356017589569, + "learning_rate": 4.0298383754662255e-06, + "loss": 0.5717, + "step": 2342 + }, + { + "epoch": 0.1941980936593452, + "grad_norm": 0.2730942666530609, + "learning_rate": 4.02942395358475e-06, + "loss": 0.5658, + "step": 2343 + }, + { + "epoch": 0.1942809780356403, + "grad_norm": 0.25725293159484863, + "learning_rate": 4.029009531703274e-06, + "loss": 0.5289, + "step": 2344 + }, + { + "epoch": 0.19436386241193535, + "grad_norm": 0.24604655802249908, + "learning_rate": 4.028595109821799e-06, + "loss": 0.5082, + "step": 2345 + }, + { + "epoch": 0.19444674678823043, + "grad_norm": 0.29015737771987915, + "learning_rate": 4.028180687940323e-06, + "loss": 0.517, + "step": 2346 + }, + { + "epoch": 0.19452963116452548, + "grad_norm": 0.28759217262268066, + "learning_rate": 4.0277662660588485e-06, + "loss": 0.5494, + "step": 2347 + }, + { + "epoch": 0.19461251554082057, + "grad_norm": 0.2596593201160431, + "learning_rate": 4.027351844177373e-06, + "loss": 0.5385, + "step": 2348 + }, + { + "epoch": 0.19469539991711562, + "grad_norm": 0.2877366840839386, + "learning_rate": 4.026937422295897e-06, + "loss": 0.5303, + "step": 2349 + }, + { + "epoch": 0.19477828429341068, + "grad_norm": 0.26650217175483704, + "learning_rate": 4.026523000414423e-06, + "loss": 0.4971, + "step": 2350 + }, + { + "epoch": 0.19486116866970576, + "grad_norm": 0.2728070020675659, + "learning_rate": 4.026108578532947e-06, + "loss": 0.5337, + "step": 2351 + }, + { + "epoch": 0.19494405304600082, + "grad_norm": 0.2842119634151459, + "learning_rate": 4.0256941566514715e-06, + "loss": 0.5441, + "step": 2352 + }, + { + "epoch": 0.1950269374222959, + "grad_norm": 0.29169192910194397, + "learning_rate": 4.025279734769996e-06, + "loss": 0.5487, + "step": 2353 + }, + { + "epoch": 0.19510982179859096, + "grad_norm": 0.29716312885284424, + "learning_rate": 4.02486531288852e-06, + "loss": 0.5287, + "step": 2354 + }, + { + "epoch": 0.19519270617488604, + "grad_norm": 0.24269232153892517, + "learning_rate": 4.024450891007046e-06, + "loss": 0.5244, + "step": 2355 + }, + { + "epoch": 0.1952755905511811, + "grad_norm": 0.29505598545074463, + "learning_rate": 4.02403646912557e-06, + "loss": 0.5258, + "step": 2356 + }, + { + "epoch": 0.19535847492747618, + "grad_norm": 0.27719250321388245, + "learning_rate": 4.0236220472440945e-06, + "loss": 0.554, + "step": 2357 + }, + { + "epoch": 0.19544135930377124, + "grad_norm": 0.26356416940689087, + "learning_rate": 4.02320762536262e-06, + "loss": 0.5546, + "step": 2358 + }, + { + "epoch": 0.1955242436800663, + "grad_norm": 0.27456551790237427, + "learning_rate": 4.022793203481144e-06, + "loss": 0.5229, + "step": 2359 + }, + { + "epoch": 0.19560712805636138, + "grad_norm": 0.2759743928909302, + "learning_rate": 4.022378781599669e-06, + "loss": 0.5199, + "step": 2360 + }, + { + "epoch": 0.19569001243265643, + "grad_norm": 0.29080289602279663, + "learning_rate": 4.021964359718193e-06, + "loss": 0.5546, + "step": 2361 + }, + { + "epoch": 0.19577289680895152, + "grad_norm": 0.25306645035743713, + "learning_rate": 4.0215499378367175e-06, + "loss": 0.5044, + "step": 2362 + }, + { + "epoch": 0.19585578118524657, + "grad_norm": 0.2581048905849457, + "learning_rate": 4.021135515955243e-06, + "loss": 0.5196, + "step": 2363 + }, + { + "epoch": 0.19593866556154166, + "grad_norm": 0.2552899420261383, + "learning_rate": 4.020721094073767e-06, + "loss": 0.5256, + "step": 2364 + }, + { + "epoch": 0.19602154993783671, + "grad_norm": 0.283420592546463, + "learning_rate": 4.0203066721922925e-06, + "loss": 0.5589, + "step": 2365 + }, + { + "epoch": 0.1961044343141318, + "grad_norm": 0.29112669825553894, + "learning_rate": 4.019892250310817e-06, + "loss": 0.5939, + "step": 2366 + }, + { + "epoch": 0.19618731869042685, + "grad_norm": 0.26751163601875305, + "learning_rate": 4.019477828429341e-06, + "loss": 0.5277, + "step": 2367 + }, + { + "epoch": 0.1962702030667219, + "grad_norm": 0.2648759186267853, + "learning_rate": 4.019063406547866e-06, + "loss": 0.5396, + "step": 2368 + }, + { + "epoch": 0.196353087443017, + "grad_norm": 0.2749149203300476, + "learning_rate": 4.01864898466639e-06, + "loss": 0.5649, + "step": 2369 + }, + { + "epoch": 0.19643597181931205, + "grad_norm": 0.2717074155807495, + "learning_rate": 4.0182345627849155e-06, + "loss": 0.4954, + "step": 2370 + }, + { + "epoch": 0.19651885619560713, + "grad_norm": 0.27630528807640076, + "learning_rate": 4.01782014090344e-06, + "loss": 0.5959, + "step": 2371 + }, + { + "epoch": 0.1966017405719022, + "grad_norm": 0.2509850263595581, + "learning_rate": 4.017405719021964e-06, + "loss": 0.5315, + "step": 2372 + }, + { + "epoch": 0.19668462494819727, + "grad_norm": 0.3004663586616516, + "learning_rate": 4.01699129714049e-06, + "loss": 0.5515, + "step": 2373 + }, + { + "epoch": 0.19676750932449233, + "grad_norm": 0.2596939206123352, + "learning_rate": 4.016576875259014e-06, + "loss": 0.5361, + "step": 2374 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 0.25939297676086426, + "learning_rate": 4.0161624533775385e-06, + "loss": 0.5648, + "step": 2375 + }, + { + "epoch": 0.19693327807708247, + "grad_norm": 0.27013036608695984, + "learning_rate": 4.015748031496064e-06, + "loss": 0.5512, + "step": 2376 + }, + { + "epoch": 0.19701616245337752, + "grad_norm": 0.247028648853302, + "learning_rate": 4.015333609614588e-06, + "loss": 0.5347, + "step": 2377 + }, + { + "epoch": 0.1970990468296726, + "grad_norm": 0.28583574295043945, + "learning_rate": 4.014919187733113e-06, + "loss": 0.5645, + "step": 2378 + }, + { + "epoch": 0.19718193120596766, + "grad_norm": 0.26603934168815613, + "learning_rate": 4.014504765851637e-06, + "loss": 0.5459, + "step": 2379 + }, + { + "epoch": 0.19726481558226275, + "grad_norm": 0.3204175531864166, + "learning_rate": 4.0140903439701615e-06, + "loss": 0.5206, + "step": 2380 + }, + { + "epoch": 0.1973476999585578, + "grad_norm": 0.28492340445518494, + "learning_rate": 4.013675922088687e-06, + "loss": 0.6171, + "step": 2381 + }, + { + "epoch": 0.1974305843348529, + "grad_norm": 0.2853088676929474, + "learning_rate": 4.013261500207211e-06, + "loss": 0.5977, + "step": 2382 + }, + { + "epoch": 0.19751346871114794, + "grad_norm": 0.2664131820201874, + "learning_rate": 4.0128470783257365e-06, + "loss": 0.5792, + "step": 2383 + }, + { + "epoch": 0.19759635308744303, + "grad_norm": 0.2932121455669403, + "learning_rate": 4.012432656444261e-06, + "loss": 0.5619, + "step": 2384 + }, + { + "epoch": 0.19767923746373808, + "grad_norm": 0.28895750641822815, + "learning_rate": 4.012018234562785e-06, + "loss": 0.5957, + "step": 2385 + }, + { + "epoch": 0.19776212184003317, + "grad_norm": 0.25806891918182373, + "learning_rate": 4.01160381268131e-06, + "loss": 0.5332, + "step": 2386 + }, + { + "epoch": 0.19784500621632822, + "grad_norm": 0.2970893979072571, + "learning_rate": 4.011189390799834e-06, + "loss": 0.6342, + "step": 2387 + }, + { + "epoch": 0.19792789059262328, + "grad_norm": 0.2573249340057373, + "learning_rate": 4.0107749689183595e-06, + "loss": 0.5336, + "step": 2388 + }, + { + "epoch": 0.19801077496891836, + "grad_norm": 0.286761999130249, + "learning_rate": 4.010360547036884e-06, + "loss": 0.5381, + "step": 2389 + }, + { + "epoch": 0.19809365934521342, + "grad_norm": 0.26823630928993225, + "learning_rate": 4.009946125155408e-06, + "loss": 0.5779, + "step": 2390 + }, + { + "epoch": 0.1981765437215085, + "grad_norm": 0.2515113353729248, + "learning_rate": 4.009531703273934e-06, + "loss": 0.4886, + "step": 2391 + }, + { + "epoch": 0.19825942809780356, + "grad_norm": 0.2518639862537384, + "learning_rate": 4.009117281392458e-06, + "loss": 0.4774, + "step": 2392 + }, + { + "epoch": 0.19834231247409864, + "grad_norm": 0.26063916087150574, + "learning_rate": 4.0087028595109825e-06, + "loss": 0.5588, + "step": 2393 + }, + { + "epoch": 0.1984251968503937, + "grad_norm": 0.2636895477771759, + "learning_rate": 4.008288437629507e-06, + "loss": 0.5265, + "step": 2394 + }, + { + "epoch": 0.19850808122668878, + "grad_norm": 0.25976434350013733, + "learning_rate": 4.007874015748031e-06, + "loss": 0.5036, + "step": 2395 + }, + { + "epoch": 0.19859096560298384, + "grad_norm": 0.2727707624435425, + "learning_rate": 4.007459593866557e-06, + "loss": 0.5156, + "step": 2396 + }, + { + "epoch": 0.1986738499792789, + "grad_norm": 0.2579866051673889, + "learning_rate": 4.007045171985081e-06, + "loss": 0.5167, + "step": 2397 + }, + { + "epoch": 0.19875673435557398, + "grad_norm": 0.2662644684314728, + "learning_rate": 4.0066307501036055e-06, + "loss": 0.5674, + "step": 2398 + }, + { + "epoch": 0.19883961873186903, + "grad_norm": 0.2610207498073578, + "learning_rate": 4.006216328222131e-06, + "loss": 0.5255, + "step": 2399 + }, + { + "epoch": 0.19892250310816412, + "grad_norm": 0.287724107503891, + "learning_rate": 4.005801906340655e-06, + "loss": 0.548, + "step": 2400 + }, + { + "epoch": 0.19900538748445917, + "grad_norm": 0.2967718541622162, + "learning_rate": 4.00538748445918e-06, + "loss": 0.5451, + "step": 2401 + }, + { + "epoch": 0.19908827186075426, + "grad_norm": 0.29087167978286743, + "learning_rate": 4.004973062577704e-06, + "loss": 0.5725, + "step": 2402 + }, + { + "epoch": 0.1991711562370493, + "grad_norm": 0.2646278440952301, + "learning_rate": 4.0045586406962285e-06, + "loss": 0.5739, + "step": 2403 + }, + { + "epoch": 0.1992540406133444, + "grad_norm": 0.2585521936416626, + "learning_rate": 4.004144218814754e-06, + "loss": 0.5605, + "step": 2404 + }, + { + "epoch": 0.19933692498963945, + "grad_norm": 0.2565876841545105, + "learning_rate": 4.003729796933278e-06, + "loss": 0.5423, + "step": 2405 + }, + { + "epoch": 0.1994198093659345, + "grad_norm": 0.27468031644821167, + "learning_rate": 4.003315375051803e-06, + "loss": 0.5542, + "step": 2406 + }, + { + "epoch": 0.1995026937422296, + "grad_norm": 0.26924753189086914, + "learning_rate": 4.002900953170328e-06, + "loss": 0.5705, + "step": 2407 + }, + { + "epoch": 0.19958557811852465, + "grad_norm": 0.2999148368835449, + "learning_rate": 4.002486531288852e-06, + "loss": 0.5606, + "step": 2408 + }, + { + "epoch": 0.19966846249481973, + "grad_norm": 0.24480214715003967, + "learning_rate": 4.002072109407377e-06, + "loss": 0.5391, + "step": 2409 + }, + { + "epoch": 0.1997513468711148, + "grad_norm": 0.2806302607059479, + "learning_rate": 4.001657687525902e-06, + "loss": 0.588, + "step": 2410 + }, + { + "epoch": 0.19983423124740987, + "grad_norm": 0.25648003816604614, + "learning_rate": 4.0012432656444265e-06, + "loss": 0.5591, + "step": 2411 + }, + { + "epoch": 0.19991711562370493, + "grad_norm": 0.25598594546318054, + "learning_rate": 4.000828843762951e-06, + "loss": 0.5021, + "step": 2412 + }, + { + "epoch": 0.2, + "grad_norm": 0.27295467257499695, + "learning_rate": 4.000414421881475e-06, + "loss": 0.5505, + "step": 2413 + }, + { + "epoch": 0.20008288437629507, + "grad_norm": 0.2458568513393402, + "learning_rate": 4.000000000000001e-06, + "loss": 0.5333, + "step": 2414 + }, + { + "epoch": 0.20016576875259012, + "grad_norm": 0.29450324177742004, + "learning_rate": 3.999585578118525e-06, + "loss": 0.5814, + "step": 2415 + }, + { + "epoch": 0.2002486531288852, + "grad_norm": 0.280734121799469, + "learning_rate": 3.9991711562370495e-06, + "loss": 0.5189, + "step": 2416 + }, + { + "epoch": 0.20033153750518026, + "grad_norm": 0.26119354367256165, + "learning_rate": 3.998756734355575e-06, + "loss": 0.5123, + "step": 2417 + }, + { + "epoch": 0.20041442188147535, + "grad_norm": 0.2705000638961792, + "learning_rate": 3.998342312474099e-06, + "loss": 0.5162, + "step": 2418 + }, + { + "epoch": 0.2004973062577704, + "grad_norm": 0.2955061197280884, + "learning_rate": 3.997927890592624e-06, + "loss": 0.5566, + "step": 2419 + }, + { + "epoch": 0.20058019063406549, + "grad_norm": 0.24515260756015778, + "learning_rate": 3.997513468711148e-06, + "loss": 0.5176, + "step": 2420 + }, + { + "epoch": 0.20066307501036054, + "grad_norm": 0.25421014428138733, + "learning_rate": 3.9970990468296725e-06, + "loss": 0.5504, + "step": 2421 + }, + { + "epoch": 0.20074595938665563, + "grad_norm": 0.28925713896751404, + "learning_rate": 3.996684624948198e-06, + "loss": 0.532, + "step": 2422 + }, + { + "epoch": 0.20082884376295068, + "grad_norm": 0.27971190214157104, + "learning_rate": 3.996270203066722e-06, + "loss": 0.5927, + "step": 2423 + }, + { + "epoch": 0.20091172813924577, + "grad_norm": 0.26966556906700134, + "learning_rate": 3.995855781185247e-06, + "loss": 0.5869, + "step": 2424 + }, + { + "epoch": 0.20099461251554082, + "grad_norm": 0.27686607837677, + "learning_rate": 3.995441359303772e-06, + "loss": 0.5438, + "step": 2425 + }, + { + "epoch": 0.20107749689183588, + "grad_norm": 0.2525418996810913, + "learning_rate": 3.995026937422296e-06, + "loss": 0.4964, + "step": 2426 + }, + { + "epoch": 0.20116038126813096, + "grad_norm": 0.2659456729888916, + "learning_rate": 3.994612515540821e-06, + "loss": 0.5182, + "step": 2427 + }, + { + "epoch": 0.20124326564442602, + "grad_norm": 0.2630992829799652, + "learning_rate": 3.994198093659345e-06, + "loss": 0.5396, + "step": 2428 + }, + { + "epoch": 0.2013261500207211, + "grad_norm": 0.26590651273727417, + "learning_rate": 3.99378367177787e-06, + "loss": 0.5584, + "step": 2429 + }, + { + "epoch": 0.20140903439701616, + "grad_norm": 0.25673818588256836, + "learning_rate": 3.993369249896395e-06, + "loss": 0.5431, + "step": 2430 + }, + { + "epoch": 0.20149191877331124, + "grad_norm": 0.268262654542923, + "learning_rate": 3.992954828014919e-06, + "loss": 0.5443, + "step": 2431 + }, + { + "epoch": 0.2015748031496063, + "grad_norm": 0.26642611622810364, + "learning_rate": 3.992540406133445e-06, + "loss": 0.5515, + "step": 2432 + }, + { + "epoch": 0.20165768752590138, + "grad_norm": 0.27177664637565613, + "learning_rate": 3.992125984251969e-06, + "loss": 0.5865, + "step": 2433 + }, + { + "epoch": 0.20174057190219644, + "grad_norm": 0.2767528295516968, + "learning_rate": 3.9917115623704935e-06, + "loss": 0.5458, + "step": 2434 + }, + { + "epoch": 0.2018234562784915, + "grad_norm": 0.28507357835769653, + "learning_rate": 3.991297140489018e-06, + "loss": 0.5168, + "step": 2435 + }, + { + "epoch": 0.20190634065478658, + "grad_norm": 0.2662017345428467, + "learning_rate": 3.990882718607542e-06, + "loss": 0.5446, + "step": 2436 + }, + { + "epoch": 0.20198922503108163, + "grad_norm": 0.27331802248954773, + "learning_rate": 3.990468296726068e-06, + "loss": 0.528, + "step": 2437 + }, + { + "epoch": 0.20207210940737672, + "grad_norm": 0.2618383765220642, + "learning_rate": 3.990053874844592e-06, + "loss": 0.5371, + "step": 2438 + }, + { + "epoch": 0.20215499378367177, + "grad_norm": 0.2639026343822479, + "learning_rate": 3.9896394529631165e-06, + "loss": 0.5236, + "step": 2439 + }, + { + "epoch": 0.20223787815996686, + "grad_norm": 0.2536134421825409, + "learning_rate": 3.989225031081642e-06, + "loss": 0.506, + "step": 2440 + }, + { + "epoch": 0.2023207625362619, + "grad_norm": 0.2843979001045227, + "learning_rate": 3.988810609200166e-06, + "loss": 0.5591, + "step": 2441 + }, + { + "epoch": 0.202403646912557, + "grad_norm": 0.2682885527610779, + "learning_rate": 3.988396187318691e-06, + "loss": 0.5335, + "step": 2442 + }, + { + "epoch": 0.20248653128885205, + "grad_norm": 0.26549550890922546, + "learning_rate": 3.987981765437215e-06, + "loss": 0.5386, + "step": 2443 + }, + { + "epoch": 0.2025694156651471, + "grad_norm": 0.2541399598121643, + "learning_rate": 3.9875673435557395e-06, + "loss": 0.5359, + "step": 2444 + }, + { + "epoch": 0.2026523000414422, + "grad_norm": 0.2619926631450653, + "learning_rate": 3.987152921674265e-06, + "loss": 0.5231, + "step": 2445 + }, + { + "epoch": 0.20273518441773725, + "grad_norm": 0.2545678913593292, + "learning_rate": 3.986738499792789e-06, + "loss": 0.5613, + "step": 2446 + }, + { + "epoch": 0.20281806879403233, + "grad_norm": 0.27846792340278625, + "learning_rate": 3.986324077911314e-06, + "loss": 0.5414, + "step": 2447 + }, + { + "epoch": 0.20290095317032739, + "grad_norm": 0.26167652010917664, + "learning_rate": 3.985909656029839e-06, + "loss": 0.528, + "step": 2448 + }, + { + "epoch": 0.20298383754662247, + "grad_norm": 0.2613655626773834, + "learning_rate": 3.985495234148363e-06, + "loss": 0.5687, + "step": 2449 + }, + { + "epoch": 0.20306672192291753, + "grad_norm": 0.2518779933452606, + "learning_rate": 3.985080812266888e-06, + "loss": 0.5369, + "step": 2450 + }, + { + "epoch": 0.2031496062992126, + "grad_norm": 0.28255903720855713, + "learning_rate": 3.984666390385413e-06, + "loss": 0.5829, + "step": 2451 + }, + { + "epoch": 0.20323249067550767, + "grad_norm": 0.23449285328388214, + "learning_rate": 3.9842519685039375e-06, + "loss": 0.5062, + "step": 2452 + }, + { + "epoch": 0.20331537505180272, + "grad_norm": 0.258163183927536, + "learning_rate": 3.983837546622462e-06, + "loss": 0.5505, + "step": 2453 + }, + { + "epoch": 0.2033982594280978, + "grad_norm": 0.2583516240119934, + "learning_rate": 3.983423124740986e-06, + "loss": 0.5574, + "step": 2454 + }, + { + "epoch": 0.20348114380439286, + "grad_norm": 0.2632848024368286, + "learning_rate": 3.983008702859512e-06, + "loss": 0.5405, + "step": 2455 + }, + { + "epoch": 0.20356402818068794, + "grad_norm": 0.25856146216392517, + "learning_rate": 3.982594280978036e-06, + "loss": 0.5563, + "step": 2456 + }, + { + "epoch": 0.203646912556983, + "grad_norm": 0.2509797513484955, + "learning_rate": 3.9821798590965606e-06, + "loss": 0.5576, + "step": 2457 + }, + { + "epoch": 0.20372979693327808, + "grad_norm": 0.2947924733161926, + "learning_rate": 3.981765437215086e-06, + "loss": 0.5357, + "step": 2458 + }, + { + "epoch": 0.20381268130957314, + "grad_norm": 0.2536151111125946, + "learning_rate": 3.98135101533361e-06, + "loss": 0.5649, + "step": 2459 + }, + { + "epoch": 0.20389556568586822, + "grad_norm": 0.28117722272872925, + "learning_rate": 3.980936593452135e-06, + "loss": 0.5713, + "step": 2460 + }, + { + "epoch": 0.20397845006216328, + "grad_norm": 0.2700258493423462, + "learning_rate": 3.980522171570659e-06, + "loss": 0.5426, + "step": 2461 + }, + { + "epoch": 0.20406133443845836, + "grad_norm": 0.2621145248413086, + "learning_rate": 3.9801077496891836e-06, + "loss": 0.5212, + "step": 2462 + }, + { + "epoch": 0.20414421881475342, + "grad_norm": 0.26133185625076294, + "learning_rate": 3.979693327807709e-06, + "loss": 0.5426, + "step": 2463 + }, + { + "epoch": 0.20422710319104848, + "grad_norm": 0.2649715840816498, + "learning_rate": 3.979278905926233e-06, + "loss": 0.5396, + "step": 2464 + }, + { + "epoch": 0.20430998756734356, + "grad_norm": 0.26538264751434326, + "learning_rate": 3.978864484044758e-06, + "loss": 0.554, + "step": 2465 + }, + { + "epoch": 0.20439287194363862, + "grad_norm": 0.26202014088630676, + "learning_rate": 3.978450062163283e-06, + "loss": 0.5211, + "step": 2466 + }, + { + "epoch": 0.2044757563199337, + "grad_norm": 0.2919391095638275, + "learning_rate": 3.978035640281807e-06, + "loss": 0.5736, + "step": 2467 + }, + { + "epoch": 0.20455864069622876, + "grad_norm": 0.2770676612854004, + "learning_rate": 3.977621218400332e-06, + "loss": 0.5582, + "step": 2468 + }, + { + "epoch": 0.20464152507252384, + "grad_norm": 0.27271661162376404, + "learning_rate": 3.977206796518856e-06, + "loss": 0.5177, + "step": 2469 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 0.29170361161231995, + "learning_rate": 3.976792374637381e-06, + "loss": 0.5411, + "step": 2470 + }, + { + "epoch": 0.20480729382511398, + "grad_norm": 0.26971083879470825, + "learning_rate": 3.976377952755906e-06, + "loss": 0.5487, + "step": 2471 + }, + { + "epoch": 0.20489017820140903, + "grad_norm": 0.27147722244262695, + "learning_rate": 3.97596353087443e-06, + "loss": 0.6135, + "step": 2472 + }, + { + "epoch": 0.2049730625777041, + "grad_norm": 0.24475936591625214, + "learning_rate": 3.975549108992956e-06, + "loss": 0.5054, + "step": 2473 + }, + { + "epoch": 0.20505594695399917, + "grad_norm": 0.27217504382133484, + "learning_rate": 3.97513468711148e-06, + "loss": 0.5345, + "step": 2474 + }, + { + "epoch": 0.20513883133029423, + "grad_norm": 0.26753705739974976, + "learning_rate": 3.9747202652300046e-06, + "loss": 0.5424, + "step": 2475 + }, + { + "epoch": 0.2052217157065893, + "grad_norm": 0.26949381828308105, + "learning_rate": 3.974305843348529e-06, + "loss": 0.5108, + "step": 2476 + }, + { + "epoch": 0.20530460008288437, + "grad_norm": 0.2563508152961731, + "learning_rate": 3.973891421467053e-06, + "loss": 0.5586, + "step": 2477 + }, + { + "epoch": 0.20538748445917945, + "grad_norm": 0.27005741000175476, + "learning_rate": 3.973476999585578e-06, + "loss": 0.5654, + "step": 2478 + }, + { + "epoch": 0.2054703688354745, + "grad_norm": 0.2672484517097473, + "learning_rate": 3.973062577704103e-06, + "loss": 0.535, + "step": 2479 + }, + { + "epoch": 0.2055532532117696, + "grad_norm": 0.24369432032108307, + "learning_rate": 3.9726481558226276e-06, + "loss": 0.4748, + "step": 2480 + }, + { + "epoch": 0.20563613758806465, + "grad_norm": 0.28356438875198364, + "learning_rate": 3.972233733941153e-06, + "loss": 0.566, + "step": 2481 + }, + { + "epoch": 0.2057190219643597, + "grad_norm": 0.2838074266910553, + "learning_rate": 3.971819312059677e-06, + "loss": 0.5822, + "step": 2482 + }, + { + "epoch": 0.2058019063406548, + "grad_norm": 0.25672680139541626, + "learning_rate": 3.971404890178202e-06, + "loss": 0.5181, + "step": 2483 + }, + { + "epoch": 0.20588479071694984, + "grad_norm": 0.2727961838245392, + "learning_rate": 3.970990468296726e-06, + "loss": 0.5467, + "step": 2484 + }, + { + "epoch": 0.20596767509324493, + "grad_norm": 0.2902170717716217, + "learning_rate": 3.9705760464152506e-06, + "loss": 0.5792, + "step": 2485 + }, + { + "epoch": 0.20605055946953998, + "grad_norm": 0.2715618908405304, + "learning_rate": 3.970161624533776e-06, + "loss": 0.57, + "step": 2486 + }, + { + "epoch": 0.20613344384583507, + "grad_norm": 0.27671393752098083, + "learning_rate": 3.9697472026523e-06, + "loss": 0.5342, + "step": 2487 + }, + { + "epoch": 0.20621632822213012, + "grad_norm": 0.2692795693874359, + "learning_rate": 3.969332780770825e-06, + "loss": 0.5434, + "step": 2488 + }, + { + "epoch": 0.2062992125984252, + "grad_norm": 0.2631853222846985, + "learning_rate": 3.96891835888935e-06, + "loss": 0.5583, + "step": 2489 + }, + { + "epoch": 0.20638209697472026, + "grad_norm": 0.29091888666152954, + "learning_rate": 3.968503937007874e-06, + "loss": 0.6046, + "step": 2490 + }, + { + "epoch": 0.20646498135101532, + "grad_norm": 0.2548522353172302, + "learning_rate": 3.968089515126399e-06, + "loss": 0.5306, + "step": 2491 + }, + { + "epoch": 0.2065478657273104, + "grad_norm": 0.2600969970226288, + "learning_rate": 3.967675093244924e-06, + "loss": 0.5664, + "step": 2492 + }, + { + "epoch": 0.20663075010360546, + "grad_norm": 0.2690175175666809, + "learning_rate": 3.9672606713634486e-06, + "loss": 0.5627, + "step": 2493 + }, + { + "epoch": 0.20671363447990054, + "grad_norm": 0.2564936578273773, + "learning_rate": 3.966846249481973e-06, + "loss": 0.5208, + "step": 2494 + }, + { + "epoch": 0.2067965188561956, + "grad_norm": 0.25396019220352173, + "learning_rate": 3.9664318276004974e-06, + "loss": 0.5214, + "step": 2495 + }, + { + "epoch": 0.20687940323249068, + "grad_norm": 0.26784664392471313, + "learning_rate": 3.966017405719022e-06, + "loss": 0.5115, + "step": 2496 + }, + { + "epoch": 0.20696228760878574, + "grad_norm": 0.29291191697120667, + "learning_rate": 3.965602983837547e-06, + "loss": 0.5934, + "step": 2497 + }, + { + "epoch": 0.20704517198508082, + "grad_norm": 0.27429720759391785, + "learning_rate": 3.9651885619560716e-06, + "loss": 0.521, + "step": 2498 + }, + { + "epoch": 0.20712805636137588, + "grad_norm": 0.2501922845840454, + "learning_rate": 3.964774140074597e-06, + "loss": 0.5495, + "step": 2499 + }, + { + "epoch": 0.20721094073767096, + "grad_norm": 0.26221996545791626, + "learning_rate": 3.964359718193121e-06, + "loss": 0.5453, + "step": 2500 + }, + { + "epoch": 0.20729382511396602, + "grad_norm": 0.27179309725761414, + "learning_rate": 3.963945296311646e-06, + "loss": 0.5289, + "step": 2501 + }, + { + "epoch": 0.20737670949026107, + "grad_norm": 0.2790723741054535, + "learning_rate": 3.96353087443017e-06, + "loss": 0.5472, + "step": 2502 + }, + { + "epoch": 0.20745959386655616, + "grad_norm": 0.2802453637123108, + "learning_rate": 3.963116452548695e-06, + "loss": 0.6219, + "step": 2503 + }, + { + "epoch": 0.2075424782428512, + "grad_norm": 0.2596874237060547, + "learning_rate": 3.96270203066722e-06, + "loss": 0.4962, + "step": 2504 + }, + { + "epoch": 0.2076253626191463, + "grad_norm": 0.26879316568374634, + "learning_rate": 3.962287608785744e-06, + "loss": 0.5681, + "step": 2505 + }, + { + "epoch": 0.20770824699544135, + "grad_norm": 0.23926183581352234, + "learning_rate": 3.961873186904269e-06, + "loss": 0.482, + "step": 2506 + }, + { + "epoch": 0.20779113137173644, + "grad_norm": 0.28089720010757446, + "learning_rate": 3.961458765022794e-06, + "loss": 0.5432, + "step": 2507 + }, + { + "epoch": 0.2078740157480315, + "grad_norm": 0.281937837600708, + "learning_rate": 3.9610443431413184e-06, + "loss": 0.5352, + "step": 2508 + }, + { + "epoch": 0.20795690012432658, + "grad_norm": 0.27514129877090454, + "learning_rate": 3.960629921259843e-06, + "loss": 0.585, + "step": 2509 + }, + { + "epoch": 0.20803978450062163, + "grad_norm": 0.27585554122924805, + "learning_rate": 3.960215499378367e-06, + "loss": 0.5842, + "step": 2510 + }, + { + "epoch": 0.2081226688769167, + "grad_norm": 0.2808743417263031, + "learning_rate": 3.959801077496892e-06, + "loss": 0.5625, + "step": 2511 + }, + { + "epoch": 0.20820555325321177, + "grad_norm": 0.26887476444244385, + "learning_rate": 3.959386655615417e-06, + "loss": 0.5321, + "step": 2512 + }, + { + "epoch": 0.20828843762950683, + "grad_norm": 0.26045581698417664, + "learning_rate": 3.9589722337339414e-06, + "loss": 0.5146, + "step": 2513 + }, + { + "epoch": 0.2083713220058019, + "grad_norm": 0.29043132066726685, + "learning_rate": 3.958557811852466e-06, + "loss": 0.6366, + "step": 2514 + }, + { + "epoch": 0.20845420638209697, + "grad_norm": 0.2598252296447754, + "learning_rate": 3.958143389970991e-06, + "loss": 0.5282, + "step": 2515 + }, + { + "epoch": 0.20853709075839205, + "grad_norm": 0.2736597955226898, + "learning_rate": 3.957728968089516e-06, + "loss": 0.5894, + "step": 2516 + }, + { + "epoch": 0.2086199751346871, + "grad_norm": 0.2793636918067932, + "learning_rate": 3.95731454620804e-06, + "loss": 0.5671, + "step": 2517 + }, + { + "epoch": 0.2087028595109822, + "grad_norm": 0.24965748190879822, + "learning_rate": 3.9569001243265644e-06, + "loss": 0.5128, + "step": 2518 + }, + { + "epoch": 0.20878574388727725, + "grad_norm": 0.2863824963569641, + "learning_rate": 3.956485702445089e-06, + "loss": 0.5974, + "step": 2519 + }, + { + "epoch": 0.2088686282635723, + "grad_norm": 0.25720325112342834, + "learning_rate": 3.956071280563614e-06, + "loss": 0.5416, + "step": 2520 + }, + { + "epoch": 0.2089515126398674, + "grad_norm": 0.26290643215179443, + "learning_rate": 3.955656858682139e-06, + "loss": 0.5541, + "step": 2521 + }, + { + "epoch": 0.20903439701616244, + "grad_norm": 0.2537500262260437, + "learning_rate": 3.955242436800664e-06, + "loss": 0.5487, + "step": 2522 + }, + { + "epoch": 0.20911728139245753, + "grad_norm": 0.2813173532485962, + "learning_rate": 3.954828014919188e-06, + "loss": 0.5376, + "step": 2523 + }, + { + "epoch": 0.20920016576875258, + "grad_norm": 0.27317333221435547, + "learning_rate": 3.954413593037713e-06, + "loss": 0.5459, + "step": 2524 + }, + { + "epoch": 0.20928305014504767, + "grad_norm": 0.26985445618629456, + "learning_rate": 3.953999171156237e-06, + "loss": 0.5409, + "step": 2525 + }, + { + "epoch": 0.20936593452134272, + "grad_norm": 0.268887460231781, + "learning_rate": 3.953584749274762e-06, + "loss": 0.5625, + "step": 2526 + }, + { + "epoch": 0.2094488188976378, + "grad_norm": 0.2637885510921478, + "learning_rate": 3.953170327393286e-06, + "loss": 0.5671, + "step": 2527 + }, + { + "epoch": 0.20953170327393286, + "grad_norm": 0.2621043920516968, + "learning_rate": 3.952755905511811e-06, + "loss": 0.5754, + "step": 2528 + }, + { + "epoch": 0.20961458765022792, + "grad_norm": 0.2581386864185333, + "learning_rate": 3.952341483630336e-06, + "loss": 0.5157, + "step": 2529 + }, + { + "epoch": 0.209697472026523, + "grad_norm": 0.27027589082717896, + "learning_rate": 3.951927061748861e-06, + "loss": 0.5991, + "step": 2530 + }, + { + "epoch": 0.20978035640281806, + "grad_norm": 0.26529431343078613, + "learning_rate": 3.9515126398673854e-06, + "loss": 0.546, + "step": 2531 + }, + { + "epoch": 0.20986324077911314, + "grad_norm": 0.2826497256755829, + "learning_rate": 3.95109821798591e-06, + "loss": 0.558, + "step": 2532 + }, + { + "epoch": 0.2099461251554082, + "grad_norm": 0.27642136812210083, + "learning_rate": 3.950683796104435e-06, + "loss": 0.5564, + "step": 2533 + }, + { + "epoch": 0.21002900953170328, + "grad_norm": 0.2601974308490753, + "learning_rate": 3.95026937422296e-06, + "loss": 0.5631, + "step": 2534 + }, + { + "epoch": 0.21011189390799834, + "grad_norm": 0.25149500370025635, + "learning_rate": 3.949854952341484e-06, + "loss": 0.5115, + "step": 2535 + }, + { + "epoch": 0.21019477828429342, + "grad_norm": 0.27119478583335876, + "learning_rate": 3.9494405304600085e-06, + "loss": 0.5281, + "step": 2536 + }, + { + "epoch": 0.21027766266058848, + "grad_norm": 0.26463571190834045, + "learning_rate": 3.949026108578533e-06, + "loss": 0.5452, + "step": 2537 + }, + { + "epoch": 0.21036054703688356, + "grad_norm": 0.27551692724227905, + "learning_rate": 3.948611686697058e-06, + "loss": 0.5675, + "step": 2538 + }, + { + "epoch": 0.21044343141317862, + "grad_norm": 0.2739616632461548, + "learning_rate": 3.948197264815583e-06, + "loss": 0.5292, + "step": 2539 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 0.2684668302536011, + "learning_rate": 3.947782842934108e-06, + "loss": 0.5482, + "step": 2540 + }, + { + "epoch": 0.21060920016576876, + "grad_norm": 0.2718030512332916, + "learning_rate": 3.947368421052632e-06, + "loss": 0.5555, + "step": 2541 + }, + { + "epoch": 0.2106920845420638, + "grad_norm": 0.2936131954193115, + "learning_rate": 3.946953999171157e-06, + "loss": 0.6006, + "step": 2542 + }, + { + "epoch": 0.2107749689183589, + "grad_norm": 0.2688820958137512, + "learning_rate": 3.946539577289681e-06, + "loss": 0.5266, + "step": 2543 + }, + { + "epoch": 0.21085785329465395, + "grad_norm": 0.2770909070968628, + "learning_rate": 3.946125155408206e-06, + "loss": 0.5176, + "step": 2544 + }, + { + "epoch": 0.21094073767094904, + "grad_norm": 0.2798771560192108, + "learning_rate": 3.94571073352673e-06, + "loss": 0.5444, + "step": 2545 + }, + { + "epoch": 0.2110236220472441, + "grad_norm": 0.2832706570625305, + "learning_rate": 3.945296311645255e-06, + "loss": 0.5364, + "step": 2546 + }, + { + "epoch": 0.21110650642353918, + "grad_norm": 0.2691981792449951, + "learning_rate": 3.94488188976378e-06, + "loss": 0.5315, + "step": 2547 + }, + { + "epoch": 0.21118939079983423, + "grad_norm": 0.2820921242237091, + "learning_rate": 3.944467467882305e-06, + "loss": 0.5571, + "step": 2548 + }, + { + "epoch": 0.2112722751761293, + "grad_norm": 0.24656818807125092, + "learning_rate": 3.9440530460008295e-06, + "loss": 0.5347, + "step": 2549 + }, + { + "epoch": 0.21135515955242437, + "grad_norm": 0.2701530158519745, + "learning_rate": 3.943638624119354e-06, + "loss": 0.5158, + "step": 2550 + }, + { + "epoch": 0.21143804392871943, + "grad_norm": 0.270619660615921, + "learning_rate": 3.943224202237878e-06, + "loss": 0.5905, + "step": 2551 + }, + { + "epoch": 0.2115209283050145, + "grad_norm": 0.27600327134132385, + "learning_rate": 3.942809780356403e-06, + "loss": 0.5604, + "step": 2552 + }, + { + "epoch": 0.21160381268130957, + "grad_norm": 0.26175829768180847, + "learning_rate": 3.942395358474928e-06, + "loss": 0.5412, + "step": 2553 + }, + { + "epoch": 0.21168669705760465, + "grad_norm": 0.28806498646736145, + "learning_rate": 3.9419809365934525e-06, + "loss": 0.5808, + "step": 2554 + }, + { + "epoch": 0.2117695814338997, + "grad_norm": 0.2653636634349823, + "learning_rate": 3.941566514711977e-06, + "loss": 0.525, + "step": 2555 + }, + { + "epoch": 0.2118524658101948, + "grad_norm": 0.2753581702709198, + "learning_rate": 3.941152092830502e-06, + "loss": 0.5351, + "step": 2556 + }, + { + "epoch": 0.21193535018648985, + "grad_norm": 0.25304460525512695, + "learning_rate": 3.940737670949027e-06, + "loss": 0.5082, + "step": 2557 + }, + { + "epoch": 0.2120182345627849, + "grad_norm": 0.25769564509391785, + "learning_rate": 3.940323249067551e-06, + "loss": 0.5187, + "step": 2558 + }, + { + "epoch": 0.21210111893907999, + "grad_norm": 0.2598414719104767, + "learning_rate": 3.9399088271860755e-06, + "loss": 0.5378, + "step": 2559 + }, + { + "epoch": 0.21218400331537504, + "grad_norm": 0.2674967050552368, + "learning_rate": 3.9394944053046e-06, + "loss": 0.4927, + "step": 2560 + }, + { + "epoch": 0.21226688769167013, + "grad_norm": 0.25099509954452515, + "learning_rate": 3.939079983423125e-06, + "loss": 0.5139, + "step": 2561 + }, + { + "epoch": 0.21234977206796518, + "grad_norm": 0.28716427087783813, + "learning_rate": 3.93866556154165e-06, + "loss": 0.5468, + "step": 2562 + }, + { + "epoch": 0.21243265644426026, + "grad_norm": 0.2701047360897064, + "learning_rate": 3.938251139660174e-06, + "loss": 0.5242, + "step": 2563 + }, + { + "epoch": 0.21251554082055532, + "grad_norm": 0.2844003438949585, + "learning_rate": 3.937836717778699e-06, + "loss": 0.5635, + "step": 2564 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 0.27404874563217163, + "learning_rate": 3.937422295897224e-06, + "loss": 0.5255, + "step": 2565 + }, + { + "epoch": 0.21268130957314546, + "grad_norm": 0.28932681679725647, + "learning_rate": 3.937007874015748e-06, + "loss": 0.6078, + "step": 2566 + }, + { + "epoch": 0.21276419394944054, + "grad_norm": 0.25723397731781006, + "learning_rate": 3.936593452134273e-06, + "loss": 0.5241, + "step": 2567 + }, + { + "epoch": 0.2128470783257356, + "grad_norm": 0.2679454982280731, + "learning_rate": 3.936179030252797e-06, + "loss": 0.5638, + "step": 2568 + }, + { + "epoch": 0.21292996270203066, + "grad_norm": 0.25768837332725525, + "learning_rate": 3.935764608371322e-06, + "loss": 0.5583, + "step": 2569 + }, + { + "epoch": 0.21301284707832574, + "grad_norm": 0.26935136318206787, + "learning_rate": 3.935350186489847e-06, + "loss": 0.5423, + "step": 2570 + }, + { + "epoch": 0.2130957314546208, + "grad_norm": 0.27592939138412476, + "learning_rate": 3.934935764608372e-06, + "loss": 0.5557, + "step": 2571 + }, + { + "epoch": 0.21317861583091588, + "grad_norm": 0.26459500193595886, + "learning_rate": 3.9345213427268965e-06, + "loss": 0.5212, + "step": 2572 + }, + { + "epoch": 0.21326150020721094, + "grad_norm": 0.2630453109741211, + "learning_rate": 3.934106920845421e-06, + "loss": 0.5307, + "step": 2573 + }, + { + "epoch": 0.21334438458350602, + "grad_norm": 0.2635488212108612, + "learning_rate": 3.933692498963946e-06, + "loss": 0.5093, + "step": 2574 + }, + { + "epoch": 0.21342726895980108, + "grad_norm": 0.26963376998901367, + "learning_rate": 3.933278077082471e-06, + "loss": 0.5797, + "step": 2575 + }, + { + "epoch": 0.21351015333609616, + "grad_norm": 0.27429741621017456, + "learning_rate": 3.932863655200995e-06, + "loss": 0.5117, + "step": 2576 + }, + { + "epoch": 0.21359303771239121, + "grad_norm": 0.3030703067779541, + "learning_rate": 3.9324492333195195e-06, + "loss": 0.5608, + "step": 2577 + }, + { + "epoch": 0.21367592208868627, + "grad_norm": 0.26431068778038025, + "learning_rate": 3.932034811438044e-06, + "loss": 0.5171, + "step": 2578 + }, + { + "epoch": 0.21375880646498135, + "grad_norm": 0.2637794017791748, + "learning_rate": 3.931620389556569e-06, + "loss": 0.5426, + "step": 2579 + }, + { + "epoch": 0.2138416908412764, + "grad_norm": 0.2578475773334503, + "learning_rate": 3.931205967675094e-06, + "loss": 0.5522, + "step": 2580 + }, + { + "epoch": 0.2139245752175715, + "grad_norm": 0.27664536237716675, + "learning_rate": 3.930791545793618e-06, + "loss": 0.5967, + "step": 2581 + }, + { + "epoch": 0.21400745959386655, + "grad_norm": 0.2732493579387665, + "learning_rate": 3.930377123912143e-06, + "loss": 0.549, + "step": 2582 + }, + { + "epoch": 0.21409034397016163, + "grad_norm": 0.2846774160861969, + "learning_rate": 3.929962702030668e-06, + "loss": 0.5922, + "step": 2583 + }, + { + "epoch": 0.2141732283464567, + "grad_norm": 0.24901965260505676, + "learning_rate": 3.929548280149192e-06, + "loss": 0.531, + "step": 2584 + }, + { + "epoch": 0.21425611272275177, + "grad_norm": 0.2709961235523224, + "learning_rate": 3.929133858267717e-06, + "loss": 0.5252, + "step": 2585 + }, + { + "epoch": 0.21433899709904683, + "grad_norm": 0.2754879295825958, + "learning_rate": 3.928719436386241e-06, + "loss": 0.5165, + "step": 2586 + }, + { + "epoch": 0.21442188147534189, + "grad_norm": 0.2830513119697571, + "learning_rate": 3.928305014504766e-06, + "loss": 0.5475, + "step": 2587 + }, + { + "epoch": 0.21450476585163697, + "grad_norm": 0.25370845198631287, + "learning_rate": 3.927890592623291e-06, + "loss": 0.5229, + "step": 2588 + }, + { + "epoch": 0.21458765022793203, + "grad_norm": 0.2711149752140045, + "learning_rate": 3.927476170741816e-06, + "loss": 0.5564, + "step": 2589 + }, + { + "epoch": 0.2146705346042271, + "grad_norm": 0.265444815158844, + "learning_rate": 3.9270617488603405e-06, + "loss": 0.5354, + "step": 2590 + }, + { + "epoch": 0.21475341898052216, + "grad_norm": 0.26024335622787476, + "learning_rate": 3.926647326978865e-06, + "loss": 0.5369, + "step": 2591 + }, + { + "epoch": 0.21483630335681725, + "grad_norm": 0.2727898955345154, + "learning_rate": 3.926232905097389e-06, + "loss": 0.504, + "step": 2592 + }, + { + "epoch": 0.2149191877331123, + "grad_norm": 0.29059669375419617, + "learning_rate": 3.925818483215914e-06, + "loss": 0.5403, + "step": 2593 + }, + { + "epoch": 0.2150020721094074, + "grad_norm": 0.2784994840621948, + "learning_rate": 3.925404061334438e-06, + "loss": 0.5623, + "step": 2594 + }, + { + "epoch": 0.21508495648570244, + "grad_norm": 0.25942981243133545, + "learning_rate": 3.9249896394529635e-06, + "loss": 0.5278, + "step": 2595 + }, + { + "epoch": 0.2151678408619975, + "grad_norm": 0.2638733983039856, + "learning_rate": 3.924575217571488e-06, + "loss": 0.541, + "step": 2596 + }, + { + "epoch": 0.21525072523829258, + "grad_norm": 0.28710174560546875, + "learning_rate": 3.924160795690013e-06, + "loss": 0.4834, + "step": 2597 + }, + { + "epoch": 0.21533360961458764, + "grad_norm": 0.26114508509635925, + "learning_rate": 3.923746373808538e-06, + "loss": 0.5651, + "step": 2598 + }, + { + "epoch": 0.21541649399088272, + "grad_norm": 0.2934890687465668, + "learning_rate": 3.923331951927062e-06, + "loss": 0.5457, + "step": 2599 + }, + { + "epoch": 0.21549937836717778, + "grad_norm": 0.2641220688819885, + "learning_rate": 3.9229175300455865e-06, + "loss": 0.5422, + "step": 2600 + }, + { + "epoch": 0.21558226274347286, + "grad_norm": 0.24862083792686462, + "learning_rate": 3.922503108164111e-06, + "loss": 0.4988, + "step": 2601 + }, + { + "epoch": 0.21566514711976792, + "grad_norm": 0.25329336524009705, + "learning_rate": 3.922088686282636e-06, + "loss": 0.5145, + "step": 2602 + }, + { + "epoch": 0.215748031496063, + "grad_norm": 0.28707355260849, + "learning_rate": 3.921674264401161e-06, + "loss": 0.5419, + "step": 2603 + }, + { + "epoch": 0.21583091587235806, + "grad_norm": 0.2729772925376892, + "learning_rate": 3.921259842519685e-06, + "loss": 0.547, + "step": 2604 + }, + { + "epoch": 0.21591380024865314, + "grad_norm": 0.29446539282798767, + "learning_rate": 3.92084542063821e-06, + "loss": 0.5634, + "step": 2605 + }, + { + "epoch": 0.2159966846249482, + "grad_norm": 0.27126237750053406, + "learning_rate": 3.920430998756735e-06, + "loss": 0.5628, + "step": 2606 + }, + { + "epoch": 0.21607956900124325, + "grad_norm": 0.2597803771495819, + "learning_rate": 3.920016576875259e-06, + "loss": 0.5341, + "step": 2607 + }, + { + "epoch": 0.21616245337753834, + "grad_norm": 0.2540409564971924, + "learning_rate": 3.919602154993784e-06, + "loss": 0.4974, + "step": 2608 + }, + { + "epoch": 0.2162453377538334, + "grad_norm": 0.26261666417121887, + "learning_rate": 3.919187733112309e-06, + "loss": 0.5593, + "step": 2609 + }, + { + "epoch": 0.21632822213012848, + "grad_norm": 0.2592675983905792, + "learning_rate": 3.918773311230833e-06, + "loss": 0.5004, + "step": 2610 + }, + { + "epoch": 0.21641110650642353, + "grad_norm": 0.30306944251060486, + "learning_rate": 3.918358889349358e-06, + "loss": 0.5511, + "step": 2611 + }, + { + "epoch": 0.21649399088271862, + "grad_norm": 0.27880358695983887, + "learning_rate": 3.917944467467882e-06, + "loss": 0.5707, + "step": 2612 + }, + { + "epoch": 0.21657687525901367, + "grad_norm": 0.27023056149482727, + "learning_rate": 3.9175300455864075e-06, + "loss": 0.5304, + "step": 2613 + }, + { + "epoch": 0.21665975963530876, + "grad_norm": 0.27046501636505127, + "learning_rate": 3.917115623704932e-06, + "loss": 0.5345, + "step": 2614 + }, + { + "epoch": 0.2167426440116038, + "grad_norm": 0.2548779547214508, + "learning_rate": 3.916701201823457e-06, + "loss": 0.5344, + "step": 2615 + }, + { + "epoch": 0.21682552838789887, + "grad_norm": 0.25914546847343445, + "learning_rate": 3.916286779941982e-06, + "loss": 0.6016, + "step": 2616 + }, + { + "epoch": 0.21690841276419395, + "grad_norm": 0.2597179114818573, + "learning_rate": 3.915872358060506e-06, + "loss": 0.5396, + "step": 2617 + }, + { + "epoch": 0.216991297140489, + "grad_norm": 0.2662237882614136, + "learning_rate": 3.9154579361790305e-06, + "loss": 0.5922, + "step": 2618 + }, + { + "epoch": 0.2170741815167841, + "grad_norm": 0.2738972008228302, + "learning_rate": 3.915043514297555e-06, + "loss": 0.5746, + "step": 2619 + }, + { + "epoch": 0.21715706589307915, + "grad_norm": 0.2805578410625458, + "learning_rate": 3.91462909241608e-06, + "loss": 0.547, + "step": 2620 + }, + { + "epoch": 0.21723995026937423, + "grad_norm": 0.25334396958351135, + "learning_rate": 3.914214670534605e-06, + "loss": 0.5309, + "step": 2621 + }, + { + "epoch": 0.2173228346456693, + "grad_norm": 0.26623499393463135, + "learning_rate": 3.913800248653129e-06, + "loss": 0.6208, + "step": 2622 + }, + { + "epoch": 0.21740571902196437, + "grad_norm": 0.25748375058174133, + "learning_rate": 3.913385826771654e-06, + "loss": 0.5288, + "step": 2623 + }, + { + "epoch": 0.21748860339825943, + "grad_norm": 0.2554318904876709, + "learning_rate": 3.912971404890179e-06, + "loss": 0.5592, + "step": 2624 + }, + { + "epoch": 0.21757148777455448, + "grad_norm": 0.27944979071617126, + "learning_rate": 3.912556983008703e-06, + "loss": 0.5677, + "step": 2625 + }, + { + "epoch": 0.21765437215084957, + "grad_norm": 0.2654993534088135, + "learning_rate": 3.912142561127228e-06, + "loss": 0.5569, + "step": 2626 + }, + { + "epoch": 0.21773725652714462, + "grad_norm": 0.25087204575538635, + "learning_rate": 3.911728139245752e-06, + "loss": 0.5303, + "step": 2627 + }, + { + "epoch": 0.2178201409034397, + "grad_norm": 0.2673639953136444, + "learning_rate": 3.911313717364277e-06, + "loss": 0.5425, + "step": 2628 + }, + { + "epoch": 0.21790302527973476, + "grad_norm": 0.2749987542629242, + "learning_rate": 3.910899295482802e-06, + "loss": 0.5885, + "step": 2629 + }, + { + "epoch": 0.21798590965602985, + "grad_norm": 0.2642422914505005, + "learning_rate": 3.910484873601326e-06, + "loss": 0.5311, + "step": 2630 + }, + { + "epoch": 0.2180687940323249, + "grad_norm": 0.2857808470726013, + "learning_rate": 3.9100704517198515e-06, + "loss": 0.579, + "step": 2631 + }, + { + "epoch": 0.21815167840862, + "grad_norm": 0.2684347927570343, + "learning_rate": 3.909656029838376e-06, + "loss": 0.5168, + "step": 2632 + }, + { + "epoch": 0.21823456278491504, + "grad_norm": 0.2711782157421112, + "learning_rate": 3.9092416079569e-06, + "loss": 0.5787, + "step": 2633 + }, + { + "epoch": 0.2183174471612101, + "grad_norm": 0.28503113985061646, + "learning_rate": 3.908827186075425e-06, + "loss": 0.5624, + "step": 2634 + }, + { + "epoch": 0.21840033153750518, + "grad_norm": 0.28900814056396484, + "learning_rate": 3.908412764193949e-06, + "loss": 0.5413, + "step": 2635 + }, + { + "epoch": 0.21848321591380024, + "grad_norm": 0.282836377620697, + "learning_rate": 3.9079983423124745e-06, + "loss": 0.5382, + "step": 2636 + }, + { + "epoch": 0.21856610029009532, + "grad_norm": 0.287103533744812, + "learning_rate": 3.907583920430999e-06, + "loss": 0.5327, + "step": 2637 + }, + { + "epoch": 0.21864898466639038, + "grad_norm": 0.27589696645736694, + "learning_rate": 3.907169498549524e-06, + "loss": 0.5608, + "step": 2638 + }, + { + "epoch": 0.21873186904268546, + "grad_norm": 0.3212207555770874, + "learning_rate": 3.906755076668049e-06, + "loss": 0.5128, + "step": 2639 + }, + { + "epoch": 0.21881475341898052, + "grad_norm": 0.2754659354686737, + "learning_rate": 3.906340654786573e-06, + "loss": 0.5219, + "step": 2640 + }, + { + "epoch": 0.2188976377952756, + "grad_norm": 0.23941028118133545, + "learning_rate": 3.9059262329050975e-06, + "loss": 0.5126, + "step": 2641 + }, + { + "epoch": 0.21898052217157066, + "grad_norm": 0.25520890951156616, + "learning_rate": 3.905511811023622e-06, + "loss": 0.4717, + "step": 2642 + }, + { + "epoch": 0.21906340654786574, + "grad_norm": 0.27138468623161316, + "learning_rate": 3.905097389142146e-06, + "loss": 0.5361, + "step": 2643 + }, + { + "epoch": 0.2191462909241608, + "grad_norm": 0.2659902274608612, + "learning_rate": 3.904682967260672e-06, + "loss": 0.5288, + "step": 2644 + }, + { + "epoch": 0.21922917530045585, + "grad_norm": 0.28885889053344727, + "learning_rate": 3.904268545379196e-06, + "loss": 0.5552, + "step": 2645 + }, + { + "epoch": 0.21931205967675094, + "grad_norm": 0.25387755036354065, + "learning_rate": 3.903854123497721e-06, + "loss": 0.5126, + "step": 2646 + }, + { + "epoch": 0.219394944053046, + "grad_norm": 0.26352256536483765, + "learning_rate": 3.903439701616246e-06, + "loss": 0.5926, + "step": 2647 + }, + { + "epoch": 0.21947782842934108, + "grad_norm": 0.25489625334739685, + "learning_rate": 3.90302527973477e-06, + "loss": 0.494, + "step": 2648 + }, + { + "epoch": 0.21956071280563613, + "grad_norm": 0.28425946831703186, + "learning_rate": 3.902610857853295e-06, + "loss": 0.5399, + "step": 2649 + }, + { + "epoch": 0.21964359718193122, + "grad_norm": 0.28920626640319824, + "learning_rate": 3.90219643597182e-06, + "loss": 0.5543, + "step": 2650 + }, + { + "epoch": 0.21972648155822627, + "grad_norm": 0.29583925008773804, + "learning_rate": 3.901782014090344e-06, + "loss": 0.5811, + "step": 2651 + }, + { + "epoch": 0.21980936593452136, + "grad_norm": 0.25883176922798157, + "learning_rate": 3.901367592208869e-06, + "loss": 0.5735, + "step": 2652 + }, + { + "epoch": 0.2198922503108164, + "grad_norm": 0.25331202149391174, + "learning_rate": 3.900953170327393e-06, + "loss": 0.5065, + "step": 2653 + }, + { + "epoch": 0.21997513468711147, + "grad_norm": 0.27876946330070496, + "learning_rate": 3.9005387484459185e-06, + "loss": 0.548, + "step": 2654 + }, + { + "epoch": 0.22005801906340655, + "grad_norm": 0.26056015491485596, + "learning_rate": 3.900124326564443e-06, + "loss": 0.4999, + "step": 2655 + }, + { + "epoch": 0.2201409034397016, + "grad_norm": 0.24665869772434235, + "learning_rate": 3.899709904682968e-06, + "loss": 0.468, + "step": 2656 + }, + { + "epoch": 0.2202237878159967, + "grad_norm": 0.262578547000885, + "learning_rate": 3.899295482801493e-06, + "loss": 0.5161, + "step": 2657 + }, + { + "epoch": 0.22030667219229175, + "grad_norm": 0.2914399802684784, + "learning_rate": 3.898881060920017e-06, + "loss": 0.5582, + "step": 2658 + }, + { + "epoch": 0.22038955656858683, + "grad_norm": 0.26272493600845337, + "learning_rate": 3.8984666390385415e-06, + "loss": 0.5563, + "step": 2659 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 0.29051658511161804, + "learning_rate": 3.898052217157066e-06, + "loss": 0.5999, + "step": 2660 + }, + { + "epoch": 0.22055532532117697, + "grad_norm": 0.2805770933628082, + "learning_rate": 3.89763779527559e-06, + "loss": 0.5687, + "step": 2661 + }, + { + "epoch": 0.22063820969747203, + "grad_norm": 0.2641628682613373, + "learning_rate": 3.897223373394116e-06, + "loss": 0.5314, + "step": 2662 + }, + { + "epoch": 0.22072109407376708, + "grad_norm": 0.27540677785873413, + "learning_rate": 3.89680895151264e-06, + "loss": 0.5479, + "step": 2663 + }, + { + "epoch": 0.22080397845006217, + "grad_norm": 0.2610776126384735, + "learning_rate": 3.896394529631165e-06, + "loss": 0.5097, + "step": 2664 + }, + { + "epoch": 0.22088686282635722, + "grad_norm": 0.27889832854270935, + "learning_rate": 3.89598010774969e-06, + "loss": 0.569, + "step": 2665 + }, + { + "epoch": 0.2209697472026523, + "grad_norm": 0.2910457253456116, + "learning_rate": 3.895565685868214e-06, + "loss": 0.5652, + "step": 2666 + }, + { + "epoch": 0.22105263157894736, + "grad_norm": 0.28419679403305054, + "learning_rate": 3.895151263986739e-06, + "loss": 0.5374, + "step": 2667 + }, + { + "epoch": 0.22113551595524245, + "grad_norm": 0.259589821100235, + "learning_rate": 3.894736842105263e-06, + "loss": 0.5312, + "step": 2668 + }, + { + "epoch": 0.2212184003315375, + "grad_norm": 0.2553180754184723, + "learning_rate": 3.894322420223788e-06, + "loss": 0.5103, + "step": 2669 + }, + { + "epoch": 0.22130128470783259, + "grad_norm": 0.26218897104263306, + "learning_rate": 3.893907998342313e-06, + "loss": 0.5627, + "step": 2670 + }, + { + "epoch": 0.22138416908412764, + "grad_norm": 0.25868839025497437, + "learning_rate": 3.893493576460837e-06, + "loss": 0.538, + "step": 2671 + }, + { + "epoch": 0.2214670534604227, + "grad_norm": 0.27202916145324707, + "learning_rate": 3.8930791545793625e-06, + "loss": 0.5354, + "step": 2672 + }, + { + "epoch": 0.22154993783671778, + "grad_norm": 0.2574353516101837, + "learning_rate": 3.892664732697887e-06, + "loss": 0.4982, + "step": 2673 + }, + { + "epoch": 0.22163282221301284, + "grad_norm": 0.24903737008571625, + "learning_rate": 3.892250310816411e-06, + "loss": 0.5279, + "step": 2674 + }, + { + "epoch": 0.22171570658930792, + "grad_norm": 0.24727319180965424, + "learning_rate": 3.891835888934936e-06, + "loss": 0.4865, + "step": 2675 + }, + { + "epoch": 0.22179859096560298, + "grad_norm": 0.2989442050457001, + "learning_rate": 3.89142146705346e-06, + "loss": 0.5776, + "step": 2676 + }, + { + "epoch": 0.22188147534189806, + "grad_norm": 0.26281124353408813, + "learning_rate": 3.8910070451719855e-06, + "loss": 0.5513, + "step": 2677 + }, + { + "epoch": 0.22196435971819312, + "grad_norm": 0.2554683983325958, + "learning_rate": 3.89059262329051e-06, + "loss": 0.5641, + "step": 2678 + }, + { + "epoch": 0.2220472440944882, + "grad_norm": 0.27294251322746277, + "learning_rate": 3.890178201409034e-06, + "loss": 0.5738, + "step": 2679 + }, + { + "epoch": 0.22213012847078326, + "grad_norm": 0.24576205015182495, + "learning_rate": 3.88976377952756e-06, + "loss": 0.5117, + "step": 2680 + }, + { + "epoch": 0.22221301284707834, + "grad_norm": 0.270347535610199, + "learning_rate": 3.889349357646084e-06, + "loss": 0.5416, + "step": 2681 + }, + { + "epoch": 0.2222958972233734, + "grad_norm": 0.2380862981081009, + "learning_rate": 3.8889349357646085e-06, + "loss": 0.54, + "step": 2682 + }, + { + "epoch": 0.22237878159966845, + "grad_norm": 0.2745005190372467, + "learning_rate": 3.888520513883133e-06, + "loss": 0.5609, + "step": 2683 + }, + { + "epoch": 0.22246166597596354, + "grad_norm": 0.2707710564136505, + "learning_rate": 3.888106092001657e-06, + "loss": 0.5704, + "step": 2684 + }, + { + "epoch": 0.2225445503522586, + "grad_norm": 0.2760825753211975, + "learning_rate": 3.887691670120183e-06, + "loss": 0.5627, + "step": 2685 + }, + { + "epoch": 0.22262743472855367, + "grad_norm": 0.2311272919178009, + "learning_rate": 3.887277248238707e-06, + "loss": 0.488, + "step": 2686 + }, + { + "epoch": 0.22271031910484873, + "grad_norm": 0.26039132475852966, + "learning_rate": 3.886862826357232e-06, + "loss": 0.5212, + "step": 2687 + }, + { + "epoch": 0.22279320348114381, + "grad_norm": 0.24569302797317505, + "learning_rate": 3.886448404475757e-06, + "loss": 0.5369, + "step": 2688 + }, + { + "epoch": 0.22287608785743887, + "grad_norm": 0.2707560658454895, + "learning_rate": 3.886033982594281e-06, + "loss": 0.5414, + "step": 2689 + }, + { + "epoch": 0.22295897223373395, + "grad_norm": 0.2486061155796051, + "learning_rate": 3.8856195607128065e-06, + "loss": 0.5069, + "step": 2690 + }, + { + "epoch": 0.223041856610029, + "grad_norm": 0.24381959438323975, + "learning_rate": 3.885205138831331e-06, + "loss": 0.4903, + "step": 2691 + }, + { + "epoch": 0.22312474098632407, + "grad_norm": 0.2686889171600342, + "learning_rate": 3.884790716949855e-06, + "loss": 0.5456, + "step": 2692 + }, + { + "epoch": 0.22320762536261915, + "grad_norm": 0.2641972601413727, + "learning_rate": 3.88437629506838e-06, + "loss": 0.5133, + "step": 2693 + }, + { + "epoch": 0.2232905097389142, + "grad_norm": 0.2704038918018341, + "learning_rate": 3.883961873186904e-06, + "loss": 0.5225, + "step": 2694 + }, + { + "epoch": 0.2233733941152093, + "grad_norm": 0.27001914381980896, + "learning_rate": 3.8835474513054295e-06, + "loss": 0.5537, + "step": 2695 + }, + { + "epoch": 0.22345627849150435, + "grad_norm": 0.26446279883384705, + "learning_rate": 3.883133029423954e-06, + "loss": 0.5437, + "step": 2696 + }, + { + "epoch": 0.22353916286779943, + "grad_norm": 0.2608036398887634, + "learning_rate": 3.882718607542478e-06, + "loss": 0.503, + "step": 2697 + }, + { + "epoch": 0.22362204724409449, + "grad_norm": 0.2751135230064392, + "learning_rate": 3.882304185661004e-06, + "loss": 0.5368, + "step": 2698 + }, + { + "epoch": 0.22370493162038957, + "grad_norm": 0.2594079077243805, + "learning_rate": 3.881889763779528e-06, + "loss": 0.5428, + "step": 2699 + }, + { + "epoch": 0.22378781599668462, + "grad_norm": 0.27205684781074524, + "learning_rate": 3.8814753418980525e-06, + "loss": 0.5815, + "step": 2700 + }, + { + "epoch": 0.22387070037297968, + "grad_norm": 0.26168036460876465, + "learning_rate": 3.881060920016577e-06, + "loss": 0.5541, + "step": 2701 + }, + { + "epoch": 0.22395358474927476, + "grad_norm": 0.2580362558364868, + "learning_rate": 3.880646498135101e-06, + "loss": 0.5183, + "step": 2702 + }, + { + "epoch": 0.22403646912556982, + "grad_norm": 0.24709725379943848, + "learning_rate": 3.880232076253627e-06, + "loss": 0.5213, + "step": 2703 + }, + { + "epoch": 0.2241193535018649, + "grad_norm": 0.2559225857257843, + "learning_rate": 3.879817654372151e-06, + "loss": 0.5756, + "step": 2704 + }, + { + "epoch": 0.22420223787815996, + "grad_norm": 0.25577399134635925, + "learning_rate": 3.879403232490676e-06, + "loss": 0.5382, + "step": 2705 + }, + { + "epoch": 0.22428512225445504, + "grad_norm": 0.2686783969402313, + "learning_rate": 3.878988810609201e-06, + "loss": 0.5646, + "step": 2706 + }, + { + "epoch": 0.2243680066307501, + "grad_norm": 0.2888002097606659, + "learning_rate": 3.878574388727725e-06, + "loss": 0.5558, + "step": 2707 + }, + { + "epoch": 0.22445089100704518, + "grad_norm": 0.26921209692955017, + "learning_rate": 3.87815996684625e-06, + "loss": 0.5544, + "step": 2708 + }, + { + "epoch": 0.22453377538334024, + "grad_norm": 0.2539839744567871, + "learning_rate": 3.877745544964774e-06, + "loss": 0.4695, + "step": 2709 + }, + { + "epoch": 0.2246166597596353, + "grad_norm": 0.27223071455955505, + "learning_rate": 3.8773311230832985e-06, + "loss": 0.4935, + "step": 2710 + }, + { + "epoch": 0.22469954413593038, + "grad_norm": 0.286913126707077, + "learning_rate": 3.876916701201824e-06, + "loss": 0.5306, + "step": 2711 + }, + { + "epoch": 0.22478242851222544, + "grad_norm": 0.25999873876571655, + "learning_rate": 3.876502279320348e-06, + "loss": 0.5513, + "step": 2712 + }, + { + "epoch": 0.22486531288852052, + "grad_norm": 0.264706552028656, + "learning_rate": 3.8760878574388735e-06, + "loss": 0.5371, + "step": 2713 + }, + { + "epoch": 0.22494819726481557, + "grad_norm": 0.2495431751012802, + "learning_rate": 3.875673435557398e-06, + "loss": 0.4906, + "step": 2714 + }, + { + "epoch": 0.22503108164111066, + "grad_norm": 0.2674497961997986, + "learning_rate": 3.875259013675922e-06, + "loss": 0.5522, + "step": 2715 + }, + { + "epoch": 0.22511396601740571, + "grad_norm": 0.27083051204681396, + "learning_rate": 3.874844591794447e-06, + "loss": 0.5606, + "step": 2716 + }, + { + "epoch": 0.2251968503937008, + "grad_norm": 0.2800511121749878, + "learning_rate": 3.874430169912971e-06, + "loss": 0.5609, + "step": 2717 + }, + { + "epoch": 0.22527973476999585, + "grad_norm": 0.2864183485507965, + "learning_rate": 3.8740157480314965e-06, + "loss": 0.5516, + "step": 2718 + }, + { + "epoch": 0.22536261914629094, + "grad_norm": 0.2619338035583496, + "learning_rate": 3.873601326150021e-06, + "loss": 0.5013, + "step": 2719 + }, + { + "epoch": 0.225445503522586, + "grad_norm": 0.28276312351226807, + "learning_rate": 3.873186904268545e-06, + "loss": 0.5611, + "step": 2720 + }, + { + "epoch": 0.22552838789888105, + "grad_norm": 0.27396249771118164, + "learning_rate": 3.872772482387071e-06, + "loss": 0.5293, + "step": 2721 + }, + { + "epoch": 0.22561127227517613, + "grad_norm": 0.24853917956352234, + "learning_rate": 3.872358060505595e-06, + "loss": 0.5195, + "step": 2722 + }, + { + "epoch": 0.2256941566514712, + "grad_norm": 0.28349313139915466, + "learning_rate": 3.8719436386241195e-06, + "loss": 0.5306, + "step": 2723 + }, + { + "epoch": 0.22577704102776627, + "grad_norm": 0.2755036950111389, + "learning_rate": 3.871529216742644e-06, + "loss": 0.5259, + "step": 2724 + }, + { + "epoch": 0.22585992540406133, + "grad_norm": 0.2622583210468292, + "learning_rate": 3.871114794861168e-06, + "loss": 0.5352, + "step": 2725 + }, + { + "epoch": 0.2259428097803564, + "grad_norm": 0.2743228077888489, + "learning_rate": 3.870700372979694e-06, + "loss": 0.5273, + "step": 2726 + }, + { + "epoch": 0.22602569415665147, + "grad_norm": 0.3026506006717682, + "learning_rate": 3.870285951098218e-06, + "loss": 0.5876, + "step": 2727 + }, + { + "epoch": 0.22610857853294655, + "grad_norm": 0.2744571566581726, + "learning_rate": 3.8698715292167426e-06, + "loss": 0.5728, + "step": 2728 + }, + { + "epoch": 0.2261914629092416, + "grad_norm": 0.2624276578426361, + "learning_rate": 3.869457107335268e-06, + "loss": 0.5524, + "step": 2729 + }, + { + "epoch": 0.22627434728553666, + "grad_norm": 0.2625750005245209, + "learning_rate": 3.869042685453792e-06, + "loss": 0.5684, + "step": 2730 + }, + { + "epoch": 0.22635723166183175, + "grad_norm": 0.26662617921829224, + "learning_rate": 3.8686282635723175e-06, + "loss": 0.5395, + "step": 2731 + }, + { + "epoch": 0.2264401160381268, + "grad_norm": 0.292371928691864, + "learning_rate": 3.868213841690842e-06, + "loss": 0.5939, + "step": 2732 + }, + { + "epoch": 0.2265230004144219, + "grad_norm": 0.33414626121520996, + "learning_rate": 3.867799419809366e-06, + "loss": 0.5835, + "step": 2733 + }, + { + "epoch": 0.22660588479071694, + "grad_norm": 0.29259616136550903, + "learning_rate": 3.867384997927891e-06, + "loss": 0.5594, + "step": 2734 + }, + { + "epoch": 0.22668876916701203, + "grad_norm": 0.314202219247818, + "learning_rate": 3.866970576046415e-06, + "loss": 0.5179, + "step": 2735 + }, + { + "epoch": 0.22677165354330708, + "grad_norm": 0.275147408246994, + "learning_rate": 3.8665561541649405e-06, + "loss": 0.5457, + "step": 2736 + }, + { + "epoch": 0.22685453791960217, + "grad_norm": 0.30637672543525696, + "learning_rate": 3.866141732283465e-06, + "loss": 0.5717, + "step": 2737 + }, + { + "epoch": 0.22693742229589722, + "grad_norm": 0.2988981008529663, + "learning_rate": 3.865727310401989e-06, + "loss": 0.5401, + "step": 2738 + }, + { + "epoch": 0.22702030667219228, + "grad_norm": 0.2821771502494812, + "learning_rate": 3.865312888520515e-06, + "loss": 0.544, + "step": 2739 + }, + { + "epoch": 0.22710319104848736, + "grad_norm": 0.2989565432071686, + "learning_rate": 3.864898466639039e-06, + "loss": 0.529, + "step": 2740 + }, + { + "epoch": 0.22718607542478242, + "grad_norm": 0.29354801774024963, + "learning_rate": 3.8644840447575636e-06, + "loss": 0.5309, + "step": 2741 + }, + { + "epoch": 0.2272689598010775, + "grad_norm": 0.26170551776885986, + "learning_rate": 3.864069622876088e-06, + "loss": 0.5246, + "step": 2742 + }, + { + "epoch": 0.22735184417737256, + "grad_norm": 0.26416274905204773, + "learning_rate": 3.863655200994612e-06, + "loss": 0.5065, + "step": 2743 + }, + { + "epoch": 0.22743472855366764, + "grad_norm": 0.2792912423610687, + "learning_rate": 3.863240779113138e-06, + "loss": 0.5808, + "step": 2744 + }, + { + "epoch": 0.2275176129299627, + "grad_norm": 0.26626503467559814, + "learning_rate": 3.862826357231662e-06, + "loss": 0.5717, + "step": 2745 + }, + { + "epoch": 0.22760049730625778, + "grad_norm": 0.26951631903648376, + "learning_rate": 3.8624119353501866e-06, + "loss": 0.5204, + "step": 2746 + }, + { + "epoch": 0.22768338168255284, + "grad_norm": 0.30946439504623413, + "learning_rate": 3.861997513468712e-06, + "loss": 0.5831, + "step": 2747 + }, + { + "epoch": 0.2277662660588479, + "grad_norm": 0.2830395996570587, + "learning_rate": 3.861583091587236e-06, + "loss": 0.5182, + "step": 2748 + }, + { + "epoch": 0.22784915043514298, + "grad_norm": 0.2654151916503906, + "learning_rate": 3.861168669705761e-06, + "loss": 0.558, + "step": 2749 + }, + { + "epoch": 0.22793203481143803, + "grad_norm": 1.245630145072937, + "learning_rate": 3.860754247824285e-06, + "loss": 0.5519, + "step": 2750 + }, + { + "epoch": 0.22801491918773312, + "grad_norm": 0.26981839537620544, + "learning_rate": 3.8603398259428096e-06, + "loss": 0.5319, + "step": 2751 + }, + { + "epoch": 0.22809780356402817, + "grad_norm": 0.26828497648239136, + "learning_rate": 3.859925404061335e-06, + "loss": 0.516, + "step": 2752 + }, + { + "epoch": 0.22818068794032326, + "grad_norm": 0.26065593957901, + "learning_rate": 3.859510982179859e-06, + "loss": 0.4972, + "step": 2753 + }, + { + "epoch": 0.2282635723166183, + "grad_norm": 0.2560385763645172, + "learning_rate": 3.8590965602983846e-06, + "loss": 0.5123, + "step": 2754 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 0.28467750549316406, + "learning_rate": 3.858682138416909e-06, + "loss": 0.5809, + "step": 2755 + }, + { + "epoch": 0.22842934106920845, + "grad_norm": 0.24604938924312592, + "learning_rate": 3.858267716535433e-06, + "loss": 0.5074, + "step": 2756 + }, + { + "epoch": 0.22851222544550354, + "grad_norm": 0.26047515869140625, + "learning_rate": 3.857853294653958e-06, + "loss": 0.5343, + "step": 2757 + }, + { + "epoch": 0.2285951098217986, + "grad_norm": 0.2639378309249878, + "learning_rate": 3.857438872772482e-06, + "loss": 0.5548, + "step": 2758 + }, + { + "epoch": 0.22867799419809365, + "grad_norm": 0.27490198612213135, + "learning_rate": 3.8570244508910076e-06, + "loss": 0.5131, + "step": 2759 + }, + { + "epoch": 0.22876087857438873, + "grad_norm": 0.25567618012428284, + "learning_rate": 3.856610029009532e-06, + "loss": 0.5094, + "step": 2760 + }, + { + "epoch": 0.2288437629506838, + "grad_norm": 0.27032604813575745, + "learning_rate": 3.8561956071280564e-06, + "loss": 0.5665, + "step": 2761 + }, + { + "epoch": 0.22892664732697887, + "grad_norm": 0.26045534014701843, + "learning_rate": 3.855781185246582e-06, + "loss": 0.5711, + "step": 2762 + }, + { + "epoch": 0.22900953170327393, + "grad_norm": 0.2647976577281952, + "learning_rate": 3.855366763365106e-06, + "loss": 0.5451, + "step": 2763 + }, + { + "epoch": 0.229092416079569, + "grad_norm": 0.28584858775138855, + "learning_rate": 3.8549523414836306e-06, + "loss": 0.5546, + "step": 2764 + }, + { + "epoch": 0.22917530045586407, + "grad_norm": 0.2629980742931366, + "learning_rate": 3.854537919602155e-06, + "loss": 0.5327, + "step": 2765 + }, + { + "epoch": 0.22925818483215915, + "grad_norm": 0.27569541335105896, + "learning_rate": 3.8541234977206794e-06, + "loss": 0.5451, + "step": 2766 + }, + { + "epoch": 0.2293410692084542, + "grad_norm": 0.2522553503513336, + "learning_rate": 3.853709075839205e-06, + "loss": 0.5493, + "step": 2767 + }, + { + "epoch": 0.22942395358474926, + "grad_norm": 0.2544832229614258, + "learning_rate": 3.853294653957729e-06, + "loss": 0.5601, + "step": 2768 + }, + { + "epoch": 0.22950683796104435, + "grad_norm": 0.25272324681282043, + "learning_rate": 3.8528802320762536e-06, + "loss": 0.5181, + "step": 2769 + }, + { + "epoch": 0.2295897223373394, + "grad_norm": 0.2891126573085785, + "learning_rate": 3.852465810194779e-06, + "loss": 0.5845, + "step": 2770 + }, + { + "epoch": 0.2296726067136345, + "grad_norm": 0.27116334438323975, + "learning_rate": 3.852051388313303e-06, + "loss": 0.5298, + "step": 2771 + }, + { + "epoch": 0.22975549108992954, + "grad_norm": 0.28947409987449646, + "learning_rate": 3.8516369664318286e-06, + "loss": 0.5623, + "step": 2772 + }, + { + "epoch": 0.22983837546622463, + "grad_norm": 0.2881045341491699, + "learning_rate": 3.851222544550353e-06, + "loss": 0.5648, + "step": 2773 + }, + { + "epoch": 0.22992125984251968, + "grad_norm": 0.2978197932243347, + "learning_rate": 3.8508081226688774e-06, + "loss": 0.5651, + "step": 2774 + }, + { + "epoch": 0.23000414421881477, + "grad_norm": 0.287738561630249, + "learning_rate": 3.850393700787402e-06, + "loss": 0.5326, + "step": 2775 + }, + { + "epoch": 0.23008702859510982, + "grad_norm": 0.27225908637046814, + "learning_rate": 3.849979278905926e-06, + "loss": 0.5521, + "step": 2776 + }, + { + "epoch": 0.23016991297140488, + "grad_norm": 0.2601177990436554, + "learning_rate": 3.849564857024451e-06, + "loss": 0.5119, + "step": 2777 + }, + { + "epoch": 0.23025279734769996, + "grad_norm": 0.2734961211681366, + "learning_rate": 3.849150435142976e-06, + "loss": 0.5737, + "step": 2778 + }, + { + "epoch": 0.23033568172399502, + "grad_norm": 0.2828013598918915, + "learning_rate": 3.8487360132615004e-06, + "loss": 0.5551, + "step": 2779 + }, + { + "epoch": 0.2304185661002901, + "grad_norm": 0.26250556111335754, + "learning_rate": 3.848321591380026e-06, + "loss": 0.5126, + "step": 2780 + }, + { + "epoch": 0.23050145047658516, + "grad_norm": 0.28368109464645386, + "learning_rate": 3.84790716949855e-06, + "loss": 0.5463, + "step": 2781 + }, + { + "epoch": 0.23058433485288024, + "grad_norm": 0.3037853240966797, + "learning_rate": 3.8474927476170746e-06, + "loss": 0.569, + "step": 2782 + }, + { + "epoch": 0.2306672192291753, + "grad_norm": 0.2652021050453186, + "learning_rate": 3.847078325735599e-06, + "loss": 0.5508, + "step": 2783 + }, + { + "epoch": 0.23075010360547038, + "grad_norm": 0.25332388281822205, + "learning_rate": 3.8466639038541234e-06, + "loss": 0.5317, + "step": 2784 + }, + { + "epoch": 0.23083298798176544, + "grad_norm": 0.2498004287481308, + "learning_rate": 3.846249481972649e-06, + "loss": 0.5099, + "step": 2785 + }, + { + "epoch": 0.2309158723580605, + "grad_norm": 0.2928227186203003, + "learning_rate": 3.845835060091173e-06, + "loss": 0.5375, + "step": 2786 + }, + { + "epoch": 0.23099875673435558, + "grad_norm": 0.26153895258903503, + "learning_rate": 3.845420638209698e-06, + "loss": 0.5267, + "step": 2787 + }, + { + "epoch": 0.23108164111065063, + "grad_norm": 0.2693178653717041, + "learning_rate": 3.845006216328223e-06, + "loss": 0.5609, + "step": 2788 + }, + { + "epoch": 0.23116452548694572, + "grad_norm": 0.27864208817481995, + "learning_rate": 3.844591794446747e-06, + "loss": 0.5815, + "step": 2789 + }, + { + "epoch": 0.23124740986324077, + "grad_norm": 0.24864749610424042, + "learning_rate": 3.844177372565272e-06, + "loss": 0.4691, + "step": 2790 + }, + { + "epoch": 0.23133029423953586, + "grad_norm": 0.29577627778053284, + "learning_rate": 3.843762950683796e-06, + "loss": 0.5655, + "step": 2791 + }, + { + "epoch": 0.2314131786158309, + "grad_norm": 0.28154247999191284, + "learning_rate": 3.843348528802321e-06, + "loss": 0.5534, + "step": 2792 + }, + { + "epoch": 0.231496062992126, + "grad_norm": 0.27039608359336853, + "learning_rate": 3.842934106920846e-06, + "loss": 0.5433, + "step": 2793 + }, + { + "epoch": 0.23157894736842105, + "grad_norm": 0.2651430070400238, + "learning_rate": 3.84251968503937e-06, + "loss": 0.5668, + "step": 2794 + }, + { + "epoch": 0.23166183174471613, + "grad_norm": 0.2480887621641159, + "learning_rate": 3.842105263157895e-06, + "loss": 0.4886, + "step": 2795 + }, + { + "epoch": 0.2317447161210112, + "grad_norm": 0.27034813165664673, + "learning_rate": 3.84169084127642e-06, + "loss": 0.5472, + "step": 2796 + }, + { + "epoch": 0.23182760049730625, + "grad_norm": 0.26590052247047424, + "learning_rate": 3.8412764193949444e-06, + "loss": 0.5758, + "step": 2797 + }, + { + "epoch": 0.23191048487360133, + "grad_norm": 0.2658424973487854, + "learning_rate": 3.840861997513469e-06, + "loss": 0.5275, + "step": 2798 + }, + { + "epoch": 0.2319933692498964, + "grad_norm": 0.29328763484954834, + "learning_rate": 3.840447575631993e-06, + "loss": 0.5274, + "step": 2799 + }, + { + "epoch": 0.23207625362619147, + "grad_norm": 0.2530929148197174, + "learning_rate": 3.840033153750518e-06, + "loss": 0.5366, + "step": 2800 + }, + { + "epoch": 0.23215913800248653, + "grad_norm": 0.25983092188835144, + "learning_rate": 3.839618731869043e-06, + "loss": 0.4974, + "step": 2801 + }, + { + "epoch": 0.2322420223787816, + "grad_norm": 0.2672503888607025, + "learning_rate": 3.8392043099875674e-06, + "loss": 0.538, + "step": 2802 + }, + { + "epoch": 0.23232490675507667, + "grad_norm": 0.2779373228549957, + "learning_rate": 3.838789888106093e-06, + "loss": 0.5286, + "step": 2803 + }, + { + "epoch": 0.23240779113137175, + "grad_norm": 0.2463812679052353, + "learning_rate": 3.838375466224617e-06, + "loss": 0.538, + "step": 2804 + }, + { + "epoch": 0.2324906755076668, + "grad_norm": 0.2624935805797577, + "learning_rate": 3.837961044343142e-06, + "loss": 0.5174, + "step": 2805 + }, + { + "epoch": 0.23257355988396186, + "grad_norm": 0.244076207280159, + "learning_rate": 3.837546622461666e-06, + "loss": 0.5271, + "step": 2806 + }, + { + "epoch": 0.23265644426025694, + "grad_norm": 0.27701687812805176, + "learning_rate": 3.8371322005801905e-06, + "loss": 0.563, + "step": 2807 + }, + { + "epoch": 0.232739328636552, + "grad_norm": 0.26544979214668274, + "learning_rate": 3.836717778698716e-06, + "loss": 0.5064, + "step": 2808 + }, + { + "epoch": 0.23282221301284708, + "grad_norm": 0.26510968804359436, + "learning_rate": 3.83630335681724e-06, + "loss": 0.5239, + "step": 2809 + }, + { + "epoch": 0.23290509738914214, + "grad_norm": 0.26427164673805237, + "learning_rate": 3.835888934935765e-06, + "loss": 0.5627, + "step": 2810 + }, + { + "epoch": 0.23298798176543722, + "grad_norm": 0.2552003860473633, + "learning_rate": 3.83547451305429e-06, + "loss": 0.5065, + "step": 2811 + }, + { + "epoch": 0.23307086614173228, + "grad_norm": 0.25704944133758545, + "learning_rate": 3.835060091172814e-06, + "loss": 0.4889, + "step": 2812 + }, + { + "epoch": 0.23315375051802736, + "grad_norm": 0.2559053301811218, + "learning_rate": 3.834645669291339e-06, + "loss": 0.5426, + "step": 2813 + }, + { + "epoch": 0.23323663489432242, + "grad_norm": 0.2538028359413147, + "learning_rate": 3.834231247409864e-06, + "loss": 0.5372, + "step": 2814 + }, + { + "epoch": 0.23331951927061748, + "grad_norm": 0.27061137557029724, + "learning_rate": 3.8338168255283884e-06, + "loss": 0.5225, + "step": 2815 + }, + { + "epoch": 0.23340240364691256, + "grad_norm": 0.2674802541732788, + "learning_rate": 3.833402403646913e-06, + "loss": 0.5506, + "step": 2816 + }, + { + "epoch": 0.23348528802320762, + "grad_norm": 0.2778453826904297, + "learning_rate": 3.832987981765437e-06, + "loss": 0.549, + "step": 2817 + }, + { + "epoch": 0.2335681723995027, + "grad_norm": 0.2739362120628357, + "learning_rate": 3.832573559883962e-06, + "loss": 0.5649, + "step": 2818 + }, + { + "epoch": 0.23365105677579776, + "grad_norm": 0.303842693567276, + "learning_rate": 3.832159138002487e-06, + "loss": 0.57, + "step": 2819 + }, + { + "epoch": 0.23373394115209284, + "grad_norm": 0.27529624104499817, + "learning_rate": 3.8317447161210115e-06, + "loss": 0.5301, + "step": 2820 + }, + { + "epoch": 0.2338168255283879, + "grad_norm": 0.25018540024757385, + "learning_rate": 3.831330294239537e-06, + "loss": 0.5205, + "step": 2821 + }, + { + "epoch": 0.23389970990468298, + "grad_norm": 0.26158878207206726, + "learning_rate": 3.830915872358061e-06, + "loss": 0.5325, + "step": 2822 + }, + { + "epoch": 0.23398259428097803, + "grad_norm": 0.27092546224594116, + "learning_rate": 3.830501450476586e-06, + "loss": 0.5526, + "step": 2823 + }, + { + "epoch": 0.2340654786572731, + "grad_norm": 0.25124770402908325, + "learning_rate": 3.83008702859511e-06, + "loss": 0.4799, + "step": 2824 + }, + { + "epoch": 0.23414836303356817, + "grad_norm": 0.26681599020957947, + "learning_rate": 3.8296726067136345e-06, + "loss": 0.5637, + "step": 2825 + }, + { + "epoch": 0.23423124740986323, + "grad_norm": 0.27541834115982056, + "learning_rate": 3.82925818483216e-06, + "loss": 0.5511, + "step": 2826 + }, + { + "epoch": 0.23431413178615831, + "grad_norm": 0.2631833255290985, + "learning_rate": 3.828843762950684e-06, + "loss": 0.5346, + "step": 2827 + }, + { + "epoch": 0.23439701616245337, + "grad_norm": 0.27299103140830994, + "learning_rate": 3.828429341069209e-06, + "loss": 0.5511, + "step": 2828 + }, + { + "epoch": 0.23447990053874845, + "grad_norm": 0.246548131108284, + "learning_rate": 3.828014919187734e-06, + "loss": 0.5369, + "step": 2829 + }, + { + "epoch": 0.2345627849150435, + "grad_norm": 0.24237950146198273, + "learning_rate": 3.827600497306258e-06, + "loss": 0.5191, + "step": 2830 + }, + { + "epoch": 0.2346456692913386, + "grad_norm": 0.28160926699638367, + "learning_rate": 3.827186075424783e-06, + "loss": 0.5555, + "step": 2831 + }, + { + "epoch": 0.23472855366763365, + "grad_norm": 0.29467475414276123, + "learning_rate": 3.826771653543307e-06, + "loss": 0.5682, + "step": 2832 + }, + { + "epoch": 0.23481143804392873, + "grad_norm": 0.254189133644104, + "learning_rate": 3.826357231661832e-06, + "loss": 0.5211, + "step": 2833 + }, + { + "epoch": 0.2348943224202238, + "grad_norm": 0.24191468954086304, + "learning_rate": 3.825942809780357e-06, + "loss": 0.5258, + "step": 2834 + }, + { + "epoch": 0.23497720679651884, + "grad_norm": 0.2612690329551697, + "learning_rate": 3.825528387898881e-06, + "loss": 0.51, + "step": 2835 + }, + { + "epoch": 0.23506009117281393, + "grad_norm": 0.25914105772972107, + "learning_rate": 3.825113966017406e-06, + "loss": 0.4762, + "step": 2836 + }, + { + "epoch": 0.23514297554910898, + "grad_norm": 0.26072949171066284, + "learning_rate": 3.824699544135931e-06, + "loss": 0.5523, + "step": 2837 + }, + { + "epoch": 0.23522585992540407, + "grad_norm": 0.2787131667137146, + "learning_rate": 3.8242851222544555e-06, + "loss": 0.5063, + "step": 2838 + }, + { + "epoch": 0.23530874430169912, + "grad_norm": 0.2695331573486328, + "learning_rate": 3.82387070037298e-06, + "loss": 0.547, + "step": 2839 + }, + { + "epoch": 0.2353916286779942, + "grad_norm": 0.2595630884170532, + "learning_rate": 3.823456278491504e-06, + "loss": 0.5662, + "step": 2840 + }, + { + "epoch": 0.23547451305428926, + "grad_norm": 0.26429247856140137, + "learning_rate": 3.823041856610029e-06, + "loss": 0.519, + "step": 2841 + }, + { + "epoch": 0.23555739743058435, + "grad_norm": 0.25085410475730896, + "learning_rate": 3.822627434728554e-06, + "loss": 0.5281, + "step": 2842 + }, + { + "epoch": 0.2356402818068794, + "grad_norm": 0.24551205337047577, + "learning_rate": 3.8222130128470785e-06, + "loss": 0.5174, + "step": 2843 + }, + { + "epoch": 0.23572316618317446, + "grad_norm": 0.26152393221855164, + "learning_rate": 3.821798590965604e-06, + "loss": 0.5233, + "step": 2844 + }, + { + "epoch": 0.23580605055946954, + "grad_norm": 0.25988054275512695, + "learning_rate": 3.821384169084128e-06, + "loss": 0.565, + "step": 2845 + }, + { + "epoch": 0.2358889349357646, + "grad_norm": 0.276864618062973, + "learning_rate": 3.820969747202653e-06, + "loss": 0.5132, + "step": 2846 + }, + { + "epoch": 0.23597181931205968, + "grad_norm": 0.256973534822464, + "learning_rate": 3.820555325321177e-06, + "loss": 0.532, + "step": 2847 + }, + { + "epoch": 0.23605470368835474, + "grad_norm": 0.2848585546016693, + "learning_rate": 3.820140903439702e-06, + "loss": 0.5421, + "step": 2848 + }, + { + "epoch": 0.23613758806464982, + "grad_norm": 0.2596903443336487, + "learning_rate": 3.819726481558227e-06, + "loss": 0.5594, + "step": 2849 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 0.2688683867454529, + "learning_rate": 3.819312059676751e-06, + "loss": 0.5255, + "step": 2850 + }, + { + "epoch": 0.23630335681723996, + "grad_norm": 0.2480214238166809, + "learning_rate": 3.818897637795276e-06, + "loss": 0.553, + "step": 2851 + }, + { + "epoch": 0.23638624119353502, + "grad_norm": 0.24191009998321533, + "learning_rate": 3.818483215913801e-06, + "loss": 0.4989, + "step": 2852 + }, + { + "epoch": 0.23646912556983007, + "grad_norm": 0.2575129270553589, + "learning_rate": 3.818068794032325e-06, + "loss": 0.4965, + "step": 2853 + }, + { + "epoch": 0.23655200994612516, + "grad_norm": 0.2618384063243866, + "learning_rate": 3.81765437215085e-06, + "loss": 0.5529, + "step": 2854 + }, + { + "epoch": 0.23663489432242021, + "grad_norm": 0.272588312625885, + "learning_rate": 3.817239950269375e-06, + "loss": 0.5956, + "step": 2855 + }, + { + "epoch": 0.2367177786987153, + "grad_norm": 0.24813981354236603, + "learning_rate": 3.8168255283878995e-06, + "loss": 0.5139, + "step": 2856 + }, + { + "epoch": 0.23680066307501035, + "grad_norm": 0.2660761773586273, + "learning_rate": 3.816411106506424e-06, + "loss": 0.5358, + "step": 2857 + }, + { + "epoch": 0.23688354745130544, + "grad_norm": 0.24960511922836304, + "learning_rate": 3.815996684624948e-06, + "loss": 0.5132, + "step": 2858 + }, + { + "epoch": 0.2369664318276005, + "grad_norm": 0.26994797587394714, + "learning_rate": 3.815582262743473e-06, + "loss": 0.5468, + "step": 2859 + }, + { + "epoch": 0.23704931620389558, + "grad_norm": 0.26443982124328613, + "learning_rate": 3.815167840861998e-06, + "loss": 0.5372, + "step": 2860 + }, + { + "epoch": 0.23713220058019063, + "grad_norm": 0.26946836709976196, + "learning_rate": 3.8147534189805225e-06, + "loss": 0.5409, + "step": 2861 + }, + { + "epoch": 0.2372150849564857, + "grad_norm": 0.26311051845550537, + "learning_rate": 3.814338997099047e-06, + "loss": 0.564, + "step": 2862 + }, + { + "epoch": 0.23729796933278077, + "grad_norm": 0.2658202052116394, + "learning_rate": 3.8139245752175718e-06, + "loss": 0.5481, + "step": 2863 + }, + { + "epoch": 0.23738085370907583, + "grad_norm": 0.26144149899482727, + "learning_rate": 3.813510153336096e-06, + "loss": 0.5231, + "step": 2864 + }, + { + "epoch": 0.2374637380853709, + "grad_norm": 0.2766946852207184, + "learning_rate": 3.8130957314546215e-06, + "loss": 0.5321, + "step": 2865 + }, + { + "epoch": 0.23754662246166597, + "grad_norm": 0.26580819487571716, + "learning_rate": 3.812681309573146e-06, + "loss": 0.5719, + "step": 2866 + }, + { + "epoch": 0.23762950683796105, + "grad_norm": 0.2608400583267212, + "learning_rate": 3.8122668876916703e-06, + "loss": 0.5427, + "step": 2867 + }, + { + "epoch": 0.2377123912142561, + "grad_norm": 0.2503049373626709, + "learning_rate": 3.811852465810195e-06, + "loss": 0.5252, + "step": 2868 + }, + { + "epoch": 0.2377952755905512, + "grad_norm": 0.2643716633319855, + "learning_rate": 3.8114380439287196e-06, + "loss": 0.5422, + "step": 2869 + }, + { + "epoch": 0.23787815996684625, + "grad_norm": 0.26059290766716003, + "learning_rate": 3.8110236220472445e-06, + "loss": 0.5329, + "step": 2870 + }, + { + "epoch": 0.23796104434314133, + "grad_norm": 0.2912364602088928, + "learning_rate": 3.810609200165769e-06, + "loss": 0.4684, + "step": 2871 + }, + { + "epoch": 0.2380439287194364, + "grad_norm": 0.2661556601524353, + "learning_rate": 3.8101947782842933e-06, + "loss": 0.5544, + "step": 2872 + }, + { + "epoch": 0.23812681309573144, + "grad_norm": 0.28191572427749634, + "learning_rate": 3.8097803564028186e-06, + "loss": 0.5354, + "step": 2873 + }, + { + "epoch": 0.23820969747202653, + "grad_norm": 0.25410979986190796, + "learning_rate": 3.809365934521343e-06, + "loss": 0.4986, + "step": 2874 + }, + { + "epoch": 0.23829258184832158, + "grad_norm": 0.29665201902389526, + "learning_rate": 3.808951512639868e-06, + "loss": 0.5552, + "step": 2875 + }, + { + "epoch": 0.23837546622461667, + "grad_norm": 0.2822943925857544, + "learning_rate": 3.8085370907583923e-06, + "loss": 0.5538, + "step": 2876 + }, + { + "epoch": 0.23845835060091172, + "grad_norm": 0.2417314052581787, + "learning_rate": 3.8081226688769168e-06, + "loss": 0.5217, + "step": 2877 + }, + { + "epoch": 0.2385412349772068, + "grad_norm": 0.27673667669296265, + "learning_rate": 3.8077082469954416e-06, + "loss": 0.53, + "step": 2878 + }, + { + "epoch": 0.23862411935350186, + "grad_norm": 0.2639002799987793, + "learning_rate": 3.8072938251139665e-06, + "loss": 0.4869, + "step": 2879 + }, + { + "epoch": 0.23870700372979695, + "grad_norm": 0.273662805557251, + "learning_rate": 3.806879403232491e-06, + "loss": 0.5873, + "step": 2880 + }, + { + "epoch": 0.238789888106092, + "grad_norm": 0.27384787797927856, + "learning_rate": 3.8064649813510158e-06, + "loss": 0.5059, + "step": 2881 + }, + { + "epoch": 0.23887277248238706, + "grad_norm": 0.2755955159664154, + "learning_rate": 3.80605055946954e-06, + "loss": 0.6012, + "step": 2882 + }, + { + "epoch": 0.23895565685868214, + "grad_norm": 0.27895885705947876, + "learning_rate": 3.805636137588065e-06, + "loss": 0.5185, + "step": 2883 + }, + { + "epoch": 0.2390385412349772, + "grad_norm": 0.26949575543403625, + "learning_rate": 3.8052217157065895e-06, + "loss": 0.5777, + "step": 2884 + }, + { + "epoch": 0.23912142561127228, + "grad_norm": 0.2733417749404907, + "learning_rate": 3.804807293825114e-06, + "loss": 0.5054, + "step": 2885 + }, + { + "epoch": 0.23920430998756734, + "grad_norm": 0.2581878900527954, + "learning_rate": 3.804392871943639e-06, + "loss": 0.5259, + "step": 2886 + }, + { + "epoch": 0.23928719436386242, + "grad_norm": 0.26191768050193787, + "learning_rate": 3.8039784500621636e-06, + "loss": 0.5591, + "step": 2887 + }, + { + "epoch": 0.23937007874015748, + "grad_norm": 0.2647910416126251, + "learning_rate": 3.8035640281806885e-06, + "loss": 0.5245, + "step": 2888 + }, + { + "epoch": 0.23945296311645256, + "grad_norm": 0.270498126745224, + "learning_rate": 3.803149606299213e-06, + "loss": 0.5249, + "step": 2889 + }, + { + "epoch": 0.23953584749274762, + "grad_norm": 0.2743970453739166, + "learning_rate": 3.8027351844177373e-06, + "loss": 0.5549, + "step": 2890 + }, + { + "epoch": 0.23961873186904267, + "grad_norm": 0.27447330951690674, + "learning_rate": 3.802320762536262e-06, + "loss": 0.5423, + "step": 2891 + }, + { + "epoch": 0.23970161624533776, + "grad_norm": 0.27858176827430725, + "learning_rate": 3.8019063406547866e-06, + "loss": 0.508, + "step": 2892 + }, + { + "epoch": 0.2397845006216328, + "grad_norm": 0.24727633595466614, + "learning_rate": 3.801491918773312e-06, + "loss": 0.4875, + "step": 2893 + }, + { + "epoch": 0.2398673849979279, + "grad_norm": 0.290696382522583, + "learning_rate": 3.8010774968918363e-06, + "loss": 0.5682, + "step": 2894 + }, + { + "epoch": 0.23995026937422295, + "grad_norm": 0.2471727728843689, + "learning_rate": 3.8006630750103608e-06, + "loss": 0.5161, + "step": 2895 + }, + { + "epoch": 0.24003315375051804, + "grad_norm": 0.26909899711608887, + "learning_rate": 3.8002486531288856e-06, + "loss": 0.533, + "step": 2896 + }, + { + "epoch": 0.2401160381268131, + "grad_norm": 0.26616325974464417, + "learning_rate": 3.79983423124741e-06, + "loss": 0.5665, + "step": 2897 + }, + { + "epoch": 0.24019892250310818, + "grad_norm": 0.28021809458732605, + "learning_rate": 3.7994198093659345e-06, + "loss": 0.5281, + "step": 2898 + }, + { + "epoch": 0.24028180687940323, + "grad_norm": 0.25037673115730286, + "learning_rate": 3.7990053874844598e-06, + "loss": 0.5436, + "step": 2899 + }, + { + "epoch": 0.2403646912556983, + "grad_norm": 0.2680245041847229, + "learning_rate": 3.798590965602984e-06, + "loss": 0.525, + "step": 2900 + }, + { + "epoch": 0.24044757563199337, + "grad_norm": 0.27185454964637756, + "learning_rate": 3.798176543721509e-06, + "loss": 0.5836, + "step": 2901 + }, + { + "epoch": 0.24053046000828843, + "grad_norm": 0.2742295265197754, + "learning_rate": 3.7977621218400335e-06, + "loss": 0.5988, + "step": 2902 + }, + { + "epoch": 0.2406133443845835, + "grad_norm": 0.2574271261692047, + "learning_rate": 3.797347699958558e-06, + "loss": 0.5206, + "step": 2903 + }, + { + "epoch": 0.24069622876087857, + "grad_norm": 0.2675509750843048, + "learning_rate": 3.7969332780770828e-06, + "loss": 0.5736, + "step": 2904 + }, + { + "epoch": 0.24077911313717365, + "grad_norm": 0.2547664940357208, + "learning_rate": 3.7965188561956072e-06, + "loss": 0.5518, + "step": 2905 + }, + { + "epoch": 0.2408619975134687, + "grad_norm": 0.25412020087242126, + "learning_rate": 3.7961044343141325e-06, + "loss": 0.4934, + "step": 2906 + }, + { + "epoch": 0.2409448818897638, + "grad_norm": 0.2758173942565918, + "learning_rate": 3.795690012432657e-06, + "loss": 0.5757, + "step": 2907 + }, + { + "epoch": 0.24102776626605885, + "grad_norm": 0.288918673992157, + "learning_rate": 3.7952755905511814e-06, + "loss": 0.5301, + "step": 2908 + }, + { + "epoch": 0.24111065064235393, + "grad_norm": 0.27973687648773193, + "learning_rate": 3.7948611686697062e-06, + "loss": 0.5026, + "step": 2909 + }, + { + "epoch": 0.24119353501864899, + "grad_norm": 0.2583288252353668, + "learning_rate": 3.7944467467882306e-06, + "loss": 0.5279, + "step": 2910 + }, + { + "epoch": 0.24127641939494404, + "grad_norm": 0.2773160934448242, + "learning_rate": 3.7940323249067555e-06, + "loss": 0.5137, + "step": 2911 + }, + { + "epoch": 0.24135930377123913, + "grad_norm": 0.25909003615379333, + "learning_rate": 3.79361790302528e-06, + "loss": 0.5513, + "step": 2912 + }, + { + "epoch": 0.24144218814753418, + "grad_norm": 0.267438679933548, + "learning_rate": 3.7932034811438044e-06, + "loss": 0.5376, + "step": 2913 + }, + { + "epoch": 0.24152507252382927, + "grad_norm": 0.3475324213504791, + "learning_rate": 3.7927890592623296e-06, + "loss": 0.5034, + "step": 2914 + }, + { + "epoch": 0.24160795690012432, + "grad_norm": 0.33398282527923584, + "learning_rate": 3.792374637380854e-06, + "loss": 0.5336, + "step": 2915 + }, + { + "epoch": 0.2416908412764194, + "grad_norm": 0.26540637016296387, + "learning_rate": 3.7919602154993785e-06, + "loss": 0.5032, + "step": 2916 + }, + { + "epoch": 0.24177372565271446, + "grad_norm": 0.24911564588546753, + "learning_rate": 3.7915457936179034e-06, + "loss": 0.5345, + "step": 2917 + }, + { + "epoch": 0.24185661002900954, + "grad_norm": 0.30641108751296997, + "learning_rate": 3.791131371736428e-06, + "loss": 0.6093, + "step": 2918 + }, + { + "epoch": 0.2419394944053046, + "grad_norm": 0.2514283359050751, + "learning_rate": 3.790716949854953e-06, + "loss": 0.522, + "step": 2919 + }, + { + "epoch": 0.24202237878159966, + "grad_norm": 0.2622119188308716, + "learning_rate": 3.7903025279734775e-06, + "loss": 0.5386, + "step": 2920 + }, + { + "epoch": 0.24210526315789474, + "grad_norm": 0.2460150420665741, + "learning_rate": 3.789888106092002e-06, + "loss": 0.503, + "step": 2921 + }, + { + "epoch": 0.2421881475341898, + "grad_norm": 0.2782851457595825, + "learning_rate": 3.789473684210527e-06, + "loss": 0.5825, + "step": 2922 + }, + { + "epoch": 0.24227103191048488, + "grad_norm": 0.24232982099056244, + "learning_rate": 3.7890592623290512e-06, + "loss": 0.5009, + "step": 2923 + }, + { + "epoch": 0.24235391628677994, + "grad_norm": 0.2683247923851013, + "learning_rate": 3.788644840447576e-06, + "loss": 0.543, + "step": 2924 + }, + { + "epoch": 0.24243680066307502, + "grad_norm": 0.2626824378967285, + "learning_rate": 3.7882304185661005e-06, + "loss": 0.5554, + "step": 2925 + }, + { + "epoch": 0.24251968503937008, + "grad_norm": 0.253456711769104, + "learning_rate": 3.787815996684625e-06, + "loss": 0.5349, + "step": 2926 + }, + { + "epoch": 0.24260256941566516, + "grad_norm": 0.3024767339229584, + "learning_rate": 3.7874015748031502e-06, + "loss": 0.5329, + "step": 2927 + }, + { + "epoch": 0.24268545379196022, + "grad_norm": 0.26231634616851807, + "learning_rate": 3.7869871529216747e-06, + "loss": 0.5149, + "step": 2928 + }, + { + "epoch": 0.24276833816825527, + "grad_norm": 0.2833091914653778, + "learning_rate": 3.786572731040199e-06, + "loss": 0.5878, + "step": 2929 + }, + { + "epoch": 0.24285122254455035, + "grad_norm": 0.24936148524284363, + "learning_rate": 3.786158309158724e-06, + "loss": 0.4908, + "step": 2930 + }, + { + "epoch": 0.2429341069208454, + "grad_norm": 0.2504430413246155, + "learning_rate": 3.7857438872772484e-06, + "loss": 0.5297, + "step": 2931 + }, + { + "epoch": 0.2430169912971405, + "grad_norm": 0.25314074754714966, + "learning_rate": 3.7853294653957732e-06, + "loss": 0.5035, + "step": 2932 + }, + { + "epoch": 0.24309987567343555, + "grad_norm": 0.2860773801803589, + "learning_rate": 3.7849150435142977e-06, + "loss": 0.5845, + "step": 2933 + }, + { + "epoch": 0.24318276004973063, + "grad_norm": 0.2637626826763153, + "learning_rate": 3.784500621632822e-06, + "loss": 0.5409, + "step": 2934 + }, + { + "epoch": 0.2432656444260257, + "grad_norm": 0.2591266334056854, + "learning_rate": 3.7840861997513474e-06, + "loss": 0.5446, + "step": 2935 + }, + { + "epoch": 0.24334852880232077, + "grad_norm": 0.24834680557250977, + "learning_rate": 3.783671777869872e-06, + "loss": 0.5605, + "step": 2936 + }, + { + "epoch": 0.24343141317861583, + "grad_norm": 0.2565367519855499, + "learning_rate": 3.7832573559883967e-06, + "loss": 0.4942, + "step": 2937 + }, + { + "epoch": 0.24351429755491089, + "grad_norm": 0.2608732581138611, + "learning_rate": 3.782842934106921e-06, + "loss": 0.5447, + "step": 2938 + }, + { + "epoch": 0.24359718193120597, + "grad_norm": 0.2801371216773987, + "learning_rate": 3.7824285122254455e-06, + "loss": 0.5268, + "step": 2939 + }, + { + "epoch": 0.24368006630750103, + "grad_norm": 0.25785061717033386, + "learning_rate": 3.782014090343971e-06, + "loss": 0.5266, + "step": 2940 + }, + { + "epoch": 0.2437629506837961, + "grad_norm": 0.2707260847091675, + "learning_rate": 3.7815996684624952e-06, + "loss": 0.5418, + "step": 2941 + }, + { + "epoch": 0.24384583506009117, + "grad_norm": 0.26634305715560913, + "learning_rate": 3.78118524658102e-06, + "loss": 0.5703, + "step": 2942 + }, + { + "epoch": 0.24392871943638625, + "grad_norm": 0.2638738751411438, + "learning_rate": 3.7807708246995445e-06, + "loss": 0.502, + "step": 2943 + }, + { + "epoch": 0.2440116038126813, + "grad_norm": 0.24248656630516052, + "learning_rate": 3.780356402818069e-06, + "loss": 0.4602, + "step": 2944 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 0.24324141442775726, + "learning_rate": 3.779941980936594e-06, + "loss": 0.522, + "step": 2945 + }, + { + "epoch": 0.24417737256527144, + "grad_norm": 0.2703179717063904, + "learning_rate": 3.7795275590551182e-06, + "loss": 0.5344, + "step": 2946 + }, + { + "epoch": 0.24426025694156653, + "grad_norm": 0.28274819254875183, + "learning_rate": 3.7791131371736427e-06, + "loss": 0.5585, + "step": 2947 + }, + { + "epoch": 0.24434314131786158, + "grad_norm": 0.2794313132762909, + "learning_rate": 3.778698715292168e-06, + "loss": 0.5395, + "step": 2948 + }, + { + "epoch": 0.24442602569415664, + "grad_norm": 0.2567288279533386, + "learning_rate": 3.7782842934106924e-06, + "loss": 0.5494, + "step": 2949 + }, + { + "epoch": 0.24450891007045172, + "grad_norm": 0.2858749330043793, + "learning_rate": 3.7778698715292172e-06, + "loss": 0.5374, + "step": 2950 + }, + { + "epoch": 0.24459179444674678, + "grad_norm": 0.2617771327495575, + "learning_rate": 3.7774554496477417e-06, + "loss": 0.5418, + "step": 2951 + }, + { + "epoch": 0.24467467882304186, + "grad_norm": 0.26723724603652954, + "learning_rate": 3.777041027766266e-06, + "loss": 0.5309, + "step": 2952 + }, + { + "epoch": 0.24475756319933692, + "grad_norm": 0.268364280462265, + "learning_rate": 3.776626605884791e-06, + "loss": 0.5494, + "step": 2953 + }, + { + "epoch": 0.244840447575632, + "grad_norm": 0.27041900157928467, + "learning_rate": 3.7762121840033154e-06, + "loss": 0.5453, + "step": 2954 + }, + { + "epoch": 0.24492333195192706, + "grad_norm": 0.2638286352157593, + "learning_rate": 3.7757977621218407e-06, + "loss": 0.5153, + "step": 2955 + }, + { + "epoch": 0.24500621632822214, + "grad_norm": 0.2652256190776825, + "learning_rate": 3.775383340240365e-06, + "loss": 0.5219, + "step": 2956 + }, + { + "epoch": 0.2450891007045172, + "grad_norm": 0.28348666429519653, + "learning_rate": 3.7749689183588895e-06, + "loss": 0.5322, + "step": 2957 + }, + { + "epoch": 0.24517198508081225, + "grad_norm": 0.26638445258140564, + "learning_rate": 3.7745544964774144e-06, + "loss": 0.5621, + "step": 2958 + }, + { + "epoch": 0.24525486945710734, + "grad_norm": 0.25459223985671997, + "learning_rate": 3.774140074595939e-06, + "loss": 0.502, + "step": 2959 + }, + { + "epoch": 0.2453377538334024, + "grad_norm": 0.28557005524635315, + "learning_rate": 3.773725652714464e-06, + "loss": 0.607, + "step": 2960 + }, + { + "epoch": 0.24542063820969748, + "grad_norm": 0.25635772943496704, + "learning_rate": 3.7733112308329885e-06, + "loss": 0.5121, + "step": 2961 + }, + { + "epoch": 0.24550352258599253, + "grad_norm": 0.2621753215789795, + "learning_rate": 3.772896808951513e-06, + "loss": 0.5317, + "step": 2962 + }, + { + "epoch": 0.24558640696228762, + "grad_norm": 0.2637782096862793, + "learning_rate": 3.772482387070038e-06, + "loss": 0.5451, + "step": 2963 + }, + { + "epoch": 0.24566929133858267, + "grad_norm": 0.27018091082572937, + "learning_rate": 3.7720679651885622e-06, + "loss": 0.503, + "step": 2964 + }, + { + "epoch": 0.24575217571487776, + "grad_norm": 0.2879149615764618, + "learning_rate": 3.7716535433070867e-06, + "loss": 0.5882, + "step": 2965 + }, + { + "epoch": 0.2458350600911728, + "grad_norm": 0.27001917362213135, + "learning_rate": 3.7712391214256115e-06, + "loss": 0.5831, + "step": 2966 + }, + { + "epoch": 0.24591794446746787, + "grad_norm": 0.24982865154743195, + "learning_rate": 3.770824699544136e-06, + "loss": 0.4897, + "step": 2967 + }, + { + "epoch": 0.24600082884376295, + "grad_norm": 0.2519347667694092, + "learning_rate": 3.7704102776626612e-06, + "loss": 0.5226, + "step": 2968 + }, + { + "epoch": 0.246083713220058, + "grad_norm": 0.27283987402915955, + "learning_rate": 3.7699958557811857e-06, + "loss": 0.5323, + "step": 2969 + }, + { + "epoch": 0.2461665975963531, + "grad_norm": 0.27165406942367554, + "learning_rate": 3.76958143389971e-06, + "loss": 0.5543, + "step": 2970 + }, + { + "epoch": 0.24624948197264815, + "grad_norm": 0.2580934166908264, + "learning_rate": 3.769167012018235e-06, + "loss": 0.5435, + "step": 2971 + }, + { + "epoch": 0.24633236634894323, + "grad_norm": 0.27948808670043945, + "learning_rate": 3.7687525901367594e-06, + "loss": 0.5474, + "step": 2972 + }, + { + "epoch": 0.2464152507252383, + "grad_norm": 0.2849757671356201, + "learning_rate": 3.7683381682552842e-06, + "loss": 0.6068, + "step": 2973 + }, + { + "epoch": 0.24649813510153337, + "grad_norm": 0.2588117718696594, + "learning_rate": 3.7679237463738087e-06, + "loss": 0.5146, + "step": 2974 + }, + { + "epoch": 0.24658101947782843, + "grad_norm": 0.27146750688552856, + "learning_rate": 3.767509324492333e-06, + "loss": 0.5263, + "step": 2975 + }, + { + "epoch": 0.24666390385412348, + "grad_norm": 0.2815910875797272, + "learning_rate": 3.7670949026108584e-06, + "loss": 0.5402, + "step": 2976 + }, + { + "epoch": 0.24674678823041857, + "grad_norm": 0.2701902687549591, + "learning_rate": 3.766680480729383e-06, + "loss": 0.54, + "step": 2977 + }, + { + "epoch": 0.24682967260671362, + "grad_norm": 0.2611358165740967, + "learning_rate": 3.7662660588479077e-06, + "loss": 0.5372, + "step": 2978 + }, + { + "epoch": 0.2469125569830087, + "grad_norm": 0.2685965299606323, + "learning_rate": 3.765851636966432e-06, + "loss": 0.5309, + "step": 2979 + }, + { + "epoch": 0.24699544135930376, + "grad_norm": 0.28486499190330505, + "learning_rate": 3.7654372150849565e-06, + "loss": 0.6033, + "step": 2980 + }, + { + "epoch": 0.24707832573559885, + "grad_norm": 0.24450184404850006, + "learning_rate": 3.765022793203482e-06, + "loss": 0.535, + "step": 2981 + }, + { + "epoch": 0.2471612101118939, + "grad_norm": 0.2611891031265259, + "learning_rate": 3.7646083713220063e-06, + "loss": 0.5556, + "step": 2982 + }, + { + "epoch": 0.247244094488189, + "grad_norm": 0.3325996398925781, + "learning_rate": 3.7641939494405307e-06, + "loss": 0.5505, + "step": 2983 + }, + { + "epoch": 0.24732697886448404, + "grad_norm": 0.2497004121541977, + "learning_rate": 3.7637795275590555e-06, + "loss": 0.529, + "step": 2984 + }, + { + "epoch": 0.24740986324077913, + "grad_norm": 0.26184993982315063, + "learning_rate": 3.76336510567758e-06, + "loss": 0.5247, + "step": 2985 + }, + { + "epoch": 0.24749274761707418, + "grad_norm": 0.2544892132282257, + "learning_rate": 3.762950683796105e-06, + "loss": 0.5152, + "step": 2986 + }, + { + "epoch": 0.24757563199336924, + "grad_norm": 0.280817449092865, + "learning_rate": 3.7625362619146293e-06, + "loss": 0.5381, + "step": 2987 + }, + { + "epoch": 0.24765851636966432, + "grad_norm": 0.2638157904148102, + "learning_rate": 3.7621218400331537e-06, + "loss": 0.5714, + "step": 2988 + }, + { + "epoch": 0.24774140074595938, + "grad_norm": 0.2631092071533203, + "learning_rate": 3.761707418151679e-06, + "loss": 0.5793, + "step": 2989 + }, + { + "epoch": 0.24782428512225446, + "grad_norm": 0.26887232065200806, + "learning_rate": 3.7612929962702034e-06, + "loss": 0.5735, + "step": 2990 + }, + { + "epoch": 0.24790716949854952, + "grad_norm": 0.26314815878868103, + "learning_rate": 3.7608785743887283e-06, + "loss": 0.5217, + "step": 2991 + }, + { + "epoch": 0.2479900538748446, + "grad_norm": 0.27854809165000916, + "learning_rate": 3.7604641525072527e-06, + "loss": 0.5572, + "step": 2992 + }, + { + "epoch": 0.24807293825113966, + "grad_norm": 0.2871966063976288, + "learning_rate": 3.760049730625777e-06, + "loss": 0.6204, + "step": 2993 + }, + { + "epoch": 0.24815582262743474, + "grad_norm": 0.25904199481010437, + "learning_rate": 3.759635308744302e-06, + "loss": 0.5366, + "step": 2994 + }, + { + "epoch": 0.2482387070037298, + "grad_norm": 0.25158798694610596, + "learning_rate": 3.7592208868628264e-06, + "loss": 0.5454, + "step": 2995 + }, + { + "epoch": 0.24832159138002485, + "grad_norm": 0.2513498365879059, + "learning_rate": 3.7588064649813517e-06, + "loss": 0.5266, + "step": 2996 + }, + { + "epoch": 0.24840447575631994, + "grad_norm": 0.25494641065597534, + "learning_rate": 3.758392043099876e-06, + "loss": 0.5144, + "step": 2997 + }, + { + "epoch": 0.248487360132615, + "grad_norm": 0.2622510492801666, + "learning_rate": 3.7579776212184005e-06, + "loss": 0.5356, + "step": 2998 + }, + { + "epoch": 0.24857024450891008, + "grad_norm": 0.2777751684188843, + "learning_rate": 3.7575631993369254e-06, + "loss": 0.5389, + "step": 2999 + }, + { + "epoch": 0.24865312888520513, + "grad_norm": 0.23744544386863708, + "learning_rate": 3.75714877745545e-06, + "loss": 0.4887, + "step": 3000 + }, + { + "epoch": 0.24873601326150022, + "grad_norm": 0.2788255214691162, + "learning_rate": 3.7567343555739743e-06, + "loss": 0.5785, + "step": 3001 + }, + { + "epoch": 0.24881889763779527, + "grad_norm": 0.2733173966407776, + "learning_rate": 3.7563199336924995e-06, + "loss": 0.5126, + "step": 3002 + }, + { + "epoch": 0.24890178201409036, + "grad_norm": 0.2606307864189148, + "learning_rate": 3.755905511811024e-06, + "loss": 0.5365, + "step": 3003 + }, + { + "epoch": 0.2489846663903854, + "grad_norm": 0.29391196370124817, + "learning_rate": 3.755491089929549e-06, + "loss": 0.624, + "step": 3004 + }, + { + "epoch": 0.24906755076668047, + "grad_norm": 0.28584280610084534, + "learning_rate": 3.7550766680480733e-06, + "loss": 0.5477, + "step": 3005 + }, + { + "epoch": 0.24915043514297555, + "grad_norm": 0.27948787808418274, + "learning_rate": 3.7546622461665977e-06, + "loss": 0.5135, + "step": 3006 + }, + { + "epoch": 0.2492333195192706, + "grad_norm": 0.366553395986557, + "learning_rate": 3.7542478242851226e-06, + "loss": 0.6236, + "step": 3007 + }, + { + "epoch": 0.2493162038955657, + "grad_norm": 0.27223947644233704, + "learning_rate": 3.753833402403647e-06, + "loss": 0.5471, + "step": 3008 + }, + { + "epoch": 0.24939908827186075, + "grad_norm": 0.27745741605758667, + "learning_rate": 3.7534189805221723e-06, + "loss": 0.5724, + "step": 3009 + }, + { + "epoch": 0.24948197264815583, + "grad_norm": 0.25414079427719116, + "learning_rate": 3.7530045586406967e-06, + "loss": 0.5258, + "step": 3010 + }, + { + "epoch": 0.2495648570244509, + "grad_norm": 0.27630192041397095, + "learning_rate": 3.752590136759221e-06, + "loss": 0.5112, + "step": 3011 + }, + { + "epoch": 0.24964774140074597, + "grad_norm": 0.2722190022468567, + "learning_rate": 3.752175714877746e-06, + "loss": 0.5805, + "step": 3012 + }, + { + "epoch": 0.24973062577704103, + "grad_norm": 0.2790640890598297, + "learning_rate": 3.7517612929962704e-06, + "loss": 0.559, + "step": 3013 + }, + { + "epoch": 0.24981351015333608, + "grad_norm": 0.2865855097770691, + "learning_rate": 3.751346871114795e-06, + "loss": 0.5864, + "step": 3014 + }, + { + "epoch": 0.24989639452963117, + "grad_norm": 0.26628485321998596, + "learning_rate": 3.7509324492333197e-06, + "loss": 0.525, + "step": 3015 + }, + { + "epoch": 0.24997927890592622, + "grad_norm": 0.27251118421554565, + "learning_rate": 3.750518027351844e-06, + "loss": 0.5432, + "step": 3016 + }, + { + "epoch": 0.2500621632822213, + "grad_norm": 0.26253893971443176, + "learning_rate": 3.7501036054703694e-06, + "loss": 0.4752, + "step": 3017 + }, + { + "epoch": 0.25014504765851636, + "grad_norm": 0.25688764452934265, + "learning_rate": 3.749689183588894e-06, + "loss": 0.5178, + "step": 3018 + }, + { + "epoch": 0.2502279320348114, + "grad_norm": 0.25001412630081177, + "learning_rate": 3.7492747617074183e-06, + "loss": 0.5641, + "step": 3019 + }, + { + "epoch": 0.25031081641110653, + "grad_norm": 0.25986945629119873, + "learning_rate": 3.748860339825943e-06, + "loss": 0.5201, + "step": 3020 + }, + { + "epoch": 0.2503937007874016, + "grad_norm": 0.2589210271835327, + "learning_rate": 3.7484459179444676e-06, + "loss": 0.5033, + "step": 3021 + }, + { + "epoch": 0.25047658516369664, + "grad_norm": 0.30750536918640137, + "learning_rate": 3.748031496062993e-06, + "loss": 0.5496, + "step": 3022 + }, + { + "epoch": 0.2505594695399917, + "grad_norm": 0.25754088163375854, + "learning_rate": 3.7476170741815173e-06, + "loss": 0.5272, + "step": 3023 + }, + { + "epoch": 0.25064235391628675, + "grad_norm": 0.2537056505680084, + "learning_rate": 3.7472026523000417e-06, + "loss": 0.5267, + "step": 3024 + }, + { + "epoch": 0.25072523829258186, + "grad_norm": 0.2624424397945404, + "learning_rate": 3.7467882304185666e-06, + "loss": 0.5266, + "step": 3025 + }, + { + "epoch": 0.2508081226688769, + "grad_norm": 0.2578641474246979, + "learning_rate": 3.746373808537091e-06, + "loss": 0.5165, + "step": 3026 + }, + { + "epoch": 0.250891007045172, + "grad_norm": 0.2584494650363922, + "learning_rate": 3.745959386655616e-06, + "loss": 0.5368, + "step": 3027 + }, + { + "epoch": 0.25097389142146703, + "grad_norm": 0.27494457364082336, + "learning_rate": 3.7455449647741403e-06, + "loss": 0.537, + "step": 3028 + }, + { + "epoch": 0.25105677579776214, + "grad_norm": 0.2910599410533905, + "learning_rate": 3.7451305428926647e-06, + "loss": 0.5298, + "step": 3029 + }, + { + "epoch": 0.2511396601740572, + "grad_norm": 0.2474503368139267, + "learning_rate": 3.74471612101119e-06, + "loss": 0.5104, + "step": 3030 + }, + { + "epoch": 0.25122254455035226, + "grad_norm": 0.2715228796005249, + "learning_rate": 3.7443016991297144e-06, + "loss": 0.5222, + "step": 3031 + }, + { + "epoch": 0.2513054289266473, + "grad_norm": 0.2661990225315094, + "learning_rate": 3.743887277248239e-06, + "loss": 0.5396, + "step": 3032 + }, + { + "epoch": 0.25138831330294237, + "grad_norm": 0.2637366056442261, + "learning_rate": 3.7434728553667637e-06, + "loss": 0.5188, + "step": 3033 + }, + { + "epoch": 0.2514711976792375, + "grad_norm": 0.27825313806533813, + "learning_rate": 3.743058433485288e-06, + "loss": 0.5646, + "step": 3034 + }, + { + "epoch": 0.25155408205553254, + "grad_norm": 0.2558795213699341, + "learning_rate": 3.742644011603813e-06, + "loss": 0.527, + "step": 3035 + }, + { + "epoch": 0.2516369664318276, + "grad_norm": 0.23890405893325806, + "learning_rate": 3.7422295897223374e-06, + "loss": 0.5175, + "step": 3036 + }, + { + "epoch": 0.25171985080812265, + "grad_norm": 0.25936874747276306, + "learning_rate": 3.7418151678408623e-06, + "loss": 0.5657, + "step": 3037 + }, + { + "epoch": 0.25180273518441776, + "grad_norm": 0.26159337162971497, + "learning_rate": 3.741400745959387e-06, + "loss": 0.5178, + "step": 3038 + }, + { + "epoch": 0.2518856195607128, + "grad_norm": 0.27023738622665405, + "learning_rate": 3.7409863240779116e-06, + "loss": 0.5143, + "step": 3039 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 0.2551354765892029, + "learning_rate": 3.7405719021964364e-06, + "loss": 0.547, + "step": 3040 + }, + { + "epoch": 0.2520513883133029, + "grad_norm": 0.2736297845840454, + "learning_rate": 3.740157480314961e-06, + "loss": 0.5456, + "step": 3041 + }, + { + "epoch": 0.25213427268959804, + "grad_norm": 0.27293774485588074, + "learning_rate": 3.7397430584334853e-06, + "loss": 0.5233, + "step": 3042 + }, + { + "epoch": 0.2522171570658931, + "grad_norm": 0.2725094258785248, + "learning_rate": 3.7393286365520106e-06, + "loss": 0.5121, + "step": 3043 + }, + { + "epoch": 0.25230004144218815, + "grad_norm": 0.2814314067363739, + "learning_rate": 3.738914214670535e-06, + "loss": 0.5617, + "step": 3044 + }, + { + "epoch": 0.2523829258184832, + "grad_norm": 0.28981003165245056, + "learning_rate": 3.73849979278906e-06, + "loss": 0.5387, + "step": 3045 + }, + { + "epoch": 0.25246581019477826, + "grad_norm": 0.2704260051250458, + "learning_rate": 3.7380853709075843e-06, + "loss": 0.5185, + "step": 3046 + }, + { + "epoch": 0.2525486945710734, + "grad_norm": 0.2719832956790924, + "learning_rate": 3.7376709490261087e-06, + "loss": 0.5308, + "step": 3047 + }, + { + "epoch": 0.25263157894736843, + "grad_norm": 0.24504654109477997, + "learning_rate": 3.7372565271446336e-06, + "loss": 0.5173, + "step": 3048 + }, + { + "epoch": 0.2527144633236635, + "grad_norm": 0.252707839012146, + "learning_rate": 3.736842105263158e-06, + "loss": 0.5522, + "step": 3049 + }, + { + "epoch": 0.25279734769995854, + "grad_norm": 0.2693573236465454, + "learning_rate": 3.7364276833816824e-06, + "loss": 0.5128, + "step": 3050 + }, + { + "epoch": 0.25288023207625365, + "grad_norm": 0.25752368569374084, + "learning_rate": 3.7360132615002077e-06, + "loss": 0.5291, + "step": 3051 + }, + { + "epoch": 0.2529631164525487, + "grad_norm": 0.2555358111858368, + "learning_rate": 3.735598839618732e-06, + "loss": 0.5174, + "step": 3052 + }, + { + "epoch": 0.25304600082884376, + "grad_norm": 0.24389436841011047, + "learning_rate": 3.735184417737257e-06, + "loss": 0.5134, + "step": 3053 + }, + { + "epoch": 0.2531288852051388, + "grad_norm": 0.25534042716026306, + "learning_rate": 3.7347699958557814e-06, + "loss": 0.5155, + "step": 3054 + }, + { + "epoch": 0.2532117695814339, + "grad_norm": 0.24968723952770233, + "learning_rate": 3.734355573974306e-06, + "loss": 0.5126, + "step": 3055 + }, + { + "epoch": 0.253294653957729, + "grad_norm": 0.2532842755317688, + "learning_rate": 3.7339411520928307e-06, + "loss": 0.5703, + "step": 3056 + }, + { + "epoch": 0.25337753833402404, + "grad_norm": 0.25002792477607727, + "learning_rate": 3.7335267302113556e-06, + "loss": 0.5311, + "step": 3057 + }, + { + "epoch": 0.2534604227103191, + "grad_norm": 0.26333415508270264, + "learning_rate": 3.7331123083298804e-06, + "loss": 0.5293, + "step": 3058 + }, + { + "epoch": 0.25354330708661416, + "grad_norm": 0.2623833417892456, + "learning_rate": 3.732697886448405e-06, + "loss": 0.527, + "step": 3059 + }, + { + "epoch": 0.25362619146290927, + "grad_norm": 0.2632240951061249, + "learning_rate": 3.7322834645669293e-06, + "loss": 0.5648, + "step": 3060 + }, + { + "epoch": 0.2537090758392043, + "grad_norm": 0.29769372940063477, + "learning_rate": 3.731869042685454e-06, + "loss": 0.5493, + "step": 3061 + }, + { + "epoch": 0.2537919602154994, + "grad_norm": 0.2858717739582062, + "learning_rate": 3.7314546208039786e-06, + "loss": 0.5403, + "step": 3062 + }, + { + "epoch": 0.25387484459179444, + "grad_norm": 0.2516707479953766, + "learning_rate": 3.731040198922504e-06, + "loss": 0.5219, + "step": 3063 + }, + { + "epoch": 0.2539577289680895, + "grad_norm": 0.2554275691509247, + "learning_rate": 3.7306257770410283e-06, + "loss": 0.5496, + "step": 3064 + }, + { + "epoch": 0.2540406133443846, + "grad_norm": 0.24546800553798676, + "learning_rate": 3.7302113551595527e-06, + "loss": 0.5258, + "step": 3065 + }, + { + "epoch": 0.25412349772067966, + "grad_norm": 0.2596164643764496, + "learning_rate": 3.7297969332780776e-06, + "loss": 0.4892, + "step": 3066 + }, + { + "epoch": 0.2542063820969747, + "grad_norm": 0.24650245904922485, + "learning_rate": 3.729382511396602e-06, + "loss": 0.4942, + "step": 3067 + }, + { + "epoch": 0.25428926647326977, + "grad_norm": 0.24945074319839478, + "learning_rate": 3.7289680895151264e-06, + "loss": 0.5082, + "step": 3068 + }, + { + "epoch": 0.2543721508495649, + "grad_norm": 0.2667863368988037, + "learning_rate": 3.7285536676336513e-06, + "loss": 0.5153, + "step": 3069 + }, + { + "epoch": 0.25445503522585994, + "grad_norm": 0.2694580554962158, + "learning_rate": 3.7281392457521757e-06, + "loss": 0.5219, + "step": 3070 + }, + { + "epoch": 0.254537919602155, + "grad_norm": 0.24955111742019653, + "learning_rate": 3.727724823870701e-06, + "loss": 0.5424, + "step": 3071 + }, + { + "epoch": 0.25462080397845005, + "grad_norm": 0.28380587697029114, + "learning_rate": 3.7273104019892254e-06, + "loss": 0.5663, + "step": 3072 + }, + { + "epoch": 0.2547036883547451, + "grad_norm": 0.2581193447113037, + "learning_rate": 3.72689598010775e-06, + "loss": 0.5396, + "step": 3073 + }, + { + "epoch": 0.2547865727310402, + "grad_norm": 0.2731954753398895, + "learning_rate": 3.7264815582262747e-06, + "loss": 0.5003, + "step": 3074 + }, + { + "epoch": 0.2548694571073353, + "grad_norm": 0.2512478530406952, + "learning_rate": 3.726067136344799e-06, + "loss": 0.5422, + "step": 3075 + }, + { + "epoch": 0.25495234148363033, + "grad_norm": 0.28452634811401367, + "learning_rate": 3.725652714463324e-06, + "loss": 0.5488, + "step": 3076 + }, + { + "epoch": 0.2550352258599254, + "grad_norm": 0.24787850677967072, + "learning_rate": 3.7252382925818484e-06, + "loss": 0.5094, + "step": 3077 + }, + { + "epoch": 0.2551181102362205, + "grad_norm": 0.2460612803697586, + "learning_rate": 3.7248238707003733e-06, + "loss": 0.5063, + "step": 3078 + }, + { + "epoch": 0.25520099461251555, + "grad_norm": 0.27951762080192566, + "learning_rate": 3.724409448818898e-06, + "loss": 0.6046, + "step": 3079 + }, + { + "epoch": 0.2552838789888106, + "grad_norm": 0.261919230222702, + "learning_rate": 3.7239950269374226e-06, + "loss": 0.5195, + "step": 3080 + }, + { + "epoch": 0.25536676336510566, + "grad_norm": 0.2645854949951172, + "learning_rate": 3.723580605055947e-06, + "loss": 0.5203, + "step": 3081 + }, + { + "epoch": 0.2554496477414007, + "grad_norm": 0.2569621801376343, + "learning_rate": 3.723166183174472e-06, + "loss": 0.4885, + "step": 3082 + }, + { + "epoch": 0.25553253211769583, + "grad_norm": 0.2677116394042969, + "learning_rate": 3.7227517612929963e-06, + "loss": 0.5444, + "step": 3083 + }, + { + "epoch": 0.2556154164939909, + "grad_norm": 0.3184617757797241, + "learning_rate": 3.7223373394115216e-06, + "loss": 0.5028, + "step": 3084 + }, + { + "epoch": 0.25569830087028594, + "grad_norm": 0.2508014738559723, + "learning_rate": 3.721922917530046e-06, + "loss": 0.4759, + "step": 3085 + }, + { + "epoch": 0.255781185246581, + "grad_norm": 0.23970019817352295, + "learning_rate": 3.7215084956485705e-06, + "loss": 0.5179, + "step": 3086 + }, + { + "epoch": 0.2558640696228761, + "grad_norm": 0.27541178464889526, + "learning_rate": 3.7210940737670953e-06, + "loss": 0.5605, + "step": 3087 + }, + { + "epoch": 0.25594695399917117, + "grad_norm": 0.257418692111969, + "learning_rate": 3.7206796518856197e-06, + "loss": 0.548, + "step": 3088 + }, + { + "epoch": 0.2560298383754662, + "grad_norm": 0.28665509819984436, + "learning_rate": 3.7202652300041446e-06, + "loss": 0.5574, + "step": 3089 + }, + { + "epoch": 0.2561127227517613, + "grad_norm": 0.2859152555465698, + "learning_rate": 3.719850808122669e-06, + "loss": 0.5293, + "step": 3090 + }, + { + "epoch": 0.25619560712805634, + "grad_norm": 0.2465820014476776, + "learning_rate": 3.7194363862411935e-06, + "loss": 0.5396, + "step": 3091 + }, + { + "epoch": 0.25627849150435145, + "grad_norm": 0.27896153926849365, + "learning_rate": 3.7190219643597187e-06, + "loss": 0.5953, + "step": 3092 + }, + { + "epoch": 0.2563613758806465, + "grad_norm": 0.2607041895389557, + "learning_rate": 3.718607542478243e-06, + "loss": 0.529, + "step": 3093 + }, + { + "epoch": 0.25644426025694156, + "grad_norm": 0.27676376700401306, + "learning_rate": 3.718193120596768e-06, + "loss": 0.5349, + "step": 3094 + }, + { + "epoch": 0.2565271446332366, + "grad_norm": 0.2605842649936676, + "learning_rate": 3.7177786987152925e-06, + "loss": 0.5294, + "step": 3095 + }, + { + "epoch": 0.2566100290095317, + "grad_norm": 0.28537148237228394, + "learning_rate": 3.717364276833817e-06, + "loss": 0.5871, + "step": 3096 + }, + { + "epoch": 0.2566929133858268, + "grad_norm": 0.2626747488975525, + "learning_rate": 3.7169498549523417e-06, + "loss": 0.5466, + "step": 3097 + }, + { + "epoch": 0.25677579776212184, + "grad_norm": 0.25452300906181335, + "learning_rate": 3.7165354330708666e-06, + "loss": 0.5479, + "step": 3098 + }, + { + "epoch": 0.2568586821384169, + "grad_norm": 0.2825677692890167, + "learning_rate": 3.716121011189391e-06, + "loss": 0.543, + "step": 3099 + }, + { + "epoch": 0.25694156651471195, + "grad_norm": 1.7786706686019897, + "learning_rate": 3.715706589307916e-06, + "loss": 0.5327, + "step": 3100 + }, + { + "epoch": 0.25702445089100706, + "grad_norm": 0.26965758204460144, + "learning_rate": 3.7152921674264403e-06, + "loss": 0.5203, + "step": 3101 + }, + { + "epoch": 0.2571073352673021, + "grad_norm": 0.2619054913520813, + "learning_rate": 3.714877745544965e-06, + "loss": 0.5368, + "step": 3102 + }, + { + "epoch": 0.2571902196435972, + "grad_norm": 0.268454909324646, + "learning_rate": 3.7144633236634896e-06, + "loss": 0.5488, + "step": 3103 + }, + { + "epoch": 0.25727310401989223, + "grad_norm": 0.26618608832359314, + "learning_rate": 3.714048901782014e-06, + "loss": 0.5436, + "step": 3104 + }, + { + "epoch": 0.25735598839618734, + "grad_norm": 0.25038501620292664, + "learning_rate": 3.7136344799005393e-06, + "loss": 0.5169, + "step": 3105 + }, + { + "epoch": 0.2574388727724824, + "grad_norm": 0.2662702798843384, + "learning_rate": 3.7132200580190637e-06, + "loss": 0.5729, + "step": 3106 + }, + { + "epoch": 0.25752175714877745, + "grad_norm": 0.27012020349502563, + "learning_rate": 3.7128056361375886e-06, + "loss": 0.5315, + "step": 3107 + }, + { + "epoch": 0.2576046415250725, + "grad_norm": 0.2706333100795746, + "learning_rate": 3.712391214256113e-06, + "loss": 0.5342, + "step": 3108 + }, + { + "epoch": 0.25768752590136756, + "grad_norm": 0.28072986006736755, + "learning_rate": 3.7119767923746375e-06, + "loss": 0.5323, + "step": 3109 + }, + { + "epoch": 0.2577704102776627, + "grad_norm": 0.2718289792537689, + "learning_rate": 3.7115623704931623e-06, + "loss": 0.5506, + "step": 3110 + }, + { + "epoch": 0.25785329465395773, + "grad_norm": 0.23362921178340912, + "learning_rate": 3.7111479486116868e-06, + "loss": 0.4856, + "step": 3111 + }, + { + "epoch": 0.2579361790302528, + "grad_norm": 0.28120511770248413, + "learning_rate": 3.710733526730212e-06, + "loss": 0.6031, + "step": 3112 + }, + { + "epoch": 0.25801906340654784, + "grad_norm": 0.2542876899242401, + "learning_rate": 3.7103191048487365e-06, + "loss": 0.5048, + "step": 3113 + }, + { + "epoch": 0.25810194778284296, + "grad_norm": 0.24785378575325012, + "learning_rate": 3.709904682967261e-06, + "loss": 0.525, + "step": 3114 + }, + { + "epoch": 0.258184832159138, + "grad_norm": 0.26902246475219727, + "learning_rate": 3.7094902610857857e-06, + "loss": 0.5137, + "step": 3115 + }, + { + "epoch": 0.25826771653543307, + "grad_norm": 0.270413339138031, + "learning_rate": 3.70907583920431e-06, + "loss": 0.5265, + "step": 3116 + }, + { + "epoch": 0.2583506009117281, + "grad_norm": 0.27837735414505005, + "learning_rate": 3.7086614173228346e-06, + "loss": 0.5143, + "step": 3117 + }, + { + "epoch": 0.25843348528802323, + "grad_norm": 0.27009308338165283, + "learning_rate": 3.70824699544136e-06, + "loss": 0.5547, + "step": 3118 + }, + { + "epoch": 0.2585163696643183, + "grad_norm": 0.2610076665878296, + "learning_rate": 3.7078325735598843e-06, + "loss": 0.5388, + "step": 3119 + }, + { + "epoch": 0.25859925404061335, + "grad_norm": 0.2643607258796692, + "learning_rate": 3.707418151678409e-06, + "loss": 0.5216, + "step": 3120 + }, + { + "epoch": 0.2586821384169084, + "grad_norm": 0.2820386290550232, + "learning_rate": 3.7070037297969336e-06, + "loss": 0.5646, + "step": 3121 + }, + { + "epoch": 0.25876502279320346, + "grad_norm": 0.25796419382095337, + "learning_rate": 3.706589307915458e-06, + "loss": 0.5124, + "step": 3122 + }, + { + "epoch": 0.25884790716949857, + "grad_norm": 0.266374796628952, + "learning_rate": 3.706174886033983e-06, + "loss": 0.5645, + "step": 3123 + }, + { + "epoch": 0.2589307915457936, + "grad_norm": 0.2405669391155243, + "learning_rate": 3.7057604641525073e-06, + "loss": 0.5105, + "step": 3124 + }, + { + "epoch": 0.2590136759220887, + "grad_norm": 0.2781530022621155, + "learning_rate": 3.7053460422710326e-06, + "loss": 0.5347, + "step": 3125 + }, + { + "epoch": 0.25909656029838374, + "grad_norm": 0.2565508484840393, + "learning_rate": 3.704931620389557e-06, + "loss": 0.5588, + "step": 3126 + }, + { + "epoch": 0.25917944467467885, + "grad_norm": 0.25046679377555847, + "learning_rate": 3.7045171985080815e-06, + "loss": 0.5245, + "step": 3127 + }, + { + "epoch": 0.2592623290509739, + "grad_norm": 0.2733306884765625, + "learning_rate": 3.7041027766266063e-06, + "loss": 0.5869, + "step": 3128 + }, + { + "epoch": 0.25934521342726896, + "grad_norm": 0.2720595598220825, + "learning_rate": 3.7036883547451308e-06, + "loss": 0.5486, + "step": 3129 + }, + { + "epoch": 0.259428097803564, + "grad_norm": 0.2553272545337677, + "learning_rate": 3.7032739328636556e-06, + "loss": 0.5455, + "step": 3130 + }, + { + "epoch": 0.2595109821798591, + "grad_norm": 0.2341809719800949, + "learning_rate": 3.70285951098218e-06, + "loss": 0.497, + "step": 3131 + }, + { + "epoch": 0.2595938665561542, + "grad_norm": 0.2638809084892273, + "learning_rate": 3.7024450891007045e-06, + "loss": 0.514, + "step": 3132 + }, + { + "epoch": 0.25967675093244924, + "grad_norm": 0.2650545537471771, + "learning_rate": 3.7020306672192298e-06, + "loss": 0.5173, + "step": 3133 + }, + { + "epoch": 0.2597596353087443, + "grad_norm": 0.24805589020252228, + "learning_rate": 3.701616245337754e-06, + "loss": 0.5232, + "step": 3134 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 0.258556604385376, + "learning_rate": 3.7012018234562786e-06, + "loss": 0.5317, + "step": 3135 + }, + { + "epoch": 0.25992540406133446, + "grad_norm": 0.24361270666122437, + "learning_rate": 3.7007874015748035e-06, + "loss": 0.474, + "step": 3136 + }, + { + "epoch": 0.2600082884376295, + "grad_norm": 0.27122732996940613, + "learning_rate": 3.700372979693328e-06, + "loss": 0.5709, + "step": 3137 + }, + { + "epoch": 0.2600911728139246, + "grad_norm": 0.2695201337337494, + "learning_rate": 3.699958557811853e-06, + "loss": 0.5761, + "step": 3138 + }, + { + "epoch": 0.26017405719021963, + "grad_norm": 0.2643957734107971, + "learning_rate": 3.6995441359303776e-06, + "loss": 0.5503, + "step": 3139 + }, + { + "epoch": 0.2602569415665147, + "grad_norm": 0.2677244246006012, + "learning_rate": 3.699129714048902e-06, + "loss": 0.5096, + "step": 3140 + }, + { + "epoch": 0.2603398259428098, + "grad_norm": 0.26654812693595886, + "learning_rate": 3.698715292167427e-06, + "loss": 0.565, + "step": 3141 + }, + { + "epoch": 0.26042271031910486, + "grad_norm": 0.29726192355155945, + "learning_rate": 3.6983008702859513e-06, + "loss": 0.5321, + "step": 3142 + }, + { + "epoch": 0.2605055946953999, + "grad_norm": 0.2876032292842865, + "learning_rate": 3.697886448404476e-06, + "loss": 0.5485, + "step": 3143 + }, + { + "epoch": 0.26058847907169497, + "grad_norm": 0.2748839259147644, + "learning_rate": 3.6974720265230006e-06, + "loss": 0.5168, + "step": 3144 + }, + { + "epoch": 0.2606713634479901, + "grad_norm": 0.2645412087440491, + "learning_rate": 3.697057604641525e-06, + "loss": 0.5526, + "step": 3145 + }, + { + "epoch": 0.26075424782428513, + "grad_norm": 0.31355464458465576, + "learning_rate": 3.6966431827600503e-06, + "loss": 0.5578, + "step": 3146 + }, + { + "epoch": 0.2608371322005802, + "grad_norm": 0.31006836891174316, + "learning_rate": 3.6962287608785748e-06, + "loss": 0.5057, + "step": 3147 + }, + { + "epoch": 0.26092001657687525, + "grad_norm": 0.2804972231388092, + "learning_rate": 3.6958143389970996e-06, + "loss": 0.5528, + "step": 3148 + }, + { + "epoch": 0.2610029009531703, + "grad_norm": 0.26445141434669495, + "learning_rate": 3.695399917115624e-06, + "loss": 0.5312, + "step": 3149 + }, + { + "epoch": 0.2610857853294654, + "grad_norm": 0.2728402018547058, + "learning_rate": 3.6949854952341485e-06, + "loss": 0.5022, + "step": 3150 + }, + { + "epoch": 0.26116866970576047, + "grad_norm": 0.4889887273311615, + "learning_rate": 3.6945710733526733e-06, + "loss": 0.5378, + "step": 3151 + }, + { + "epoch": 0.2612515540820555, + "grad_norm": 0.26813146471977234, + "learning_rate": 3.6941566514711978e-06, + "loss": 0.5697, + "step": 3152 + }, + { + "epoch": 0.2613344384583506, + "grad_norm": 0.28469353914260864, + "learning_rate": 3.693742229589722e-06, + "loss": 0.5059, + "step": 3153 + }, + { + "epoch": 0.2614173228346457, + "grad_norm": 0.2495679408311844, + "learning_rate": 3.6933278077082475e-06, + "loss": 0.5097, + "step": 3154 + }, + { + "epoch": 0.26150020721094075, + "grad_norm": 0.2699187099933624, + "learning_rate": 3.692913385826772e-06, + "loss": 0.5463, + "step": 3155 + }, + { + "epoch": 0.2615830915872358, + "grad_norm": 0.28471797704696655, + "learning_rate": 3.6924989639452968e-06, + "loss": 0.5459, + "step": 3156 + }, + { + "epoch": 0.26166597596353086, + "grad_norm": 0.27029719948768616, + "learning_rate": 3.692084542063821e-06, + "loss": 0.5543, + "step": 3157 + }, + { + "epoch": 0.2617488603398259, + "grad_norm": 0.2829173505306244, + "learning_rate": 3.6916701201823456e-06, + "loss": 0.5655, + "step": 3158 + }, + { + "epoch": 0.26183174471612103, + "grad_norm": 0.26450082659721375, + "learning_rate": 3.691255698300871e-06, + "loss": 0.5387, + "step": 3159 + }, + { + "epoch": 0.2619146290924161, + "grad_norm": 0.26737889647483826, + "learning_rate": 3.6908412764193953e-06, + "loss": 0.5487, + "step": 3160 + }, + { + "epoch": 0.26199751346871114, + "grad_norm": 0.27600374817848206, + "learning_rate": 3.69042685453792e-06, + "loss": 0.5521, + "step": 3161 + }, + { + "epoch": 0.2620803978450062, + "grad_norm": 0.27743270993232727, + "learning_rate": 3.6900124326564446e-06, + "loss": 0.4885, + "step": 3162 + }, + { + "epoch": 0.2621632822213013, + "grad_norm": 0.26010194420814514, + "learning_rate": 3.689598010774969e-06, + "loss": 0.5057, + "step": 3163 + }, + { + "epoch": 0.26224616659759636, + "grad_norm": 0.23886984586715698, + "learning_rate": 3.689183588893494e-06, + "loss": 0.5023, + "step": 3164 + }, + { + "epoch": 0.2623290509738914, + "grad_norm": 0.2476005256175995, + "learning_rate": 3.6887691670120184e-06, + "loss": 0.5383, + "step": 3165 + }, + { + "epoch": 0.2624119353501865, + "grad_norm": 0.25278595089912415, + "learning_rate": 3.6883547451305428e-06, + "loss": 0.4857, + "step": 3166 + }, + { + "epoch": 0.26249481972648153, + "grad_norm": 0.2644527852535248, + "learning_rate": 3.687940323249068e-06, + "loss": 0.569, + "step": 3167 + }, + { + "epoch": 0.26257770410277664, + "grad_norm": 0.2576541006565094, + "learning_rate": 3.6875259013675925e-06, + "loss": 0.5521, + "step": 3168 + }, + { + "epoch": 0.2626605884790717, + "grad_norm": 0.25094401836395264, + "learning_rate": 3.6871114794861173e-06, + "loss": 0.517, + "step": 3169 + }, + { + "epoch": 0.26274347285536676, + "grad_norm": 0.2630549967288971, + "learning_rate": 3.6866970576046418e-06, + "loss": 0.5688, + "step": 3170 + }, + { + "epoch": 0.2628263572316618, + "grad_norm": 0.25869327783584595, + "learning_rate": 3.6862826357231662e-06, + "loss": 0.5934, + "step": 3171 + }, + { + "epoch": 0.2629092416079569, + "grad_norm": 0.2533027231693268, + "learning_rate": 3.685868213841691e-06, + "loss": 0.5403, + "step": 3172 + }, + { + "epoch": 0.262992125984252, + "grad_norm": 0.2664099633693695, + "learning_rate": 3.6854537919602155e-06, + "loss": 0.5521, + "step": 3173 + }, + { + "epoch": 0.26307501036054703, + "grad_norm": 0.2755780816078186, + "learning_rate": 3.6850393700787408e-06, + "loss": 0.5303, + "step": 3174 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 0.24543897807598114, + "learning_rate": 3.684624948197265e-06, + "loss": 0.5088, + "step": 3175 + }, + { + "epoch": 0.26324077911313715, + "grad_norm": 0.269807904958725, + "learning_rate": 3.6842105263157896e-06, + "loss": 0.5616, + "step": 3176 + }, + { + "epoch": 0.26332366348943226, + "grad_norm": 0.257368803024292, + "learning_rate": 3.6837961044343145e-06, + "loss": 0.5423, + "step": 3177 + }, + { + "epoch": 0.2634065478657273, + "grad_norm": 0.25569191575050354, + "learning_rate": 3.683381682552839e-06, + "loss": 0.5515, + "step": 3178 + }, + { + "epoch": 0.26348943224202237, + "grad_norm": 0.2704939842224121, + "learning_rate": 3.682967260671364e-06, + "loss": 0.5551, + "step": 3179 + }, + { + "epoch": 0.2635723166183174, + "grad_norm": 0.25408923625946045, + "learning_rate": 3.6825528387898886e-06, + "loss": 0.5191, + "step": 3180 + }, + { + "epoch": 0.26365520099461254, + "grad_norm": 0.2691519260406494, + "learning_rate": 3.682138416908413e-06, + "loss": 0.5552, + "step": 3181 + }, + { + "epoch": 0.2637380853709076, + "grad_norm": 0.28387781977653503, + "learning_rate": 3.681723995026938e-06, + "loss": 0.5354, + "step": 3182 + }, + { + "epoch": 0.26382096974720265, + "grad_norm": 0.249093696475029, + "learning_rate": 3.6813095731454624e-06, + "loss": 0.5285, + "step": 3183 + }, + { + "epoch": 0.2639038541234977, + "grad_norm": 0.2553102672100067, + "learning_rate": 3.680895151263987e-06, + "loss": 0.5221, + "step": 3184 + }, + { + "epoch": 0.2639867384997928, + "grad_norm": 0.330753356218338, + "learning_rate": 3.6804807293825116e-06, + "loss": 0.5419, + "step": 3185 + }, + { + "epoch": 0.2640696228760879, + "grad_norm": 0.2903355360031128, + "learning_rate": 3.680066307501036e-06, + "loss": 0.5363, + "step": 3186 + }, + { + "epoch": 0.26415250725238293, + "grad_norm": 0.2665586769580841, + "learning_rate": 3.6796518856195614e-06, + "loss": 0.5391, + "step": 3187 + }, + { + "epoch": 0.264235391628678, + "grad_norm": 0.2656952440738678, + "learning_rate": 3.6792374637380858e-06, + "loss": 0.5377, + "step": 3188 + }, + { + "epoch": 0.26431827600497304, + "grad_norm": 0.2722485065460205, + "learning_rate": 3.6788230418566102e-06, + "loss": 0.4867, + "step": 3189 + }, + { + "epoch": 0.26440116038126815, + "grad_norm": 0.24169817566871643, + "learning_rate": 3.678408619975135e-06, + "loss": 0.4874, + "step": 3190 + }, + { + "epoch": 0.2644840447575632, + "grad_norm": 0.2641121745109558, + "learning_rate": 3.6779941980936595e-06, + "loss": 0.5049, + "step": 3191 + }, + { + "epoch": 0.26456692913385826, + "grad_norm": 0.26329848170280457, + "learning_rate": 3.6775797762121844e-06, + "loss": 0.5582, + "step": 3192 + }, + { + "epoch": 0.2646498135101533, + "grad_norm": 0.2718062102794647, + "learning_rate": 3.677165354330709e-06, + "loss": 0.534, + "step": 3193 + }, + { + "epoch": 0.26473269788644843, + "grad_norm": 0.2664894461631775, + "learning_rate": 3.6767509324492332e-06, + "loss": 0.5283, + "step": 3194 + }, + { + "epoch": 0.2648155822627435, + "grad_norm": 0.26946836709976196, + "learning_rate": 3.6763365105677585e-06, + "loss": 0.5282, + "step": 3195 + }, + { + "epoch": 0.26489846663903854, + "grad_norm": 0.260154128074646, + "learning_rate": 3.675922088686283e-06, + "loss": 0.5468, + "step": 3196 + }, + { + "epoch": 0.2649813510153336, + "grad_norm": 0.28437334299087524, + "learning_rate": 3.675507666804808e-06, + "loss": 0.5632, + "step": 3197 + }, + { + "epoch": 0.26506423539162866, + "grad_norm": 0.26601502299308777, + "learning_rate": 3.6750932449233322e-06, + "loss": 0.5327, + "step": 3198 + }, + { + "epoch": 0.26514711976792377, + "grad_norm": 0.23289211094379425, + "learning_rate": 3.6746788230418567e-06, + "loss": 0.4725, + "step": 3199 + }, + { + "epoch": 0.2652300041442188, + "grad_norm": 0.25009438395500183, + "learning_rate": 3.674264401160382e-06, + "loss": 0.4909, + "step": 3200 + }, + { + "epoch": 0.2653128885205139, + "grad_norm": 0.2572276294231415, + "learning_rate": 3.6738499792789064e-06, + "loss": 0.535, + "step": 3201 + }, + { + "epoch": 0.26539577289680893, + "grad_norm": 0.2730824649333954, + "learning_rate": 3.673435557397431e-06, + "loss": 0.5099, + "step": 3202 + }, + { + "epoch": 0.26547865727310405, + "grad_norm": 0.2734759747982025, + "learning_rate": 3.6730211355159557e-06, + "loss": 0.5457, + "step": 3203 + }, + { + "epoch": 0.2655615416493991, + "grad_norm": 0.2571609616279602, + "learning_rate": 3.67260671363448e-06, + "loss": 0.5485, + "step": 3204 + }, + { + "epoch": 0.26564442602569416, + "grad_norm": 0.27025526762008667, + "learning_rate": 3.672192291753005e-06, + "loss": 0.5924, + "step": 3205 + }, + { + "epoch": 0.2657273104019892, + "grad_norm": 0.2706720232963562, + "learning_rate": 3.6717778698715294e-06, + "loss": 0.5704, + "step": 3206 + }, + { + "epoch": 0.26581019477828427, + "grad_norm": 0.31309282779693604, + "learning_rate": 3.671363447990054e-06, + "loss": 0.5969, + "step": 3207 + }, + { + "epoch": 0.2658930791545794, + "grad_norm": 0.2969683110713959, + "learning_rate": 3.670949026108579e-06, + "loss": 0.5419, + "step": 3208 + }, + { + "epoch": 0.26597596353087444, + "grad_norm": 0.2661495804786682, + "learning_rate": 3.6705346042271035e-06, + "loss": 0.531, + "step": 3209 + }, + { + "epoch": 0.2660588479071695, + "grad_norm": 0.25665903091430664, + "learning_rate": 3.6701201823456284e-06, + "loss": 0.5226, + "step": 3210 + }, + { + "epoch": 0.26614173228346455, + "grad_norm": 0.26062047481536865, + "learning_rate": 3.669705760464153e-06, + "loss": 0.5159, + "step": 3211 + }, + { + "epoch": 0.26622461665975966, + "grad_norm": 0.654873251914978, + "learning_rate": 3.6692913385826772e-06, + "loss": 0.5513, + "step": 3212 + }, + { + "epoch": 0.2663075010360547, + "grad_norm": 0.25731515884399414, + "learning_rate": 3.668876916701202e-06, + "loss": 0.5206, + "step": 3213 + }, + { + "epoch": 0.2663903854123498, + "grad_norm": 0.27909332513809204, + "learning_rate": 3.6684624948197265e-06, + "loss": 0.5542, + "step": 3214 + }, + { + "epoch": 0.26647326978864483, + "grad_norm": 0.29114291071891785, + "learning_rate": 3.668048072938252e-06, + "loss": 0.5478, + "step": 3215 + }, + { + "epoch": 0.2665561541649399, + "grad_norm": 0.259702205657959, + "learning_rate": 3.6676336510567762e-06, + "loss": 0.5377, + "step": 3216 + }, + { + "epoch": 0.266639038541235, + "grad_norm": 0.28196263313293457, + "learning_rate": 3.6672192291753007e-06, + "loss": 0.5827, + "step": 3217 + }, + { + "epoch": 0.26672192291753005, + "grad_norm": 0.2827322781085968, + "learning_rate": 3.6668048072938255e-06, + "loss": 0.5527, + "step": 3218 + }, + { + "epoch": 0.2668048072938251, + "grad_norm": 0.25926443934440613, + "learning_rate": 3.66639038541235e-06, + "loss": 0.4986, + "step": 3219 + }, + { + "epoch": 0.26688769167012016, + "grad_norm": 0.24585379660129547, + "learning_rate": 3.6659759635308744e-06, + "loss": 0.5, + "step": 3220 + }, + { + "epoch": 0.2669705760464153, + "grad_norm": 0.2718033194541931, + "learning_rate": 3.6655615416493997e-06, + "loss": 0.5356, + "step": 3221 + }, + { + "epoch": 0.26705346042271033, + "grad_norm": 0.2727328836917877, + "learning_rate": 3.665147119767924e-06, + "loss": 0.5262, + "step": 3222 + }, + { + "epoch": 0.2671363447990054, + "grad_norm": 0.25982099771499634, + "learning_rate": 3.664732697886449e-06, + "loss": 0.4971, + "step": 3223 + }, + { + "epoch": 0.26721922917530044, + "grad_norm": 0.3812170624732971, + "learning_rate": 3.6643182760049734e-06, + "loss": 0.5387, + "step": 3224 + }, + { + "epoch": 0.2673021135515955, + "grad_norm": 0.28327298164367676, + "learning_rate": 3.663903854123498e-06, + "loss": 0.541, + "step": 3225 + }, + { + "epoch": 0.2673849979278906, + "grad_norm": 0.27957025170326233, + "learning_rate": 3.6634894322420227e-06, + "loss": 0.5524, + "step": 3226 + }, + { + "epoch": 0.26746788230418567, + "grad_norm": 0.27696070075035095, + "learning_rate": 3.663075010360547e-06, + "loss": 0.5697, + "step": 3227 + }, + { + "epoch": 0.2675507666804807, + "grad_norm": 0.27289557456970215, + "learning_rate": 3.6626605884790724e-06, + "loss": 0.5607, + "step": 3228 + }, + { + "epoch": 0.2676336510567758, + "grad_norm": 0.2810351848602295, + "learning_rate": 3.662246166597597e-06, + "loss": 0.5853, + "step": 3229 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 0.26448026299476624, + "learning_rate": 3.6618317447161212e-06, + "loss": 0.5262, + "step": 3230 + }, + { + "epoch": 0.26779941980936595, + "grad_norm": 0.2581854462623596, + "learning_rate": 3.661417322834646e-06, + "loss": 0.5688, + "step": 3231 + }, + { + "epoch": 0.267882304185661, + "grad_norm": 0.267996221780777, + "learning_rate": 3.6610029009531705e-06, + "loss": 0.5317, + "step": 3232 + }, + { + "epoch": 0.26796518856195606, + "grad_norm": 0.27860796451568604, + "learning_rate": 3.660588479071695e-06, + "loss": 0.5274, + "step": 3233 + }, + { + "epoch": 0.2680480729382511, + "grad_norm": 0.2650493085384369, + "learning_rate": 3.66017405719022e-06, + "loss": 0.5514, + "step": 3234 + }, + { + "epoch": 0.2681309573145462, + "grad_norm": 0.24025940895080566, + "learning_rate": 3.6597596353087442e-06, + "loss": 0.5089, + "step": 3235 + }, + { + "epoch": 0.2682138416908413, + "grad_norm": 0.2596697211265564, + "learning_rate": 3.6593452134272695e-06, + "loss": 0.5385, + "step": 3236 + }, + { + "epoch": 0.26829672606713634, + "grad_norm": 0.2638474404811859, + "learning_rate": 3.658930791545794e-06, + "loss": 0.5345, + "step": 3237 + }, + { + "epoch": 0.2683796104434314, + "grad_norm": 0.2426549643278122, + "learning_rate": 3.6585163696643184e-06, + "loss": 0.5339, + "step": 3238 + }, + { + "epoch": 0.2684624948197265, + "grad_norm": 0.2972433567047119, + "learning_rate": 3.6581019477828432e-06, + "loss": 0.5596, + "step": 3239 + }, + { + "epoch": 0.26854537919602156, + "grad_norm": 0.26828065514564514, + "learning_rate": 3.6576875259013677e-06, + "loss": 0.5524, + "step": 3240 + }, + { + "epoch": 0.2686282635723166, + "grad_norm": 0.29720404744148254, + "learning_rate": 3.657273104019893e-06, + "loss": 0.5527, + "step": 3241 + }, + { + "epoch": 0.2687111479486117, + "grad_norm": 0.27918171882629395, + "learning_rate": 3.6568586821384174e-06, + "loss": 0.5599, + "step": 3242 + }, + { + "epoch": 0.26879403232490673, + "grad_norm": 0.2612060010433197, + "learning_rate": 3.656444260256942e-06, + "loss": 0.4965, + "step": 3243 + }, + { + "epoch": 0.26887691670120184, + "grad_norm": 0.27383488416671753, + "learning_rate": 3.6560298383754667e-06, + "loss": 0.5348, + "step": 3244 + }, + { + "epoch": 0.2689598010774969, + "grad_norm": 0.3119625747203827, + "learning_rate": 3.655615416493991e-06, + "loss": 0.6097, + "step": 3245 + }, + { + "epoch": 0.26904268545379195, + "grad_norm": 0.2668245732784271, + "learning_rate": 3.655200994612516e-06, + "loss": 0.497, + "step": 3246 + }, + { + "epoch": 0.269125569830087, + "grad_norm": 0.2697283923625946, + "learning_rate": 3.6547865727310404e-06, + "loss": 0.5143, + "step": 3247 + }, + { + "epoch": 0.2692084542063821, + "grad_norm": 0.305082231760025, + "learning_rate": 3.654372150849565e-06, + "loss": 0.5891, + "step": 3248 + }, + { + "epoch": 0.2692913385826772, + "grad_norm": 0.24437610805034637, + "learning_rate": 3.65395772896809e-06, + "loss": 0.506, + "step": 3249 + }, + { + "epoch": 0.26937422295897223, + "grad_norm": 0.2590849995613098, + "learning_rate": 3.6535433070866145e-06, + "loss": 0.5575, + "step": 3250 + }, + { + "epoch": 0.2694571073352673, + "grad_norm": 0.24600419402122498, + "learning_rate": 3.653128885205139e-06, + "loss": 0.5206, + "step": 3251 + }, + { + "epoch": 0.26953999171156234, + "grad_norm": 0.23583145439624786, + "learning_rate": 3.652714463323664e-06, + "loss": 0.5343, + "step": 3252 + }, + { + "epoch": 0.26962287608785745, + "grad_norm": 0.24866090714931488, + "learning_rate": 3.6523000414421883e-06, + "loss": 0.5515, + "step": 3253 + }, + { + "epoch": 0.2697057604641525, + "grad_norm": 0.28021690249443054, + "learning_rate": 3.651885619560713e-06, + "loss": 0.5468, + "step": 3254 + }, + { + "epoch": 0.26978864484044757, + "grad_norm": 0.24913452565670013, + "learning_rate": 3.6514711976792375e-06, + "loss": 0.5587, + "step": 3255 + }, + { + "epoch": 0.2698715292167426, + "grad_norm": 0.25216925144195557, + "learning_rate": 3.6510567757977624e-06, + "loss": 0.5659, + "step": 3256 + }, + { + "epoch": 0.26995441359303773, + "grad_norm": 0.25497540831565857, + "learning_rate": 3.6506423539162873e-06, + "loss": 0.5202, + "step": 3257 + }, + { + "epoch": 0.2700372979693328, + "grad_norm": 0.264384925365448, + "learning_rate": 3.6502279320348117e-06, + "loss": 0.5476, + "step": 3258 + }, + { + "epoch": 0.27012018234562785, + "grad_norm": 0.28308188915252686, + "learning_rate": 3.6498135101533365e-06, + "loss": 0.6087, + "step": 3259 + }, + { + "epoch": 0.2702030667219229, + "grad_norm": 0.24286812543869019, + "learning_rate": 3.649399088271861e-06, + "loss": 0.5018, + "step": 3260 + }, + { + "epoch": 0.270285951098218, + "grad_norm": 0.2650483548641205, + "learning_rate": 3.6489846663903854e-06, + "loss": 0.5535, + "step": 3261 + }, + { + "epoch": 0.27036883547451307, + "grad_norm": 0.29214850068092346, + "learning_rate": 3.6485702445089107e-06, + "loss": 0.5873, + "step": 3262 + }, + { + "epoch": 0.2704517198508081, + "grad_norm": 0.29781123995780945, + "learning_rate": 3.648155822627435e-06, + "loss": 0.5305, + "step": 3263 + }, + { + "epoch": 0.2705346042271032, + "grad_norm": 0.2612508535385132, + "learning_rate": 3.64774140074596e-06, + "loss": 0.5225, + "step": 3264 + }, + { + "epoch": 0.27061748860339824, + "grad_norm": 0.24499386548995972, + "learning_rate": 3.6473269788644844e-06, + "loss": 0.4903, + "step": 3265 + }, + { + "epoch": 0.27070037297969335, + "grad_norm": 0.24758480489253998, + "learning_rate": 3.646912556983009e-06, + "loss": 0.5461, + "step": 3266 + }, + { + "epoch": 0.2707832573559884, + "grad_norm": 0.27177053689956665, + "learning_rate": 3.6464981351015337e-06, + "loss": 0.5148, + "step": 3267 + }, + { + "epoch": 0.27086614173228346, + "grad_norm": 0.25692319869995117, + "learning_rate": 3.646083713220058e-06, + "loss": 0.5207, + "step": 3268 + }, + { + "epoch": 0.2709490261085785, + "grad_norm": 0.28959211707115173, + "learning_rate": 3.6456692913385826e-06, + "loss": 0.5441, + "step": 3269 + }, + { + "epoch": 0.27103191048487363, + "grad_norm": 0.2557172477245331, + "learning_rate": 3.645254869457108e-06, + "loss": 0.5375, + "step": 3270 + }, + { + "epoch": 0.2711147948611687, + "grad_norm": 0.2605079114437103, + "learning_rate": 3.6448404475756323e-06, + "loss": 0.5057, + "step": 3271 + }, + { + "epoch": 0.27119767923746374, + "grad_norm": 0.2476402223110199, + "learning_rate": 3.644426025694157e-06, + "loss": 0.5204, + "step": 3272 + }, + { + "epoch": 0.2712805636137588, + "grad_norm": 0.24936015903949738, + "learning_rate": 3.6440116038126815e-06, + "loss": 0.5262, + "step": 3273 + }, + { + "epoch": 0.27136344799005385, + "grad_norm": 0.2748275697231293, + "learning_rate": 3.643597181931206e-06, + "loss": 0.5403, + "step": 3274 + }, + { + "epoch": 0.27144633236634896, + "grad_norm": 0.28269127011299133, + "learning_rate": 3.643182760049731e-06, + "loss": 0.5721, + "step": 3275 + }, + { + "epoch": 0.271529216742644, + "grad_norm": 0.2985675632953644, + "learning_rate": 3.6427683381682557e-06, + "loss": 0.5961, + "step": 3276 + }, + { + "epoch": 0.2716121011189391, + "grad_norm": 0.2531624436378479, + "learning_rate": 3.6423539162867805e-06, + "loss": 0.538, + "step": 3277 + }, + { + "epoch": 0.27169498549523413, + "grad_norm": 0.28852277994155884, + "learning_rate": 3.641939494405305e-06, + "loss": 0.5419, + "step": 3278 + }, + { + "epoch": 0.27177786987152924, + "grad_norm": 0.24205638468265533, + "learning_rate": 3.6415250725238294e-06, + "loss": 0.4786, + "step": 3279 + }, + { + "epoch": 0.2718607542478243, + "grad_norm": 0.2584994435310364, + "learning_rate": 3.6411106506423543e-06, + "loss": 0.5363, + "step": 3280 + }, + { + "epoch": 0.27194363862411935, + "grad_norm": 0.6588530540466309, + "learning_rate": 3.6406962287608787e-06, + "loss": 0.5781, + "step": 3281 + }, + { + "epoch": 0.2720265230004144, + "grad_norm": 0.24518026411533356, + "learning_rate": 3.640281806879404e-06, + "loss": 0.4842, + "step": 3282 + }, + { + "epoch": 0.27210940737670947, + "grad_norm": 0.2902783453464508, + "learning_rate": 3.6398673849979284e-06, + "loss": 0.5548, + "step": 3283 + }, + { + "epoch": 0.2721922917530046, + "grad_norm": 0.25223594903945923, + "learning_rate": 3.639452963116453e-06, + "loss": 0.477, + "step": 3284 + }, + { + "epoch": 0.27227517612929963, + "grad_norm": 0.26571986079216003, + "learning_rate": 3.6390385412349777e-06, + "loss": 0.5635, + "step": 3285 + }, + { + "epoch": 0.2723580605055947, + "grad_norm": 0.271987646818161, + "learning_rate": 3.638624119353502e-06, + "loss": 0.5502, + "step": 3286 + }, + { + "epoch": 0.27244094488188975, + "grad_norm": 0.2799866199493408, + "learning_rate": 3.6382096974720266e-06, + "loss": 0.59, + "step": 3287 + }, + { + "epoch": 0.27252382925818486, + "grad_norm": 0.2561870217323303, + "learning_rate": 3.6377952755905514e-06, + "loss": 0.5436, + "step": 3288 + }, + { + "epoch": 0.2726067136344799, + "grad_norm": 0.25710529088974, + "learning_rate": 3.637380853709076e-06, + "loss": 0.5181, + "step": 3289 + }, + { + "epoch": 0.27268959801077497, + "grad_norm": 0.2455517202615738, + "learning_rate": 3.636966431827601e-06, + "loss": 0.5335, + "step": 3290 + }, + { + "epoch": 0.27277248238707, + "grad_norm": 0.2689892053604126, + "learning_rate": 3.6365520099461256e-06, + "loss": 0.5241, + "step": 3291 + }, + { + "epoch": 0.2728553667633651, + "grad_norm": 0.29115694761276245, + "learning_rate": 3.63613758806465e-06, + "loss": 0.5666, + "step": 3292 + }, + { + "epoch": 0.2729382511396602, + "grad_norm": 0.26688170433044434, + "learning_rate": 3.635723166183175e-06, + "loss": 0.528, + "step": 3293 + }, + { + "epoch": 0.27302113551595525, + "grad_norm": 0.2663537263870239, + "learning_rate": 3.6353087443016993e-06, + "loss": 0.5482, + "step": 3294 + }, + { + "epoch": 0.2731040198922503, + "grad_norm": 0.27020344138145447, + "learning_rate": 3.634894322420224e-06, + "loss": 0.5607, + "step": 3295 + }, + { + "epoch": 0.27318690426854536, + "grad_norm": 0.2746204435825348, + "learning_rate": 3.6344799005387486e-06, + "loss": 0.5367, + "step": 3296 + }, + { + "epoch": 0.2732697886448405, + "grad_norm": 0.2684139311313629, + "learning_rate": 3.6340654786572734e-06, + "loss": 0.5522, + "step": 3297 + }, + { + "epoch": 0.27335267302113553, + "grad_norm": 0.27451443672180176, + "learning_rate": 3.6336510567757983e-06, + "loss": 0.5346, + "step": 3298 + }, + { + "epoch": 0.2734355573974306, + "grad_norm": 0.2861509323120117, + "learning_rate": 3.6332366348943227e-06, + "loss": 0.5174, + "step": 3299 + }, + { + "epoch": 0.27351844177372564, + "grad_norm": 0.26436367630958557, + "learning_rate": 3.632822213012847e-06, + "loss": 0.569, + "step": 3300 + }, + { + "epoch": 0.2736013261500207, + "grad_norm": 0.2814600467681885, + "learning_rate": 3.632407791131372e-06, + "loss": 0.5525, + "step": 3301 + }, + { + "epoch": 0.2736842105263158, + "grad_norm": 0.2610165774822235, + "learning_rate": 3.6319933692498964e-06, + "loss": 0.5527, + "step": 3302 + }, + { + "epoch": 0.27376709490261086, + "grad_norm": 0.28233882784843445, + "learning_rate": 3.6315789473684217e-06, + "loss": 0.5445, + "step": 3303 + }, + { + "epoch": 0.2738499792789059, + "grad_norm": 0.24947503209114075, + "learning_rate": 3.631164525486946e-06, + "loss": 0.5513, + "step": 3304 + }, + { + "epoch": 0.273932863655201, + "grad_norm": 0.2667756676673889, + "learning_rate": 3.6307501036054706e-06, + "loss": 0.5393, + "step": 3305 + }, + { + "epoch": 0.2740157480314961, + "grad_norm": 0.2574373483657837, + "learning_rate": 3.6303356817239954e-06, + "loss": 0.5568, + "step": 3306 + }, + { + "epoch": 0.27409863240779114, + "grad_norm": 0.2513679563999176, + "learning_rate": 3.62992125984252e-06, + "loss": 0.5257, + "step": 3307 + }, + { + "epoch": 0.2741815167840862, + "grad_norm": 0.2827610969543457, + "learning_rate": 3.6295068379610447e-06, + "loss": 0.5764, + "step": 3308 + }, + { + "epoch": 0.27426440116038125, + "grad_norm": 0.2503167986869812, + "learning_rate": 3.629092416079569e-06, + "loss": 0.4938, + "step": 3309 + }, + { + "epoch": 0.2743472855366763, + "grad_norm": 0.2770426571369171, + "learning_rate": 3.6286779941980936e-06, + "loss": 0.5255, + "step": 3310 + }, + { + "epoch": 0.2744301699129714, + "grad_norm": 0.2629266083240509, + "learning_rate": 3.628263572316619e-06, + "loss": 0.5593, + "step": 3311 + }, + { + "epoch": 0.2745130542892665, + "grad_norm": 0.2773684561252594, + "learning_rate": 3.6278491504351433e-06, + "loss": 0.5291, + "step": 3312 + }, + { + "epoch": 0.27459593866556153, + "grad_norm": 0.274880051612854, + "learning_rate": 3.627434728553668e-06, + "loss": 0.5094, + "step": 3313 + }, + { + "epoch": 0.2746788230418566, + "grad_norm": 0.27290043234825134, + "learning_rate": 3.6270203066721926e-06, + "loss": 0.5339, + "step": 3314 + }, + { + "epoch": 0.2747617074181517, + "grad_norm": 0.2647908926010132, + "learning_rate": 3.626605884790717e-06, + "loss": 0.5162, + "step": 3315 + }, + { + "epoch": 0.27484459179444676, + "grad_norm": 0.26653990149497986, + "learning_rate": 3.626191462909242e-06, + "loss": 0.5308, + "step": 3316 + }, + { + "epoch": 0.2749274761707418, + "grad_norm": 0.26479971408843994, + "learning_rate": 3.6257770410277667e-06, + "loss": 0.55, + "step": 3317 + }, + { + "epoch": 0.27501036054703687, + "grad_norm": 0.25940966606140137, + "learning_rate": 3.625362619146291e-06, + "loss": 0.5385, + "step": 3318 + }, + { + "epoch": 0.2750932449233319, + "grad_norm": 0.267103374004364, + "learning_rate": 3.624948197264816e-06, + "loss": 0.5142, + "step": 3319 + }, + { + "epoch": 0.27517612929962704, + "grad_norm": 0.28381460905075073, + "learning_rate": 3.6245337753833404e-06, + "loss": 0.5232, + "step": 3320 + }, + { + "epoch": 0.2752590136759221, + "grad_norm": 0.29292982816696167, + "learning_rate": 3.6241193535018653e-06, + "loss": 0.5632, + "step": 3321 + }, + { + "epoch": 0.27534189805221715, + "grad_norm": 0.26223093271255493, + "learning_rate": 3.6237049316203897e-06, + "loss": 0.5072, + "step": 3322 + }, + { + "epoch": 0.2754247824285122, + "grad_norm": 0.2742306888103485, + "learning_rate": 3.623290509738914e-06, + "loss": 0.5333, + "step": 3323 + }, + { + "epoch": 0.2755076668048073, + "grad_norm": 0.28094029426574707, + "learning_rate": 3.6228760878574394e-06, + "loss": 0.5505, + "step": 3324 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 0.2863244414329529, + "learning_rate": 3.622461665975964e-06, + "loss": 0.5718, + "step": 3325 + }, + { + "epoch": 0.27567343555739743, + "grad_norm": 0.29289743304252625, + "learning_rate": 3.6220472440944887e-06, + "loss": 0.5702, + "step": 3326 + }, + { + "epoch": 0.2757563199336925, + "grad_norm": 0.253286212682724, + "learning_rate": 3.621632822213013e-06, + "loss": 0.4892, + "step": 3327 + }, + { + "epoch": 0.27583920430998754, + "grad_norm": 0.2703787088394165, + "learning_rate": 3.6212184003315376e-06, + "loss": 0.5247, + "step": 3328 + }, + { + "epoch": 0.27592208868628265, + "grad_norm": 0.2757149338722229, + "learning_rate": 3.6208039784500624e-06, + "loss": 0.5765, + "step": 3329 + }, + { + "epoch": 0.2760049730625777, + "grad_norm": 0.27984195947647095, + "learning_rate": 3.620389556568587e-06, + "loss": 0.5377, + "step": 3330 + }, + { + "epoch": 0.27608785743887276, + "grad_norm": 0.2550954818725586, + "learning_rate": 3.619975134687112e-06, + "loss": 0.5278, + "step": 3331 + }, + { + "epoch": 0.2761707418151678, + "grad_norm": 0.24379074573516846, + "learning_rate": 3.6195607128056366e-06, + "loss": 0.5209, + "step": 3332 + }, + { + "epoch": 0.27625362619146293, + "grad_norm": 0.24554383754730225, + "learning_rate": 3.619146290924161e-06, + "loss": 0.5167, + "step": 3333 + }, + { + "epoch": 0.276336510567758, + "grad_norm": 0.2559818625450134, + "learning_rate": 3.618731869042686e-06, + "loss": 0.5076, + "step": 3334 + }, + { + "epoch": 0.27641939494405304, + "grad_norm": 0.2840614914894104, + "learning_rate": 3.6183174471612103e-06, + "loss": 0.5882, + "step": 3335 + }, + { + "epoch": 0.2765022793203481, + "grad_norm": 0.25210943818092346, + "learning_rate": 3.6179030252797347e-06, + "loss": 0.5508, + "step": 3336 + }, + { + "epoch": 0.2765851636966432, + "grad_norm": 0.24920335412025452, + "learning_rate": 3.61748860339826e-06, + "loss": 0.5071, + "step": 3337 + }, + { + "epoch": 0.27666804807293827, + "grad_norm": 0.24258174002170563, + "learning_rate": 3.6170741815167844e-06, + "loss": 0.524, + "step": 3338 + }, + { + "epoch": 0.2767509324492333, + "grad_norm": 0.2744062840938568, + "learning_rate": 3.6166597596353093e-06, + "loss": 0.5737, + "step": 3339 + }, + { + "epoch": 0.2768338168255284, + "grad_norm": 0.278959721326828, + "learning_rate": 3.6162453377538337e-06, + "loss": 0.5288, + "step": 3340 + }, + { + "epoch": 0.27691670120182343, + "grad_norm": 0.25274500250816345, + "learning_rate": 3.615830915872358e-06, + "loss": 0.5361, + "step": 3341 + }, + { + "epoch": 0.27699958557811855, + "grad_norm": 0.2722002863883972, + "learning_rate": 3.615416493990883e-06, + "loss": 0.5015, + "step": 3342 + }, + { + "epoch": 0.2770824699544136, + "grad_norm": 0.25043874979019165, + "learning_rate": 3.6150020721094074e-06, + "loss": 0.5151, + "step": 3343 + }, + { + "epoch": 0.27716535433070866, + "grad_norm": 0.26863259077072144, + "learning_rate": 3.6145876502279327e-06, + "loss": 0.5181, + "step": 3344 + }, + { + "epoch": 0.2772482387070037, + "grad_norm": 0.2635591924190521, + "learning_rate": 3.614173228346457e-06, + "loss": 0.5399, + "step": 3345 + }, + { + "epoch": 0.2773311230832988, + "grad_norm": 0.26627859473228455, + "learning_rate": 3.6137588064649816e-06, + "loss": 0.5605, + "step": 3346 + }, + { + "epoch": 0.2774140074595939, + "grad_norm": 0.26863381266593933, + "learning_rate": 3.6133443845835064e-06, + "loss": 0.5024, + "step": 3347 + }, + { + "epoch": 0.27749689183588894, + "grad_norm": 0.2648603022098541, + "learning_rate": 3.612929962702031e-06, + "loss": 0.5288, + "step": 3348 + }, + { + "epoch": 0.277579776212184, + "grad_norm": 0.2515513300895691, + "learning_rate": 3.6125155408205557e-06, + "loss": 0.5571, + "step": 3349 + }, + { + "epoch": 0.27766266058847905, + "grad_norm": 0.28456535935401917, + "learning_rate": 3.61210111893908e-06, + "loss": 0.5232, + "step": 3350 + }, + { + "epoch": 0.27774554496477416, + "grad_norm": 0.26049545407295227, + "learning_rate": 3.6116866970576046e-06, + "loss": 0.5621, + "step": 3351 + }, + { + "epoch": 0.2778284293410692, + "grad_norm": 0.2720368504524231, + "learning_rate": 3.61127227517613e-06, + "loss": 0.5336, + "step": 3352 + }, + { + "epoch": 0.2779113137173643, + "grad_norm": 0.2629505693912506, + "learning_rate": 3.6108578532946543e-06, + "loss": 0.5397, + "step": 3353 + }, + { + "epoch": 0.27799419809365933, + "grad_norm": 0.25255706906318665, + "learning_rate": 3.6104434314131787e-06, + "loss": 0.5153, + "step": 3354 + }, + { + "epoch": 0.27807708246995444, + "grad_norm": 0.2505296766757965, + "learning_rate": 3.6100290095317036e-06, + "loss": 0.5229, + "step": 3355 + }, + { + "epoch": 0.2781599668462495, + "grad_norm": 0.2885300815105438, + "learning_rate": 3.609614587650228e-06, + "loss": 0.5662, + "step": 3356 + }, + { + "epoch": 0.27824285122254455, + "grad_norm": 0.2607544958591461, + "learning_rate": 3.6092001657687533e-06, + "loss": 0.5459, + "step": 3357 + }, + { + "epoch": 0.2783257355988396, + "grad_norm": 0.2904371917247772, + "learning_rate": 3.6087857438872777e-06, + "loss": 0.5477, + "step": 3358 + }, + { + "epoch": 0.27840861997513466, + "grad_norm": 0.24505092203617096, + "learning_rate": 3.608371322005802e-06, + "loss": 0.5124, + "step": 3359 + }, + { + "epoch": 0.2784915043514298, + "grad_norm": 0.2754078209400177, + "learning_rate": 3.607956900124327e-06, + "loss": 0.5118, + "step": 3360 + }, + { + "epoch": 0.27857438872772483, + "grad_norm": 0.2715921700000763, + "learning_rate": 3.6075424782428515e-06, + "loss": 0.5415, + "step": 3361 + }, + { + "epoch": 0.2786572731040199, + "grad_norm": 0.2616950571537018, + "learning_rate": 3.6071280563613763e-06, + "loss": 0.4882, + "step": 3362 + }, + { + "epoch": 0.27874015748031494, + "grad_norm": 0.2766316831111908, + "learning_rate": 3.6067136344799007e-06, + "loss": 0.586, + "step": 3363 + }, + { + "epoch": 0.27882304185661005, + "grad_norm": 0.25219428539276123, + "learning_rate": 3.606299212598425e-06, + "loss": 0.5507, + "step": 3364 + }, + { + "epoch": 0.2789059262329051, + "grad_norm": 0.25201138854026794, + "learning_rate": 3.6058847907169504e-06, + "loss": 0.5328, + "step": 3365 + }, + { + "epoch": 0.27898881060920017, + "grad_norm": 0.265338659286499, + "learning_rate": 3.605470368835475e-06, + "loss": 0.5512, + "step": 3366 + }, + { + "epoch": 0.2790716949854952, + "grad_norm": 0.2986564040184021, + "learning_rate": 3.6050559469539997e-06, + "loss": 0.5622, + "step": 3367 + }, + { + "epoch": 0.2791545793617903, + "grad_norm": 0.28237733244895935, + "learning_rate": 3.604641525072524e-06, + "loss": 0.534, + "step": 3368 + }, + { + "epoch": 0.2792374637380854, + "grad_norm": 0.2498088777065277, + "learning_rate": 3.6042271031910486e-06, + "loss": 0.5317, + "step": 3369 + }, + { + "epoch": 0.27932034811438045, + "grad_norm": 0.24229489266872406, + "learning_rate": 3.6038126813095735e-06, + "loss": 0.5544, + "step": 3370 + }, + { + "epoch": 0.2794032324906755, + "grad_norm": 0.2673484683036804, + "learning_rate": 3.603398259428098e-06, + "loss": 0.5471, + "step": 3371 + }, + { + "epoch": 0.27948611686697056, + "grad_norm": 0.2519961893558502, + "learning_rate": 3.6029838375466223e-06, + "loss": 0.5146, + "step": 3372 + }, + { + "epoch": 0.27956900124326567, + "grad_norm": 0.29486083984375, + "learning_rate": 3.6025694156651476e-06, + "loss": 0.5112, + "step": 3373 + }, + { + "epoch": 0.2796518856195607, + "grad_norm": 0.26752975583076477, + "learning_rate": 3.602154993783672e-06, + "loss": 0.5509, + "step": 3374 + }, + { + "epoch": 0.2797347699958558, + "grad_norm": 0.2460704892873764, + "learning_rate": 3.601740571902197e-06, + "loss": 0.5033, + "step": 3375 + }, + { + "epoch": 0.27981765437215084, + "grad_norm": 0.2544277310371399, + "learning_rate": 3.6013261500207213e-06, + "loss": 0.5494, + "step": 3376 + }, + { + "epoch": 0.2799005387484459, + "grad_norm": 0.2617429494857788, + "learning_rate": 3.6009117281392457e-06, + "loss": 0.5435, + "step": 3377 + }, + { + "epoch": 0.279983423124741, + "grad_norm": 0.2740265429019928, + "learning_rate": 3.600497306257771e-06, + "loss": 0.5341, + "step": 3378 + }, + { + "epoch": 0.28006630750103606, + "grad_norm": 0.265863835811615, + "learning_rate": 3.6000828843762955e-06, + "loss": 0.5181, + "step": 3379 + }, + { + "epoch": 0.2801491918773311, + "grad_norm": 0.2806549668312073, + "learning_rate": 3.5996684624948203e-06, + "loss": 0.5955, + "step": 3380 + }, + { + "epoch": 0.28023207625362617, + "grad_norm": 0.25572532415390015, + "learning_rate": 3.5992540406133447e-06, + "loss": 0.529, + "step": 3381 + }, + { + "epoch": 0.2803149606299213, + "grad_norm": 0.26112398505210876, + "learning_rate": 3.598839618731869e-06, + "loss": 0.5497, + "step": 3382 + }, + { + "epoch": 0.28039784500621634, + "grad_norm": 0.2814580500125885, + "learning_rate": 3.598425196850394e-06, + "loss": 0.5503, + "step": 3383 + }, + { + "epoch": 0.2804807293825114, + "grad_norm": 0.25916147232055664, + "learning_rate": 3.5980107749689185e-06, + "loss": 0.5164, + "step": 3384 + }, + { + "epoch": 0.28056361375880645, + "grad_norm": 0.2715447247028351, + "learning_rate": 3.597596353087443e-06, + "loss": 0.5516, + "step": 3385 + }, + { + "epoch": 0.2806464981351015, + "grad_norm": 0.2713638246059418, + "learning_rate": 3.597181931205968e-06, + "loss": 0.529, + "step": 3386 + }, + { + "epoch": 0.2807293825113966, + "grad_norm": 0.2618071138858795, + "learning_rate": 3.5967675093244926e-06, + "loss": 0.5424, + "step": 3387 + }, + { + "epoch": 0.2808122668876917, + "grad_norm": 0.2781781852245331, + "learning_rate": 3.5963530874430175e-06, + "loss": 0.5232, + "step": 3388 + }, + { + "epoch": 0.28089515126398673, + "grad_norm": 0.2917286455631256, + "learning_rate": 3.595938665561542e-06, + "loss": 0.5637, + "step": 3389 + }, + { + "epoch": 0.2809780356402818, + "grad_norm": 0.2579343020915985, + "learning_rate": 3.5955242436800663e-06, + "loss": 0.5381, + "step": 3390 + }, + { + "epoch": 0.2810609200165769, + "grad_norm": 0.25213122367858887, + "learning_rate": 3.595109821798591e-06, + "loss": 0.5376, + "step": 3391 + }, + { + "epoch": 0.28114380439287195, + "grad_norm": 0.23439733684062958, + "learning_rate": 3.5946953999171156e-06, + "loss": 0.4801, + "step": 3392 + }, + { + "epoch": 0.281226688769167, + "grad_norm": 0.25374144315719604, + "learning_rate": 3.594280978035641e-06, + "loss": 0.5462, + "step": 3393 + }, + { + "epoch": 0.28130957314546207, + "grad_norm": 0.27467337250709534, + "learning_rate": 3.5938665561541653e-06, + "loss": 0.6024, + "step": 3394 + }, + { + "epoch": 0.2813924575217571, + "grad_norm": 0.2660697400569916, + "learning_rate": 3.5934521342726898e-06, + "loss": 0.5405, + "step": 3395 + }, + { + "epoch": 0.28147534189805223, + "grad_norm": 0.25057893991470337, + "learning_rate": 3.5930377123912146e-06, + "loss": 0.4672, + "step": 3396 + }, + { + "epoch": 0.2815582262743473, + "grad_norm": 0.25016650557518005, + "learning_rate": 3.592623290509739e-06, + "loss": 0.5201, + "step": 3397 + }, + { + "epoch": 0.28164111065064235, + "grad_norm": 0.2798374593257904, + "learning_rate": 3.5922088686282643e-06, + "loss": 0.5542, + "step": 3398 + }, + { + "epoch": 0.2817239950269374, + "grad_norm": 0.2697664201259613, + "learning_rate": 3.5917944467467888e-06, + "loss": 0.5243, + "step": 3399 + }, + { + "epoch": 0.2818068794032325, + "grad_norm": 0.26883307099342346, + "learning_rate": 3.591380024865313e-06, + "loss": 0.5539, + "step": 3400 + }, + { + "epoch": 0.28188976377952757, + "grad_norm": 0.25594112277030945, + "learning_rate": 3.590965602983838e-06, + "loss": 0.5349, + "step": 3401 + }, + { + "epoch": 0.2819726481558226, + "grad_norm": 0.2628462314605713, + "learning_rate": 3.5905511811023625e-06, + "loss": 0.5566, + "step": 3402 + }, + { + "epoch": 0.2820555325321177, + "grad_norm": 0.2486766129732132, + "learning_rate": 3.590136759220887e-06, + "loss": 0.4567, + "step": 3403 + }, + { + "epoch": 0.28213841690841274, + "grad_norm": 0.2539660930633545, + "learning_rate": 3.5897223373394118e-06, + "loss": 0.5469, + "step": 3404 + }, + { + "epoch": 0.28222130128470785, + "grad_norm": 0.2925092577934265, + "learning_rate": 3.589307915457936e-06, + "loss": 0.5461, + "step": 3405 + }, + { + "epoch": 0.2823041856610029, + "grad_norm": 0.2632187008857727, + "learning_rate": 3.5888934935764615e-06, + "loss": 0.5546, + "step": 3406 + }, + { + "epoch": 0.28238707003729796, + "grad_norm": 0.25102391839027405, + "learning_rate": 3.588479071694986e-06, + "loss": 0.5303, + "step": 3407 + }, + { + "epoch": 0.282469954413593, + "grad_norm": 0.2659199833869934, + "learning_rate": 3.5880646498135103e-06, + "loss": 0.5231, + "step": 3408 + }, + { + "epoch": 0.2825528387898881, + "grad_norm": 0.25651228427886963, + "learning_rate": 3.587650227932035e-06, + "loss": 0.5299, + "step": 3409 + }, + { + "epoch": 0.2826357231661832, + "grad_norm": 0.2754357159137726, + "learning_rate": 3.5872358060505596e-06, + "loss": 0.5304, + "step": 3410 + }, + { + "epoch": 0.28271860754247824, + "grad_norm": 0.24869628250598907, + "learning_rate": 3.5868213841690845e-06, + "loss": 0.5285, + "step": 3411 + }, + { + "epoch": 0.2828014919187733, + "grad_norm": 0.27901631593704224, + "learning_rate": 3.586406962287609e-06, + "loss": 0.5986, + "step": 3412 + }, + { + "epoch": 0.2828843762950684, + "grad_norm": 0.25622957944869995, + "learning_rate": 3.5859925404061333e-06, + "loss": 0.5339, + "step": 3413 + }, + { + "epoch": 0.28296726067136346, + "grad_norm": 0.2570708096027374, + "learning_rate": 3.5855781185246586e-06, + "loss": 0.5804, + "step": 3414 + }, + { + "epoch": 0.2830501450476585, + "grad_norm": 0.25116467475891113, + "learning_rate": 3.585163696643183e-06, + "loss": 0.5299, + "step": 3415 + }, + { + "epoch": 0.2831330294239536, + "grad_norm": 0.2533397376537323, + "learning_rate": 3.584749274761708e-06, + "loss": 0.5134, + "step": 3416 + }, + { + "epoch": 0.28321591380024863, + "grad_norm": 0.25543299317359924, + "learning_rate": 3.5843348528802323e-06, + "loss": 0.5211, + "step": 3417 + }, + { + "epoch": 0.28329879817654374, + "grad_norm": 0.2647010385990143, + "learning_rate": 3.5839204309987568e-06, + "loss": 0.5563, + "step": 3418 + }, + { + "epoch": 0.2833816825528388, + "grad_norm": 0.2634742856025696, + "learning_rate": 3.583506009117282e-06, + "loss": 0.53, + "step": 3419 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 0.23999042809009552, + "learning_rate": 3.5830915872358065e-06, + "loss": 0.5113, + "step": 3420 + }, + { + "epoch": 0.2835474513054289, + "grad_norm": 0.2554287612438202, + "learning_rate": 3.582677165354331e-06, + "loss": 0.5184, + "step": 3421 + }, + { + "epoch": 0.283630335681724, + "grad_norm": 0.2431136965751648, + "learning_rate": 3.5822627434728558e-06, + "loss": 0.4851, + "step": 3422 + }, + { + "epoch": 0.2837132200580191, + "grad_norm": 0.2780151665210724, + "learning_rate": 3.58184832159138e-06, + "loss": 0.5596, + "step": 3423 + }, + { + "epoch": 0.28379610443431413, + "grad_norm": 0.2610645592212677, + "learning_rate": 3.581433899709905e-06, + "loss": 0.5222, + "step": 3424 + }, + { + "epoch": 0.2838789888106092, + "grad_norm": 0.2451491355895996, + "learning_rate": 3.5810194778284295e-06, + "loss": 0.5337, + "step": 3425 + }, + { + "epoch": 0.28396187318690425, + "grad_norm": 0.283757746219635, + "learning_rate": 3.580605055946954e-06, + "loss": 0.5605, + "step": 3426 + }, + { + "epoch": 0.28404475756319936, + "grad_norm": 0.26295149326324463, + "learning_rate": 3.580190634065479e-06, + "loss": 0.5571, + "step": 3427 + }, + { + "epoch": 0.2841276419394944, + "grad_norm": 0.28955045342445374, + "learning_rate": 3.5797762121840036e-06, + "loss": 0.5188, + "step": 3428 + }, + { + "epoch": 0.28421052631578947, + "grad_norm": 0.2555926442146301, + "learning_rate": 3.5793617903025285e-06, + "loss": 0.5078, + "step": 3429 + }, + { + "epoch": 0.2842934106920845, + "grad_norm": 0.25781485438346863, + "learning_rate": 3.578947368421053e-06, + "loss": 0.581, + "step": 3430 + }, + { + "epoch": 0.28437629506837964, + "grad_norm": 0.28395652770996094, + "learning_rate": 3.5785329465395773e-06, + "loss": 0.5648, + "step": 3431 + }, + { + "epoch": 0.2844591794446747, + "grad_norm": 0.26100051403045654, + "learning_rate": 3.578118524658102e-06, + "loss": 0.5536, + "step": 3432 + }, + { + "epoch": 0.28454206382096975, + "grad_norm": 0.2682129740715027, + "learning_rate": 3.5777041027766266e-06, + "loss": 0.4991, + "step": 3433 + }, + { + "epoch": 0.2846249481972648, + "grad_norm": 0.24439740180969238, + "learning_rate": 3.577289680895152e-06, + "loss": 0.4773, + "step": 3434 + }, + { + "epoch": 0.28470783257355986, + "grad_norm": 0.28736257553100586, + "learning_rate": 3.5768752590136763e-06, + "loss": 0.554, + "step": 3435 + }, + { + "epoch": 0.28479071694985497, + "grad_norm": 0.2520984411239624, + "learning_rate": 3.5764608371322008e-06, + "loss": 0.4991, + "step": 3436 + }, + { + "epoch": 0.28487360132615, + "grad_norm": 0.2683401107788086, + "learning_rate": 3.5760464152507256e-06, + "loss": 0.5134, + "step": 3437 + }, + { + "epoch": 0.2849564857024451, + "grad_norm": 0.25807833671569824, + "learning_rate": 3.57563199336925e-06, + "loss": 0.5341, + "step": 3438 + }, + { + "epoch": 0.28503937007874014, + "grad_norm": 0.23446807265281677, + "learning_rate": 3.5752175714877745e-06, + "loss": 0.4929, + "step": 3439 + }, + { + "epoch": 0.28512225445503525, + "grad_norm": 0.26148325204849243, + "learning_rate": 3.5748031496062998e-06, + "loss": 0.544, + "step": 3440 + }, + { + "epoch": 0.2852051388313303, + "grad_norm": 0.249294713139534, + "learning_rate": 3.574388727724824e-06, + "loss": 0.4853, + "step": 3441 + }, + { + "epoch": 0.28528802320762536, + "grad_norm": 0.2695107161998749, + "learning_rate": 3.573974305843349e-06, + "loss": 0.4856, + "step": 3442 + }, + { + "epoch": 0.2853709075839204, + "grad_norm": 0.2565538287162781, + "learning_rate": 3.5735598839618735e-06, + "loss": 0.514, + "step": 3443 + }, + { + "epoch": 0.2854537919602155, + "grad_norm": 0.2646479904651642, + "learning_rate": 3.573145462080398e-06, + "loss": 0.5586, + "step": 3444 + }, + { + "epoch": 0.2855366763365106, + "grad_norm": 0.2555069029331207, + "learning_rate": 3.5727310401989228e-06, + "loss": 0.5352, + "step": 3445 + }, + { + "epoch": 0.28561956071280564, + "grad_norm": 0.25913041830062866, + "learning_rate": 3.5723166183174472e-06, + "loss": 0.5341, + "step": 3446 + }, + { + "epoch": 0.2857024450891007, + "grad_norm": 0.2660740315914154, + "learning_rate": 3.5719021964359725e-06, + "loss": 0.5315, + "step": 3447 + }, + { + "epoch": 0.28578532946539575, + "grad_norm": 0.28834402561187744, + "learning_rate": 3.571487774554497e-06, + "loss": 0.5779, + "step": 3448 + }, + { + "epoch": 0.28586821384169087, + "grad_norm": 0.2629520893096924, + "learning_rate": 3.5710733526730214e-06, + "loss": 0.5127, + "step": 3449 + }, + { + "epoch": 0.2859510982179859, + "grad_norm": 0.2964678406715393, + "learning_rate": 3.570658930791546e-06, + "loss": 0.5237, + "step": 3450 + }, + { + "epoch": 0.286033982594281, + "grad_norm": 0.2941250801086426, + "learning_rate": 3.5702445089100706e-06, + "loss": 0.5317, + "step": 3451 + }, + { + "epoch": 0.28611686697057603, + "grad_norm": 0.2616561949253082, + "learning_rate": 3.569830087028595e-06, + "loss": 0.5759, + "step": 3452 + }, + { + "epoch": 0.2861997513468711, + "grad_norm": 0.2745438814163208, + "learning_rate": 3.56941566514712e-06, + "loss": 0.5196, + "step": 3453 + }, + { + "epoch": 0.2862826357231662, + "grad_norm": 0.25127720832824707, + "learning_rate": 3.5690012432656444e-06, + "loss": 0.5177, + "step": 3454 + }, + { + "epoch": 0.28636552009946126, + "grad_norm": 0.2859114706516266, + "learning_rate": 3.5685868213841696e-06, + "loss": 0.5625, + "step": 3455 + }, + { + "epoch": 0.2864484044757563, + "grad_norm": 0.3042738139629364, + "learning_rate": 3.568172399502694e-06, + "loss": 0.6033, + "step": 3456 + }, + { + "epoch": 0.28653128885205137, + "grad_norm": 0.2689189314842224, + "learning_rate": 3.5677579776212185e-06, + "loss": 0.5129, + "step": 3457 + }, + { + "epoch": 0.2866141732283465, + "grad_norm": 0.2911272644996643, + "learning_rate": 3.5673435557397434e-06, + "loss": 0.5652, + "step": 3458 + }, + { + "epoch": 0.28669705760464154, + "grad_norm": 0.2642197906970978, + "learning_rate": 3.566929133858268e-06, + "loss": 0.5517, + "step": 3459 + }, + { + "epoch": 0.2867799419809366, + "grad_norm": 0.2639409601688385, + "learning_rate": 3.566514711976793e-06, + "loss": 0.5562, + "step": 3460 + }, + { + "epoch": 0.28686282635723165, + "grad_norm": 0.2718166410923004, + "learning_rate": 3.5661002900953175e-06, + "loss": 0.5578, + "step": 3461 + }, + { + "epoch": 0.2869457107335267, + "grad_norm": 0.31310558319091797, + "learning_rate": 3.565685868213842e-06, + "loss": 0.5568, + "step": 3462 + }, + { + "epoch": 0.2870285951098218, + "grad_norm": 0.26321300864219666, + "learning_rate": 3.5652714463323668e-06, + "loss": 0.5461, + "step": 3463 + }, + { + "epoch": 0.28711147948611687, + "grad_norm": 0.27305716276168823, + "learning_rate": 3.5648570244508912e-06, + "loss": 0.5319, + "step": 3464 + }, + { + "epoch": 0.2871943638624119, + "grad_norm": 0.2554228901863098, + "learning_rate": 3.564442602569416e-06, + "loss": 0.5352, + "step": 3465 + }, + { + "epoch": 0.287277248238707, + "grad_norm": 0.312426894903183, + "learning_rate": 3.5640281806879405e-06, + "loss": 0.5426, + "step": 3466 + }, + { + "epoch": 0.2873601326150021, + "grad_norm": 0.2604530155658722, + "learning_rate": 3.563613758806465e-06, + "loss": 0.5103, + "step": 3467 + }, + { + "epoch": 0.28744301699129715, + "grad_norm": 0.2767626643180847, + "learning_rate": 3.5631993369249902e-06, + "loss": 0.5193, + "step": 3468 + }, + { + "epoch": 0.2875259013675922, + "grad_norm": 0.252722829580307, + "learning_rate": 3.5627849150435146e-06, + "loss": 0.5143, + "step": 3469 + }, + { + "epoch": 0.28760878574388726, + "grad_norm": 0.28009268641471863, + "learning_rate": 3.562370493162039e-06, + "loss": 0.5211, + "step": 3470 + }, + { + "epoch": 0.2876916701201823, + "grad_norm": 0.24928592145442963, + "learning_rate": 3.561956071280564e-06, + "loss": 0.5204, + "step": 3471 + }, + { + "epoch": 0.28777455449647743, + "grad_norm": 0.2851804792881012, + "learning_rate": 3.5615416493990884e-06, + "loss": 0.5071, + "step": 3472 + }, + { + "epoch": 0.2878574388727725, + "grad_norm": 0.2682110071182251, + "learning_rate": 3.5611272275176132e-06, + "loss": 0.5251, + "step": 3473 + }, + { + "epoch": 0.28794032324906754, + "grad_norm": 0.2794106602668762, + "learning_rate": 3.5607128056361377e-06, + "loss": 0.5182, + "step": 3474 + }, + { + "epoch": 0.2880232076253626, + "grad_norm": 0.5931112170219421, + "learning_rate": 3.5602983837546625e-06, + "loss": 0.5101, + "step": 3475 + }, + { + "epoch": 0.2881060920016577, + "grad_norm": 0.27141281962394714, + "learning_rate": 3.5598839618731874e-06, + "loss": 0.494, + "step": 3476 + }, + { + "epoch": 0.28818897637795277, + "grad_norm": 0.2821704149246216, + "learning_rate": 3.559469539991712e-06, + "loss": 0.5505, + "step": 3477 + }, + { + "epoch": 0.2882718607542478, + "grad_norm": 0.26055386662483215, + "learning_rate": 3.5590551181102367e-06, + "loss": 0.5168, + "step": 3478 + }, + { + "epoch": 0.2883547451305429, + "grad_norm": 0.2530941665172577, + "learning_rate": 3.558640696228761e-06, + "loss": 0.5491, + "step": 3479 + }, + { + "epoch": 0.28843762950683793, + "grad_norm": 0.2574920058250427, + "learning_rate": 3.5582262743472855e-06, + "loss": 0.4951, + "step": 3480 + }, + { + "epoch": 0.28852051388313305, + "grad_norm": 0.29240044951438904, + "learning_rate": 3.557811852465811e-06, + "loss": 0.525, + "step": 3481 + }, + { + "epoch": 0.2886033982594281, + "grad_norm": 0.2772268056869507, + "learning_rate": 3.5573974305843352e-06, + "loss": 0.5598, + "step": 3482 + }, + { + "epoch": 0.28868628263572316, + "grad_norm": 0.27283719182014465, + "learning_rate": 3.55698300870286e-06, + "loss": 0.4935, + "step": 3483 + }, + { + "epoch": 0.2887691670120182, + "grad_norm": 0.25878429412841797, + "learning_rate": 3.5565685868213845e-06, + "loss": 0.5165, + "step": 3484 + }, + { + "epoch": 0.2888520513883133, + "grad_norm": 0.25588926672935486, + "learning_rate": 3.556154164939909e-06, + "loss": 0.5256, + "step": 3485 + }, + { + "epoch": 0.2889349357646084, + "grad_norm": 0.2672945261001587, + "learning_rate": 3.555739743058434e-06, + "loss": 0.5547, + "step": 3486 + }, + { + "epoch": 0.28901782014090344, + "grad_norm": 0.25291943550109863, + "learning_rate": 3.5553253211769582e-06, + "loss": 0.5128, + "step": 3487 + }, + { + "epoch": 0.2891007045171985, + "grad_norm": 0.2686325013637543, + "learning_rate": 3.5549108992954827e-06, + "loss": 0.5158, + "step": 3488 + }, + { + "epoch": 0.2891835888934936, + "grad_norm": 0.29024064540863037, + "learning_rate": 3.554496477414008e-06, + "loss": 0.5513, + "step": 3489 + }, + { + "epoch": 0.28926647326978866, + "grad_norm": 0.2630784809589386, + "learning_rate": 3.5540820555325324e-06, + "loss": 0.5438, + "step": 3490 + }, + { + "epoch": 0.2893493576460837, + "grad_norm": 0.27259525656700134, + "learning_rate": 3.5536676336510572e-06, + "loss": 0.5668, + "step": 3491 + }, + { + "epoch": 0.28943224202237877, + "grad_norm": 0.250681072473526, + "learning_rate": 3.5532532117695817e-06, + "loss": 0.4577, + "step": 3492 + }, + { + "epoch": 0.2895151263986738, + "grad_norm": 0.25700584053993225, + "learning_rate": 3.552838789888106e-06, + "loss": 0.503, + "step": 3493 + }, + { + "epoch": 0.28959801077496894, + "grad_norm": 0.27181464433670044, + "learning_rate": 3.552424368006631e-06, + "loss": 0.5548, + "step": 3494 + }, + { + "epoch": 0.289680895151264, + "grad_norm": 0.25433585047721863, + "learning_rate": 3.552009946125156e-06, + "loss": 0.5564, + "step": 3495 + }, + { + "epoch": 0.28976377952755905, + "grad_norm": 0.2601119875907898, + "learning_rate": 3.5515955242436807e-06, + "loss": 0.5244, + "step": 3496 + }, + { + "epoch": 0.2898466639038541, + "grad_norm": 0.2564889192581177, + "learning_rate": 3.551181102362205e-06, + "loss": 0.521, + "step": 3497 + }, + { + "epoch": 0.2899295482801492, + "grad_norm": 0.25750288367271423, + "learning_rate": 3.5507666804807295e-06, + "loss": 0.5004, + "step": 3498 + }, + { + "epoch": 0.2900124326564443, + "grad_norm": 0.26950803399086, + "learning_rate": 3.5503522585992544e-06, + "loss": 0.5879, + "step": 3499 + }, + { + "epoch": 0.29009531703273933, + "grad_norm": 0.2482549548149109, + "learning_rate": 3.549937836717779e-06, + "loss": 0.4882, + "step": 3500 + }, + { + "epoch": 0.2901782014090344, + "grad_norm": 0.2553381025791168, + "learning_rate": 3.549523414836304e-06, + "loss": 0.496, + "step": 3501 + }, + { + "epoch": 0.29026108578532944, + "grad_norm": 0.2884554862976074, + "learning_rate": 3.5491089929548285e-06, + "loss": 0.5436, + "step": 3502 + }, + { + "epoch": 0.29034397016162455, + "grad_norm": 0.2677091658115387, + "learning_rate": 3.548694571073353e-06, + "loss": 0.5656, + "step": 3503 + }, + { + "epoch": 0.2904268545379196, + "grad_norm": 0.2561146318912506, + "learning_rate": 3.548280149191878e-06, + "loss": 0.5065, + "step": 3504 + }, + { + "epoch": 0.29050973891421467, + "grad_norm": 0.25775739550590515, + "learning_rate": 3.5478657273104022e-06, + "loss": 0.5418, + "step": 3505 + }, + { + "epoch": 0.2905926232905097, + "grad_norm": 0.26649266481399536, + "learning_rate": 3.5474513054289267e-06, + "loss": 0.5743, + "step": 3506 + }, + { + "epoch": 0.29067550766680483, + "grad_norm": 0.2766686975955963, + "learning_rate": 3.5470368835474515e-06, + "loss": 0.5366, + "step": 3507 + }, + { + "epoch": 0.2907583920430999, + "grad_norm": 0.2647007405757904, + "learning_rate": 3.546622461665976e-06, + "loss": 0.5395, + "step": 3508 + }, + { + "epoch": 0.29084127641939495, + "grad_norm": 0.2594122588634491, + "learning_rate": 3.5462080397845012e-06, + "loss": 0.5418, + "step": 3509 + }, + { + "epoch": 0.29092416079569, + "grad_norm": 0.2769795358181, + "learning_rate": 3.5457936179030257e-06, + "loss": 0.5135, + "step": 3510 + }, + { + "epoch": 0.29100704517198506, + "grad_norm": 0.266630619764328, + "learning_rate": 3.54537919602155e-06, + "loss": 0.5027, + "step": 3511 + }, + { + "epoch": 0.29108992954828017, + "grad_norm": 0.28297463059425354, + "learning_rate": 3.544964774140075e-06, + "loss": 0.5585, + "step": 3512 + }, + { + "epoch": 0.2911728139245752, + "grad_norm": 0.2639424502849579, + "learning_rate": 3.5445503522585994e-06, + "loss": 0.5383, + "step": 3513 + }, + { + "epoch": 0.2912556983008703, + "grad_norm": 0.24074816703796387, + "learning_rate": 3.5441359303771242e-06, + "loss": 0.4663, + "step": 3514 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 0.2603960931301117, + "learning_rate": 3.5437215084956487e-06, + "loss": 0.5436, + "step": 3515 + }, + { + "epoch": 0.29142146705346045, + "grad_norm": 0.2467331439256668, + "learning_rate": 3.5433070866141735e-06, + "loss": 0.5131, + "step": 3516 + }, + { + "epoch": 0.2915043514297555, + "grad_norm": 0.2339547574520111, + "learning_rate": 3.5428926647326984e-06, + "loss": 0.4539, + "step": 3517 + }, + { + "epoch": 0.29158723580605056, + "grad_norm": 0.2595500946044922, + "learning_rate": 3.542478242851223e-06, + "loss": 0.5429, + "step": 3518 + }, + { + "epoch": 0.2916701201823456, + "grad_norm": 0.25910189747810364, + "learning_rate": 3.5420638209697477e-06, + "loss": 0.5502, + "step": 3519 + }, + { + "epoch": 0.29175300455864067, + "grad_norm": 0.26983827352523804, + "learning_rate": 3.541649399088272e-06, + "loss": 0.5383, + "step": 3520 + }, + { + "epoch": 0.2918358889349358, + "grad_norm": 0.2724289298057556, + "learning_rate": 3.5412349772067965e-06, + "loss": 0.5348, + "step": 3521 + }, + { + "epoch": 0.29191877331123084, + "grad_norm": 0.24919529259204865, + "learning_rate": 3.540820555325322e-06, + "loss": 0.5143, + "step": 3522 + }, + { + "epoch": 0.2920016576875259, + "grad_norm": 0.23712977766990662, + "learning_rate": 3.5404061334438462e-06, + "loss": 0.5212, + "step": 3523 + }, + { + "epoch": 0.29208454206382095, + "grad_norm": 0.2514236271381378, + "learning_rate": 3.5399917115623707e-06, + "loss": 0.5464, + "step": 3524 + }, + { + "epoch": 0.29216742644011606, + "grad_norm": 0.26576754450798035, + "learning_rate": 3.5395772896808955e-06, + "loss": 0.5406, + "step": 3525 + }, + { + "epoch": 0.2922503108164111, + "grad_norm": 0.27268028259277344, + "learning_rate": 3.53916286779942e-06, + "loss": 0.5616, + "step": 3526 + }, + { + "epoch": 0.2923331951927062, + "grad_norm": 0.2606621980667114, + "learning_rate": 3.538748445917945e-06, + "loss": 0.5031, + "step": 3527 + }, + { + "epoch": 0.29241607956900123, + "grad_norm": 0.257281631231308, + "learning_rate": 3.5383340240364693e-06, + "loss": 0.5204, + "step": 3528 + }, + { + "epoch": 0.2924989639452963, + "grad_norm": 0.2951712906360626, + "learning_rate": 3.5379196021549937e-06, + "loss": 0.5796, + "step": 3529 + }, + { + "epoch": 0.2925818483215914, + "grad_norm": 0.27941036224365234, + "learning_rate": 3.537505180273519e-06, + "loss": 0.5177, + "step": 3530 + }, + { + "epoch": 0.29266473269788645, + "grad_norm": 0.27466773986816406, + "learning_rate": 3.5370907583920434e-06, + "loss": 0.5245, + "step": 3531 + }, + { + "epoch": 0.2927476170741815, + "grad_norm": 0.253654807806015, + "learning_rate": 3.5366763365105683e-06, + "loss": 0.5472, + "step": 3532 + }, + { + "epoch": 0.29283050145047657, + "grad_norm": 0.2628018856048584, + "learning_rate": 3.5362619146290927e-06, + "loss": 0.5557, + "step": 3533 + }, + { + "epoch": 0.2929133858267717, + "grad_norm": 0.25340792536735535, + "learning_rate": 3.535847492747617e-06, + "loss": 0.5012, + "step": 3534 + }, + { + "epoch": 0.29299627020306673, + "grad_norm": 0.24291154742240906, + "learning_rate": 3.535433070866142e-06, + "loss": 0.5414, + "step": 3535 + }, + { + "epoch": 0.2930791545793618, + "grad_norm": 0.26792070269584656, + "learning_rate": 3.535018648984667e-06, + "loss": 0.5695, + "step": 3536 + }, + { + "epoch": 0.29316203895565685, + "grad_norm": 0.245859295129776, + "learning_rate": 3.5346042271031913e-06, + "loss": 0.5287, + "step": 3537 + }, + { + "epoch": 0.2932449233319519, + "grad_norm": 0.25867289304733276, + "learning_rate": 3.534189805221716e-06, + "loss": 0.5128, + "step": 3538 + }, + { + "epoch": 0.293327807708247, + "grad_norm": 0.28369954228401184, + "learning_rate": 3.5337753833402405e-06, + "loss": 0.5435, + "step": 3539 + }, + { + "epoch": 0.29341069208454207, + "grad_norm": 0.24838389456272125, + "learning_rate": 3.5333609614587654e-06, + "loss": 0.5304, + "step": 3540 + }, + { + "epoch": 0.2934935764608371, + "grad_norm": 0.26694756746292114, + "learning_rate": 3.53294653957729e-06, + "loss": 0.5155, + "step": 3541 + }, + { + "epoch": 0.2935764608371322, + "grad_norm": 0.2642602026462555, + "learning_rate": 3.5325321176958143e-06, + "loss": 0.5113, + "step": 3542 + }, + { + "epoch": 0.2936593452134273, + "grad_norm": 0.2570878267288208, + "learning_rate": 3.5321176958143395e-06, + "loss": 0.5377, + "step": 3543 + }, + { + "epoch": 0.29374222958972235, + "grad_norm": 0.26003727316856384, + "learning_rate": 3.531703273932864e-06, + "loss": 0.5345, + "step": 3544 + }, + { + "epoch": 0.2938251139660174, + "grad_norm": 0.28346142172813416, + "learning_rate": 3.531288852051389e-06, + "loss": 0.53, + "step": 3545 + }, + { + "epoch": 0.29390799834231246, + "grad_norm": 0.2557445764541626, + "learning_rate": 3.5308744301699133e-06, + "loss": 0.5223, + "step": 3546 + }, + { + "epoch": 0.2939908827186075, + "grad_norm": 0.24387088418006897, + "learning_rate": 3.5304600082884377e-06, + "loss": 0.5018, + "step": 3547 + }, + { + "epoch": 0.2940737670949026, + "grad_norm": 0.27852463722229004, + "learning_rate": 3.5300455864069625e-06, + "loss": 0.5666, + "step": 3548 + }, + { + "epoch": 0.2941566514711977, + "grad_norm": 0.2785398066043854, + "learning_rate": 3.529631164525487e-06, + "loss": 0.5399, + "step": 3549 + }, + { + "epoch": 0.29423953584749274, + "grad_norm": 0.27801910042762756, + "learning_rate": 3.5292167426440123e-06, + "loss": 0.5445, + "step": 3550 + }, + { + "epoch": 0.2943224202237878, + "grad_norm": 0.25138363242149353, + "learning_rate": 3.5288023207625367e-06, + "loss": 0.5007, + "step": 3551 + }, + { + "epoch": 0.2944053046000829, + "grad_norm": 0.23798352479934692, + "learning_rate": 3.528387898881061e-06, + "loss": 0.5187, + "step": 3552 + }, + { + "epoch": 0.29448818897637796, + "grad_norm": 0.25968125462532043, + "learning_rate": 3.527973476999586e-06, + "loss": 0.5349, + "step": 3553 + }, + { + "epoch": 0.294571073352673, + "grad_norm": 0.25081634521484375, + "learning_rate": 3.5275590551181104e-06, + "loss": 0.5314, + "step": 3554 + }, + { + "epoch": 0.2946539577289681, + "grad_norm": 0.24935947358608246, + "learning_rate": 3.527144633236635e-06, + "loss": 0.5176, + "step": 3555 + }, + { + "epoch": 0.29473684210526313, + "grad_norm": 0.2539975345134735, + "learning_rate": 3.52673021135516e-06, + "loss": 0.5008, + "step": 3556 + }, + { + "epoch": 0.29481972648155824, + "grad_norm": 0.23919211328029633, + "learning_rate": 3.5263157894736846e-06, + "loss": 0.4952, + "step": 3557 + }, + { + "epoch": 0.2949026108578533, + "grad_norm": 0.25720128417015076, + "learning_rate": 3.5259013675922094e-06, + "loss": 0.4799, + "step": 3558 + }, + { + "epoch": 0.29498549523414835, + "grad_norm": 0.2700645625591278, + "learning_rate": 3.525486945710734e-06, + "loss": 0.5522, + "step": 3559 + }, + { + "epoch": 0.2950683796104434, + "grad_norm": 0.254911869764328, + "learning_rate": 3.5250725238292583e-06, + "loss": 0.5783, + "step": 3560 + }, + { + "epoch": 0.2951512639867385, + "grad_norm": 0.25263211131095886, + "learning_rate": 3.524658101947783e-06, + "loss": 0.5444, + "step": 3561 + }, + { + "epoch": 0.2952341483630336, + "grad_norm": 0.2674219608306885, + "learning_rate": 3.5242436800663076e-06, + "loss": 0.5821, + "step": 3562 + }, + { + "epoch": 0.29531703273932863, + "grad_norm": 0.27177971601486206, + "learning_rate": 3.523829258184833e-06, + "loss": 0.5079, + "step": 3563 + }, + { + "epoch": 0.2953999171156237, + "grad_norm": 0.2629162669181824, + "learning_rate": 3.5234148363033573e-06, + "loss": 0.5039, + "step": 3564 + }, + { + "epoch": 0.2954828014919188, + "grad_norm": 0.2850857079029083, + "learning_rate": 3.5230004144218817e-06, + "loss": 0.5923, + "step": 3565 + }, + { + "epoch": 0.29556568586821386, + "grad_norm": 0.27113690972328186, + "learning_rate": 3.5225859925404066e-06, + "loss": 0.5612, + "step": 3566 + }, + { + "epoch": 0.2956485702445089, + "grad_norm": 0.2849039137363434, + "learning_rate": 3.522171570658931e-06, + "loss": 0.5629, + "step": 3567 + }, + { + "epoch": 0.29573145462080397, + "grad_norm": 0.265720009803772, + "learning_rate": 3.521757148777456e-06, + "loss": 0.5349, + "step": 3568 + }, + { + "epoch": 0.295814338997099, + "grad_norm": 0.2583088278770447, + "learning_rate": 3.5213427268959803e-06, + "loss": 0.5331, + "step": 3569 + }, + { + "epoch": 0.29589722337339414, + "grad_norm": 0.26170575618743896, + "learning_rate": 3.5209283050145047e-06, + "loss": 0.537, + "step": 3570 + }, + { + "epoch": 0.2959801077496892, + "grad_norm": 0.265774667263031, + "learning_rate": 3.52051388313303e-06, + "loss": 0.5278, + "step": 3571 + }, + { + "epoch": 0.29606299212598425, + "grad_norm": 0.2708953320980072, + "learning_rate": 3.5200994612515544e-06, + "loss": 0.5268, + "step": 3572 + }, + { + "epoch": 0.2961458765022793, + "grad_norm": 0.25735148787498474, + "learning_rate": 3.519685039370079e-06, + "loss": 0.5334, + "step": 3573 + }, + { + "epoch": 0.2962287608785744, + "grad_norm": 0.26449501514434814, + "learning_rate": 3.5192706174886037e-06, + "loss": 0.5256, + "step": 3574 + }, + { + "epoch": 0.29631164525486947, + "grad_norm": 0.2764492928981781, + "learning_rate": 3.518856195607128e-06, + "loss": 0.5579, + "step": 3575 + }, + { + "epoch": 0.2963945296311645, + "grad_norm": 0.2701444625854492, + "learning_rate": 3.5184417737256534e-06, + "loss": 0.5609, + "step": 3576 + }, + { + "epoch": 0.2964774140074596, + "grad_norm": 0.2542991638183594, + "learning_rate": 3.518027351844178e-06, + "loss": 0.5006, + "step": 3577 + }, + { + "epoch": 0.29656029838375464, + "grad_norm": 0.2620658278465271, + "learning_rate": 3.5176129299627023e-06, + "loss": 0.5099, + "step": 3578 + }, + { + "epoch": 0.29664318276004975, + "grad_norm": 0.2538573741912842, + "learning_rate": 3.517198508081227e-06, + "loss": 0.5011, + "step": 3579 + }, + { + "epoch": 0.2967260671363448, + "grad_norm": 0.2561744153499603, + "learning_rate": 3.5167840861997516e-06, + "loss": 0.5047, + "step": 3580 + }, + { + "epoch": 0.29680895151263986, + "grad_norm": 0.24899566173553467, + "learning_rate": 3.5163696643182764e-06, + "loss": 0.5258, + "step": 3581 + }, + { + "epoch": 0.2968918358889349, + "grad_norm": 0.26993992924690247, + "learning_rate": 3.515955242436801e-06, + "loss": 0.5245, + "step": 3582 + }, + { + "epoch": 0.29697472026523003, + "grad_norm": 0.29178690910339355, + "learning_rate": 3.5155408205553253e-06, + "loss": 0.5795, + "step": 3583 + }, + { + "epoch": 0.2970576046415251, + "grad_norm": 0.27187466621398926, + "learning_rate": 3.5151263986738506e-06, + "loss": 0.5527, + "step": 3584 + }, + { + "epoch": 0.29714048901782014, + "grad_norm": 0.2347342073917389, + "learning_rate": 3.514711976792375e-06, + "loss": 0.4944, + "step": 3585 + }, + { + "epoch": 0.2972233733941152, + "grad_norm": 0.2739121615886688, + "learning_rate": 3.5142975549109e-06, + "loss": 0.5525, + "step": 3586 + }, + { + "epoch": 0.29730625777041025, + "grad_norm": 0.2492673546075821, + "learning_rate": 3.5138831330294243e-06, + "loss": 0.5161, + "step": 3587 + }, + { + "epoch": 0.29738914214670537, + "grad_norm": 0.24306507408618927, + "learning_rate": 3.5134687111479487e-06, + "loss": 0.5198, + "step": 3588 + }, + { + "epoch": 0.2974720265230004, + "grad_norm": 0.28127774596214294, + "learning_rate": 3.5130542892664736e-06, + "loss": 0.5776, + "step": 3589 + }, + { + "epoch": 0.2975549108992955, + "grad_norm": 0.272644966840744, + "learning_rate": 3.512639867384998e-06, + "loss": 0.5123, + "step": 3590 + }, + { + "epoch": 0.29763779527559053, + "grad_norm": 0.2625287175178528, + "learning_rate": 3.5122254455035224e-06, + "loss": 0.5103, + "step": 3591 + }, + { + "epoch": 0.29772067965188564, + "grad_norm": 0.24663716554641724, + "learning_rate": 3.5118110236220477e-06, + "loss": 0.497, + "step": 3592 + }, + { + "epoch": 0.2978035640281807, + "grad_norm": 0.2678804099559784, + "learning_rate": 3.511396601740572e-06, + "loss": 0.5252, + "step": 3593 + }, + { + "epoch": 0.29788644840447576, + "grad_norm": 0.23896239697933197, + "learning_rate": 3.510982179859097e-06, + "loss": 0.491, + "step": 3594 + }, + { + "epoch": 0.2979693327807708, + "grad_norm": 0.2709214985370636, + "learning_rate": 3.5105677579776214e-06, + "loss": 0.4863, + "step": 3595 + }, + { + "epoch": 0.29805221715706587, + "grad_norm": 0.2601209878921509, + "learning_rate": 3.510153336096146e-06, + "loss": 0.553, + "step": 3596 + }, + { + "epoch": 0.298135101533361, + "grad_norm": 0.2649509906768799, + "learning_rate": 3.509738914214671e-06, + "loss": 0.5507, + "step": 3597 + }, + { + "epoch": 0.29821798590965604, + "grad_norm": 0.25925853848457336, + "learning_rate": 3.5093244923331956e-06, + "loss": 0.5568, + "step": 3598 + }, + { + "epoch": 0.2983008702859511, + "grad_norm": 0.2640743851661682, + "learning_rate": 3.5089100704517204e-06, + "loss": 0.5287, + "step": 3599 + }, + { + "epoch": 0.29838375466224615, + "grad_norm": 0.27651694416999817, + "learning_rate": 3.508495648570245e-06, + "loss": 0.5282, + "step": 3600 + }, + { + "epoch": 0.29846663903854126, + "grad_norm": 0.3763650357723236, + "learning_rate": 3.5080812266887693e-06, + "loss": 0.5368, + "step": 3601 + }, + { + "epoch": 0.2985495234148363, + "grad_norm": 0.25675880908966064, + "learning_rate": 3.507666804807294e-06, + "loss": 0.5181, + "step": 3602 + }, + { + "epoch": 0.29863240779113137, + "grad_norm": 0.25794726610183716, + "learning_rate": 3.5072523829258186e-06, + "loss": 0.5155, + "step": 3603 + }, + { + "epoch": 0.2987152921674264, + "grad_norm": 0.26466622948646545, + "learning_rate": 3.506837961044343e-06, + "loss": 0.526, + "step": 3604 + }, + { + "epoch": 0.2987981765437215, + "grad_norm": 0.26618504524230957, + "learning_rate": 3.5064235391628683e-06, + "loss": 0.553, + "step": 3605 + }, + { + "epoch": 0.2988810609200166, + "grad_norm": 0.26045259833335876, + "learning_rate": 3.5060091172813927e-06, + "loss": 0.54, + "step": 3606 + }, + { + "epoch": 0.29896394529631165, + "grad_norm": 0.2657877802848816, + "learning_rate": 3.5055946953999176e-06, + "loss": 0.5439, + "step": 3607 + }, + { + "epoch": 0.2990468296726067, + "grad_norm": 0.2416057586669922, + "learning_rate": 3.505180273518442e-06, + "loss": 0.5067, + "step": 3608 + }, + { + "epoch": 0.29912971404890176, + "grad_norm": 0.2953014075756073, + "learning_rate": 3.5047658516369664e-06, + "loss": 0.5537, + "step": 3609 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 0.2581882178783417, + "learning_rate": 3.5043514297554913e-06, + "loss": 0.5286, + "step": 3610 + }, + { + "epoch": 0.29929548280149193, + "grad_norm": 0.3019474744796753, + "learning_rate": 3.5039370078740157e-06, + "loss": 0.5782, + "step": 3611 + }, + { + "epoch": 0.299378367177787, + "grad_norm": 0.23505936563014984, + "learning_rate": 3.503522585992541e-06, + "loss": 0.4933, + "step": 3612 + }, + { + "epoch": 0.29946125155408204, + "grad_norm": 0.26247841119766235, + "learning_rate": 3.5031081641110654e-06, + "loss": 0.522, + "step": 3613 + }, + { + "epoch": 0.2995441359303771, + "grad_norm": 0.2473120242357254, + "learning_rate": 3.50269374222959e-06, + "loss": 0.5263, + "step": 3614 + }, + { + "epoch": 0.2996270203066722, + "grad_norm": 0.2713656723499298, + "learning_rate": 3.5022793203481147e-06, + "loss": 0.5335, + "step": 3615 + }, + { + "epoch": 0.29970990468296727, + "grad_norm": 0.25797590613365173, + "learning_rate": 3.501864898466639e-06, + "loss": 0.5024, + "step": 3616 + }, + { + "epoch": 0.2997927890592623, + "grad_norm": 0.268822580575943, + "learning_rate": 3.5014504765851644e-06, + "loss": 0.5109, + "step": 3617 + }, + { + "epoch": 0.2998756734355574, + "grad_norm": 0.2777380049228668, + "learning_rate": 3.501036054703689e-06, + "loss": 0.5547, + "step": 3618 + }, + { + "epoch": 0.2999585578118525, + "grad_norm": 0.2661244869232178, + "learning_rate": 3.5006216328222133e-06, + "loss": 0.5576, + "step": 3619 + }, + { + "epoch": 0.30004144218814754, + "grad_norm": 0.2980940639972687, + "learning_rate": 3.500207210940738e-06, + "loss": 0.5503, + "step": 3620 + }, + { + "epoch": 0.3001243265644426, + "grad_norm": 0.2588352560997009, + "learning_rate": 3.4997927890592626e-06, + "loss": 0.5205, + "step": 3621 + }, + { + "epoch": 0.30020721094073766, + "grad_norm": 0.25429990887641907, + "learning_rate": 3.499378367177787e-06, + "loss": 0.4763, + "step": 3622 + }, + { + "epoch": 0.3002900953170327, + "grad_norm": 0.2531398832798004, + "learning_rate": 3.498963945296312e-06, + "loss": 0.5451, + "step": 3623 + }, + { + "epoch": 0.3003729796933278, + "grad_norm": 0.26273277401924133, + "learning_rate": 3.4985495234148363e-06, + "loss": 0.5606, + "step": 3624 + }, + { + "epoch": 0.3004558640696229, + "grad_norm": 0.2788519859313965, + "learning_rate": 3.4981351015333616e-06, + "loss": 0.5748, + "step": 3625 + }, + { + "epoch": 0.30053874844591794, + "grad_norm": 0.24456974864006042, + "learning_rate": 3.497720679651886e-06, + "loss": 0.4987, + "step": 3626 + }, + { + "epoch": 0.300621632822213, + "grad_norm": 0.26474425196647644, + "learning_rate": 3.4973062577704104e-06, + "loss": 0.5279, + "step": 3627 + }, + { + "epoch": 0.3007045171985081, + "grad_norm": 0.2635611593723297, + "learning_rate": 3.4968918358889353e-06, + "loss": 0.5282, + "step": 3628 + }, + { + "epoch": 0.30078740157480316, + "grad_norm": 0.2533359229564667, + "learning_rate": 3.4964774140074597e-06, + "loss": 0.5146, + "step": 3629 + }, + { + "epoch": 0.3008702859510982, + "grad_norm": 0.2660093307495117, + "learning_rate": 3.4960629921259846e-06, + "loss": 0.539, + "step": 3630 + }, + { + "epoch": 0.30095317032739327, + "grad_norm": 0.267210453748703, + "learning_rate": 3.495648570244509e-06, + "loss": 0.5239, + "step": 3631 + }, + { + "epoch": 0.3010360547036883, + "grad_norm": 0.26287907361984253, + "learning_rate": 3.4952341483630335e-06, + "loss": 0.5219, + "step": 3632 + }, + { + "epoch": 0.30111893907998344, + "grad_norm": 0.2828295826911926, + "learning_rate": 3.4948197264815587e-06, + "loss": 0.5838, + "step": 3633 + }, + { + "epoch": 0.3012018234562785, + "grad_norm": 0.24726659059524536, + "learning_rate": 3.494405304600083e-06, + "loss": 0.4854, + "step": 3634 + }, + { + "epoch": 0.30128470783257355, + "grad_norm": 0.27660083770751953, + "learning_rate": 3.493990882718608e-06, + "loss": 0.5528, + "step": 3635 + }, + { + "epoch": 0.3013675922088686, + "grad_norm": 0.26825836300849915, + "learning_rate": 3.4935764608371325e-06, + "loss": 0.5264, + "step": 3636 + }, + { + "epoch": 0.3014504765851637, + "grad_norm": 0.25454211235046387, + "learning_rate": 3.493162038955657e-06, + "loss": 0.5473, + "step": 3637 + }, + { + "epoch": 0.3015333609614588, + "grad_norm": 0.2704048752784729, + "learning_rate": 3.492747617074182e-06, + "loss": 0.5613, + "step": 3638 + }, + { + "epoch": 0.30161624533775383, + "grad_norm": 0.26984933018684387, + "learning_rate": 3.4923331951927066e-06, + "loss": 0.5771, + "step": 3639 + }, + { + "epoch": 0.3016991297140489, + "grad_norm": 0.2731933891773224, + "learning_rate": 3.491918773311231e-06, + "loss": 0.5389, + "step": 3640 + }, + { + "epoch": 0.301782014090344, + "grad_norm": 0.2790355086326599, + "learning_rate": 3.491504351429756e-06, + "loss": 0.5589, + "step": 3641 + }, + { + "epoch": 0.30186489846663905, + "grad_norm": 0.2749725580215454, + "learning_rate": 3.4910899295482803e-06, + "loss": 0.5431, + "step": 3642 + }, + { + "epoch": 0.3019477828429341, + "grad_norm": 0.26873400807380676, + "learning_rate": 3.490675507666805e-06, + "loss": 0.4928, + "step": 3643 + }, + { + "epoch": 0.30203066721922917, + "grad_norm": 0.26647162437438965, + "learning_rate": 3.4902610857853296e-06, + "loss": 0.501, + "step": 3644 + }, + { + "epoch": 0.3021135515955242, + "grad_norm": 0.26816776394844055, + "learning_rate": 3.489846663903854e-06, + "loss": 0.5097, + "step": 3645 + }, + { + "epoch": 0.30219643597181933, + "grad_norm": 0.2644834816455841, + "learning_rate": 3.4894322420223793e-06, + "loss": 0.5381, + "step": 3646 + }, + { + "epoch": 0.3022793203481144, + "grad_norm": 0.25453630089759827, + "learning_rate": 3.4890178201409037e-06, + "loss": 0.5052, + "step": 3647 + }, + { + "epoch": 0.30236220472440944, + "grad_norm": 0.24534066021442413, + "learning_rate": 3.4886033982594286e-06, + "loss": 0.4812, + "step": 3648 + }, + { + "epoch": 0.3024450891007045, + "grad_norm": 0.2685686945915222, + "learning_rate": 3.488188976377953e-06, + "loss": 0.5397, + "step": 3649 + }, + { + "epoch": 0.3025279734769996, + "grad_norm": 0.26028314232826233, + "learning_rate": 3.4877745544964775e-06, + "loss": 0.5142, + "step": 3650 + }, + { + "epoch": 0.30261085785329467, + "grad_norm": 0.2643410563468933, + "learning_rate": 3.4873601326150023e-06, + "loss": 0.5515, + "step": 3651 + }, + { + "epoch": 0.3026937422295897, + "grad_norm": 0.2443723976612091, + "learning_rate": 3.4869457107335267e-06, + "loss": 0.5273, + "step": 3652 + }, + { + "epoch": 0.3027766266058848, + "grad_norm": 0.2519228756427765, + "learning_rate": 3.486531288852052e-06, + "loss": 0.5588, + "step": 3653 + }, + { + "epoch": 0.30285951098217984, + "grad_norm": 0.26830461621284485, + "learning_rate": 3.4861168669705765e-06, + "loss": 0.5526, + "step": 3654 + }, + { + "epoch": 0.30294239535847495, + "grad_norm": 0.25509536266326904, + "learning_rate": 3.485702445089101e-06, + "loss": 0.5216, + "step": 3655 + }, + { + "epoch": 0.30302527973477, + "grad_norm": 0.2298627644777298, + "learning_rate": 3.4852880232076257e-06, + "loss": 0.458, + "step": 3656 + }, + { + "epoch": 0.30310816411106506, + "grad_norm": 0.2559512257575989, + "learning_rate": 3.48487360132615e-06, + "loss": 0.56, + "step": 3657 + }, + { + "epoch": 0.3031910484873601, + "grad_norm": 0.27143028378486633, + "learning_rate": 3.4844591794446746e-06, + "loss": 0.5348, + "step": 3658 + }, + { + "epoch": 0.3032739328636552, + "grad_norm": 0.28687939047813416, + "learning_rate": 3.4840447575632e-06, + "loss": 0.5718, + "step": 3659 + }, + { + "epoch": 0.3033568172399503, + "grad_norm": 0.25256186723709106, + "learning_rate": 3.4836303356817243e-06, + "loss": 0.5024, + "step": 3660 + }, + { + "epoch": 0.30343970161624534, + "grad_norm": 0.26691505312919617, + "learning_rate": 3.483215913800249e-06, + "loss": 0.5632, + "step": 3661 + }, + { + "epoch": 0.3035225859925404, + "grad_norm": 0.2697103023529053, + "learning_rate": 3.4828014919187736e-06, + "loss": 0.5739, + "step": 3662 + }, + { + "epoch": 0.30360547036883545, + "grad_norm": 0.28462013602256775, + "learning_rate": 3.482387070037298e-06, + "loss": 0.5895, + "step": 3663 + }, + { + "epoch": 0.30368835474513056, + "grad_norm": 0.25508642196655273, + "learning_rate": 3.481972648155823e-06, + "loss": 0.5587, + "step": 3664 + }, + { + "epoch": 0.3037712391214256, + "grad_norm": 0.2631795108318329, + "learning_rate": 3.4815582262743473e-06, + "loss": 0.5593, + "step": 3665 + }, + { + "epoch": 0.3038541234977207, + "grad_norm": 0.2986525297164917, + "learning_rate": 3.4811438043928726e-06, + "loss": 0.495, + "step": 3666 + }, + { + "epoch": 0.30393700787401573, + "grad_norm": 0.2814001739025116, + "learning_rate": 3.480729382511397e-06, + "loss": 0.5408, + "step": 3667 + }, + { + "epoch": 0.30401989225031084, + "grad_norm": 0.27535882592201233, + "learning_rate": 3.4803149606299215e-06, + "loss": 0.5339, + "step": 3668 + }, + { + "epoch": 0.3041027766266059, + "grad_norm": 0.23902229964733124, + "learning_rate": 3.4799005387484463e-06, + "loss": 0.4858, + "step": 3669 + }, + { + "epoch": 0.30418566100290095, + "grad_norm": 0.26968663930892944, + "learning_rate": 3.4794861168669708e-06, + "loss": 0.5297, + "step": 3670 + }, + { + "epoch": 0.304268545379196, + "grad_norm": 0.2596248686313629, + "learning_rate": 3.4790716949854956e-06, + "loss": 0.5293, + "step": 3671 + }, + { + "epoch": 0.30435142975549107, + "grad_norm": 0.2612520456314087, + "learning_rate": 3.47865727310402e-06, + "loss": 0.5194, + "step": 3672 + }, + { + "epoch": 0.3044343141317862, + "grad_norm": 0.2683924734592438, + "learning_rate": 3.4782428512225445e-06, + "loss": 0.5176, + "step": 3673 + }, + { + "epoch": 0.30451719850808123, + "grad_norm": 0.29606130719184875, + "learning_rate": 3.4778284293410698e-06, + "loss": 0.5486, + "step": 3674 + }, + { + "epoch": 0.3046000828843763, + "grad_norm": 0.25238773226737976, + "learning_rate": 3.477414007459594e-06, + "loss": 0.517, + "step": 3675 + }, + { + "epoch": 0.30468296726067134, + "grad_norm": 0.2724360525608063, + "learning_rate": 3.4769995855781186e-06, + "loss": 0.5839, + "step": 3676 + }, + { + "epoch": 0.30476585163696646, + "grad_norm": 0.2437765896320343, + "learning_rate": 3.4765851636966435e-06, + "loss": 0.4872, + "step": 3677 + }, + { + "epoch": 0.3048487360132615, + "grad_norm": 0.2518393397331238, + "learning_rate": 3.476170741815168e-06, + "loss": 0.5149, + "step": 3678 + }, + { + "epoch": 0.30493162038955657, + "grad_norm": 0.2657361328601837, + "learning_rate": 3.475756319933693e-06, + "loss": 0.5439, + "step": 3679 + }, + { + "epoch": 0.3050145047658516, + "grad_norm": 0.2779373228549957, + "learning_rate": 3.4753418980522176e-06, + "loss": 0.5358, + "step": 3680 + }, + { + "epoch": 0.3050973891421467, + "grad_norm": 0.25099581480026245, + "learning_rate": 3.474927476170742e-06, + "loss": 0.5284, + "step": 3681 + }, + { + "epoch": 0.3051802735184418, + "grad_norm": 0.26496320962905884, + "learning_rate": 3.474513054289267e-06, + "loss": 0.5508, + "step": 3682 + }, + { + "epoch": 0.30526315789473685, + "grad_norm": 0.26970601081848145, + "learning_rate": 3.4740986324077913e-06, + "loss": 0.5356, + "step": 3683 + }, + { + "epoch": 0.3053460422710319, + "grad_norm": 0.271840363740921, + "learning_rate": 3.473684210526316e-06, + "loss": 0.5274, + "step": 3684 + }, + { + "epoch": 0.30542892664732696, + "grad_norm": 0.2584269046783447, + "learning_rate": 3.4732697886448406e-06, + "loss": 0.5518, + "step": 3685 + }, + { + "epoch": 0.30551181102362207, + "grad_norm": 0.25066640973091125, + "learning_rate": 3.472855366763365e-06, + "loss": 0.5079, + "step": 3686 + }, + { + "epoch": 0.3055946953999171, + "grad_norm": 0.2477996051311493, + "learning_rate": 3.4724409448818903e-06, + "loss": 0.5577, + "step": 3687 + }, + { + "epoch": 0.3056775797762122, + "grad_norm": 0.26340317726135254, + "learning_rate": 3.4720265230004148e-06, + "loss": 0.5105, + "step": 3688 + }, + { + "epoch": 0.30576046415250724, + "grad_norm": 0.29224979877471924, + "learning_rate": 3.471612101118939e-06, + "loss": 0.4717, + "step": 3689 + }, + { + "epoch": 0.3058433485288023, + "grad_norm": 0.25287020206451416, + "learning_rate": 3.471197679237464e-06, + "loss": 0.4898, + "step": 3690 + }, + { + "epoch": 0.3059262329050974, + "grad_norm": 0.2648696005344391, + "learning_rate": 3.4707832573559885e-06, + "loss": 0.5402, + "step": 3691 + }, + { + "epoch": 0.30600911728139246, + "grad_norm": 0.2830617129802704, + "learning_rate": 3.4703688354745133e-06, + "loss": 0.5643, + "step": 3692 + }, + { + "epoch": 0.3060920016576875, + "grad_norm": 0.2489991933107376, + "learning_rate": 3.4699544135930378e-06, + "loss": 0.4944, + "step": 3693 + }, + { + "epoch": 0.3061748860339826, + "grad_norm": 0.26465025544166565, + "learning_rate": 3.4695399917115626e-06, + "loss": 0.5605, + "step": 3694 + }, + { + "epoch": 0.3062577704102777, + "grad_norm": 0.30262869596481323, + "learning_rate": 3.4691255698300875e-06, + "loss": 0.5181, + "step": 3695 + }, + { + "epoch": 0.30634065478657274, + "grad_norm": 0.2870234251022339, + "learning_rate": 3.468711147948612e-06, + "loss": 0.6028, + "step": 3696 + }, + { + "epoch": 0.3064235391628678, + "grad_norm": 0.2630198895931244, + "learning_rate": 3.4682967260671368e-06, + "loss": 0.5172, + "step": 3697 + }, + { + "epoch": 0.30650642353916285, + "grad_norm": 0.25717660784721375, + "learning_rate": 3.467882304185661e-06, + "loss": 0.5063, + "step": 3698 + }, + { + "epoch": 0.3065893079154579, + "grad_norm": 0.2615622580051422, + "learning_rate": 3.4674678823041856e-06, + "loss": 0.5344, + "step": 3699 + }, + { + "epoch": 0.306672192291753, + "grad_norm": 0.2629668414592743, + "learning_rate": 3.467053460422711e-06, + "loss": 0.575, + "step": 3700 + }, + { + "epoch": 0.3067550766680481, + "grad_norm": 0.26519399881362915, + "learning_rate": 3.4666390385412353e-06, + "loss": 0.5244, + "step": 3701 + }, + { + "epoch": 0.30683796104434313, + "grad_norm": 0.2629600763320923, + "learning_rate": 3.46622461665976e-06, + "loss": 0.5231, + "step": 3702 + }, + { + "epoch": 0.3069208454206382, + "grad_norm": 0.25094330310821533, + "learning_rate": 3.4658101947782846e-06, + "loss": 0.4788, + "step": 3703 + }, + { + "epoch": 0.3070037297969333, + "grad_norm": 0.25591200590133667, + "learning_rate": 3.465395772896809e-06, + "loss": 0.5386, + "step": 3704 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 0.29421278834342957, + "learning_rate": 3.464981351015334e-06, + "loss": 0.5768, + "step": 3705 + }, + { + "epoch": 0.3071694985495234, + "grad_norm": 0.2544909119606018, + "learning_rate": 3.4645669291338583e-06, + "loss": 0.5463, + "step": 3706 + }, + { + "epoch": 0.30725238292581847, + "grad_norm": 0.26416805386543274, + "learning_rate": 3.4641525072523828e-06, + "loss": 0.5341, + "step": 3707 + }, + { + "epoch": 0.3073352673021135, + "grad_norm": 0.2653396427631378, + "learning_rate": 3.463738085370908e-06, + "loss": 0.524, + "step": 3708 + }, + { + "epoch": 0.30741815167840864, + "grad_norm": 0.24102959036827087, + "learning_rate": 3.4633236634894325e-06, + "loss": 0.4708, + "step": 3709 + }, + { + "epoch": 0.3075010360547037, + "grad_norm": 0.2856687009334564, + "learning_rate": 3.4629092416079573e-06, + "loss": 0.5904, + "step": 3710 + }, + { + "epoch": 0.30758392043099875, + "grad_norm": 0.2940810024738312, + "learning_rate": 3.4624948197264818e-06, + "loss": 0.5325, + "step": 3711 + }, + { + "epoch": 0.3076668048072938, + "grad_norm": 0.2912777066230774, + "learning_rate": 3.462080397845006e-06, + "loss": 0.524, + "step": 3712 + }, + { + "epoch": 0.3077496891835889, + "grad_norm": 0.2651403844356537, + "learning_rate": 3.461665975963531e-06, + "loss": 0.5889, + "step": 3713 + }, + { + "epoch": 0.30783257355988397, + "grad_norm": 0.26105204224586487, + "learning_rate": 3.461251554082056e-06, + "loss": 0.5334, + "step": 3714 + }, + { + "epoch": 0.307915457936179, + "grad_norm": 0.25293874740600586, + "learning_rate": 3.4608371322005808e-06, + "loss": 0.5285, + "step": 3715 + }, + { + "epoch": 0.3079983423124741, + "grad_norm": 0.2938259541988373, + "learning_rate": 3.460422710319105e-06, + "loss": 0.544, + "step": 3716 + }, + { + "epoch": 0.3080812266887692, + "grad_norm": 0.2660636305809021, + "learning_rate": 3.4600082884376296e-06, + "loss": 0.501, + "step": 3717 + }, + { + "epoch": 0.30816411106506425, + "grad_norm": 0.2645530104637146, + "learning_rate": 3.4595938665561545e-06, + "loss": 0.4838, + "step": 3718 + }, + { + "epoch": 0.3082469954413593, + "grad_norm": 0.2519286572933197, + "learning_rate": 3.459179444674679e-06, + "loss": 0.5011, + "step": 3719 + }, + { + "epoch": 0.30832987981765436, + "grad_norm": 0.25133153796195984, + "learning_rate": 3.458765022793204e-06, + "loss": 0.548, + "step": 3720 + }, + { + "epoch": 0.3084127641939494, + "grad_norm": 0.2767024338245392, + "learning_rate": 3.4583506009117286e-06, + "loss": 0.5207, + "step": 3721 + }, + { + "epoch": 0.30849564857024453, + "grad_norm": 0.27269768714904785, + "learning_rate": 3.457936179030253e-06, + "loss": 0.5604, + "step": 3722 + }, + { + "epoch": 0.3085785329465396, + "grad_norm": 0.23850327730178833, + "learning_rate": 3.457521757148778e-06, + "loss": 0.5268, + "step": 3723 + }, + { + "epoch": 0.30866141732283464, + "grad_norm": 0.2588243782520294, + "learning_rate": 3.4571073352673024e-06, + "loss": 0.4959, + "step": 3724 + }, + { + "epoch": 0.3087443016991297, + "grad_norm": 0.28515201807022095, + "learning_rate": 3.4566929133858268e-06, + "loss": 0.5294, + "step": 3725 + }, + { + "epoch": 0.3088271860754248, + "grad_norm": 0.2697819769382477, + "learning_rate": 3.4562784915043516e-06, + "loss": 0.5251, + "step": 3726 + }, + { + "epoch": 0.30891007045171986, + "grad_norm": 0.2565769553184509, + "learning_rate": 3.455864069622876e-06, + "loss": 0.5146, + "step": 3727 + }, + { + "epoch": 0.3089929548280149, + "grad_norm": 0.27540600299835205, + "learning_rate": 3.4554496477414014e-06, + "loss": 0.576, + "step": 3728 + }, + { + "epoch": 0.30907583920431, + "grad_norm": 0.2667192220687866, + "learning_rate": 3.4550352258599258e-06, + "loss": 0.5508, + "step": 3729 + }, + { + "epoch": 0.30915872358060503, + "grad_norm": 0.23820802569389343, + "learning_rate": 3.4546208039784502e-06, + "loss": 0.5276, + "step": 3730 + }, + { + "epoch": 0.30924160795690014, + "grad_norm": 0.2611593008041382, + "learning_rate": 3.454206382096975e-06, + "loss": 0.5583, + "step": 3731 + }, + { + "epoch": 0.3093244923331952, + "grad_norm": 0.2614177167415619, + "learning_rate": 3.4537919602154995e-06, + "loss": 0.5457, + "step": 3732 + }, + { + "epoch": 0.30940737670949026, + "grad_norm": 0.25888949632644653, + "learning_rate": 3.4533775383340244e-06, + "loss": 0.51, + "step": 3733 + }, + { + "epoch": 0.3094902610857853, + "grad_norm": 0.2869052588939667, + "learning_rate": 3.452963116452549e-06, + "loss": 0.5739, + "step": 3734 + }, + { + "epoch": 0.3095731454620804, + "grad_norm": 0.2743317484855652, + "learning_rate": 3.4525486945710736e-06, + "loss": 0.5724, + "step": 3735 + }, + { + "epoch": 0.3096560298383755, + "grad_norm": 0.2633558511734009, + "learning_rate": 3.4521342726895985e-06, + "loss": 0.5396, + "step": 3736 + }, + { + "epoch": 0.30973891421467054, + "grad_norm": 0.26768290996551514, + "learning_rate": 3.451719850808123e-06, + "loss": 0.5403, + "step": 3737 + }, + { + "epoch": 0.3098217985909656, + "grad_norm": 0.25239092111587524, + "learning_rate": 3.451305428926648e-06, + "loss": 0.4647, + "step": 3738 + }, + { + "epoch": 0.30990468296726065, + "grad_norm": 0.24558165669441223, + "learning_rate": 3.4508910070451722e-06, + "loss": 0.5194, + "step": 3739 + }, + { + "epoch": 0.30998756734355576, + "grad_norm": 0.23826411366462708, + "learning_rate": 3.4504765851636967e-06, + "loss": 0.5168, + "step": 3740 + }, + { + "epoch": 0.3100704517198508, + "grad_norm": 0.2599188983440399, + "learning_rate": 3.450062163282222e-06, + "loss": 0.546, + "step": 3741 + }, + { + "epoch": 0.31015333609614587, + "grad_norm": 0.2405068278312683, + "learning_rate": 3.4496477414007464e-06, + "loss": 0.494, + "step": 3742 + }, + { + "epoch": 0.3102362204724409, + "grad_norm": 0.25308579206466675, + "learning_rate": 3.449233319519271e-06, + "loss": 0.5266, + "step": 3743 + }, + { + "epoch": 0.31031910484873604, + "grad_norm": 0.30091509222984314, + "learning_rate": 3.4488188976377956e-06, + "loss": 0.5787, + "step": 3744 + }, + { + "epoch": 0.3104019892250311, + "grad_norm": 0.27628466486930847, + "learning_rate": 3.44840447575632e-06, + "loss": 0.4996, + "step": 3745 + }, + { + "epoch": 0.31048487360132615, + "grad_norm": 0.2484862357378006, + "learning_rate": 3.447990053874845e-06, + "loss": 0.4995, + "step": 3746 + }, + { + "epoch": 0.3105677579776212, + "grad_norm": 0.2639239728450775, + "learning_rate": 3.4475756319933694e-06, + "loss": 0.5485, + "step": 3747 + }, + { + "epoch": 0.31065064235391626, + "grad_norm": 0.27669063210487366, + "learning_rate": 3.447161210111894e-06, + "loss": 0.523, + "step": 3748 + }, + { + "epoch": 0.3107335267302114, + "grad_norm": 0.24115891754627228, + "learning_rate": 3.446746788230419e-06, + "loss": 0.5221, + "step": 3749 + }, + { + "epoch": 0.31081641110650643, + "grad_norm": 0.25548845529556274, + "learning_rate": 3.4463323663489435e-06, + "loss": 0.5313, + "step": 3750 + }, + { + "epoch": 0.3108992954828015, + "grad_norm": 0.2599942088127136, + "learning_rate": 3.4459179444674684e-06, + "loss": 0.5462, + "step": 3751 + }, + { + "epoch": 0.31098217985909654, + "grad_norm": 0.25827473402023315, + "learning_rate": 3.445503522585993e-06, + "loss": 0.5195, + "step": 3752 + }, + { + "epoch": 0.31106506423539165, + "grad_norm": 0.25553178787231445, + "learning_rate": 3.4450891007045172e-06, + "loss": 0.5364, + "step": 3753 + }, + { + "epoch": 0.3111479486116867, + "grad_norm": 0.2525707185268402, + "learning_rate": 3.444674678823042e-06, + "loss": 0.5416, + "step": 3754 + }, + { + "epoch": 0.31123083298798176, + "grad_norm": 0.28033000230789185, + "learning_rate": 3.444260256941567e-06, + "loss": 0.5331, + "step": 3755 + }, + { + "epoch": 0.3113137173642768, + "grad_norm": 0.2700800597667694, + "learning_rate": 3.4438458350600914e-06, + "loss": 0.5606, + "step": 3756 + }, + { + "epoch": 0.3113966017405719, + "grad_norm": 0.2589953541755676, + "learning_rate": 3.4434314131786162e-06, + "loss": 0.494, + "step": 3757 + }, + { + "epoch": 0.311479486116867, + "grad_norm": 0.24488019943237305, + "learning_rate": 3.4430169912971407e-06, + "loss": 0.5151, + "step": 3758 + }, + { + "epoch": 0.31156237049316204, + "grad_norm": 0.30320480465888977, + "learning_rate": 3.4426025694156655e-06, + "loss": 0.5152, + "step": 3759 + }, + { + "epoch": 0.3116452548694571, + "grad_norm": 0.26470452547073364, + "learning_rate": 3.44218814753419e-06, + "loss": 0.5389, + "step": 3760 + }, + { + "epoch": 0.31172813924575216, + "grad_norm": 0.2673277258872986, + "learning_rate": 3.4417737256527144e-06, + "loss": 0.5332, + "step": 3761 + }, + { + "epoch": 0.31181102362204727, + "grad_norm": 0.25452423095703125, + "learning_rate": 3.4413593037712397e-06, + "loss": 0.5248, + "step": 3762 + }, + { + "epoch": 0.3118939079983423, + "grad_norm": 0.26297539472579956, + "learning_rate": 3.440944881889764e-06, + "loss": 0.5981, + "step": 3763 + }, + { + "epoch": 0.3119767923746374, + "grad_norm": 0.27562716603279114, + "learning_rate": 3.440530460008289e-06, + "loss": 0.5775, + "step": 3764 + }, + { + "epoch": 0.31205967675093244, + "grad_norm": 0.24740520119667053, + "learning_rate": 3.4401160381268134e-06, + "loss": 0.5279, + "step": 3765 + }, + { + "epoch": 0.3121425611272275, + "grad_norm": 0.2565785348415375, + "learning_rate": 3.439701616245338e-06, + "loss": 0.5239, + "step": 3766 + }, + { + "epoch": 0.3122254455035226, + "grad_norm": 0.2542875111103058, + "learning_rate": 3.4392871943638627e-06, + "loss": 0.5444, + "step": 3767 + }, + { + "epoch": 0.31230832987981766, + "grad_norm": 0.2624465525150299, + "learning_rate": 3.438872772482387e-06, + "loss": 0.5436, + "step": 3768 + }, + { + "epoch": 0.3123912142561127, + "grad_norm": 0.25606656074523926, + "learning_rate": 3.4384583506009124e-06, + "loss": 0.5044, + "step": 3769 + }, + { + "epoch": 0.31247409863240777, + "grad_norm": 0.2447153478860855, + "learning_rate": 3.438043928719437e-06, + "loss": 0.5297, + "step": 3770 + }, + { + "epoch": 0.3125569830087029, + "grad_norm": 0.25946736335754395, + "learning_rate": 3.4376295068379612e-06, + "loss": 0.4998, + "step": 3771 + }, + { + "epoch": 0.31263986738499794, + "grad_norm": 0.26057615876197815, + "learning_rate": 3.437215084956486e-06, + "loss": 0.5012, + "step": 3772 + }, + { + "epoch": 0.312722751761293, + "grad_norm": 0.2730274200439453, + "learning_rate": 3.4368006630750105e-06, + "loss": 0.5075, + "step": 3773 + }, + { + "epoch": 0.31280563613758805, + "grad_norm": 0.2905065715312958, + "learning_rate": 3.436386241193535e-06, + "loss": 0.5288, + "step": 3774 + }, + { + "epoch": 0.3128885205138831, + "grad_norm": 0.27243682742118835, + "learning_rate": 3.4359718193120602e-06, + "loss": 0.5453, + "step": 3775 + }, + { + "epoch": 0.3129714048901782, + "grad_norm": 0.2627275586128235, + "learning_rate": 3.4355573974305847e-06, + "loss": 0.4985, + "step": 3776 + }, + { + "epoch": 0.3130542892664733, + "grad_norm": 0.2736596167087555, + "learning_rate": 3.4351429755491095e-06, + "loss": 0.5499, + "step": 3777 + }, + { + "epoch": 0.31313717364276833, + "grad_norm": 0.2796374559402466, + "learning_rate": 3.434728553667634e-06, + "loss": 0.574, + "step": 3778 + }, + { + "epoch": 0.3132200580190634, + "grad_norm": 0.26440441608428955, + "learning_rate": 3.4343141317861584e-06, + "loss": 0.5204, + "step": 3779 + }, + { + "epoch": 0.3133029423953585, + "grad_norm": 0.26491525769233704, + "learning_rate": 3.4338997099046832e-06, + "loss": 0.5538, + "step": 3780 + }, + { + "epoch": 0.31338582677165355, + "grad_norm": 0.2341463714838028, + "learning_rate": 3.4334852880232077e-06, + "loss": 0.4879, + "step": 3781 + }, + { + "epoch": 0.3134687111479486, + "grad_norm": 0.27948492765426636, + "learning_rate": 3.433070866141733e-06, + "loss": 0.5619, + "step": 3782 + }, + { + "epoch": 0.31355159552424366, + "grad_norm": 0.282562255859375, + "learning_rate": 3.4326564442602574e-06, + "loss": 0.5714, + "step": 3783 + }, + { + "epoch": 0.3136344799005387, + "grad_norm": 0.24220354855060577, + "learning_rate": 3.432242022378782e-06, + "loss": 0.5379, + "step": 3784 + }, + { + "epoch": 0.31371736427683383, + "grad_norm": 0.2508859634399414, + "learning_rate": 3.4318276004973067e-06, + "loss": 0.5115, + "step": 3785 + }, + { + "epoch": 0.3138002486531289, + "grad_norm": 0.26520103216171265, + "learning_rate": 3.431413178615831e-06, + "loss": 0.5474, + "step": 3786 + }, + { + "epoch": 0.31388313302942394, + "grad_norm": 0.2953545153141022, + "learning_rate": 3.430998756734356e-06, + "loss": 0.6177, + "step": 3787 + }, + { + "epoch": 0.313966017405719, + "grad_norm": 0.24866093695163727, + "learning_rate": 3.4305843348528804e-06, + "loss": 0.5579, + "step": 3788 + }, + { + "epoch": 0.3140489017820141, + "grad_norm": 0.25600680708885193, + "learning_rate": 3.430169912971405e-06, + "loss": 0.5044, + "step": 3789 + }, + { + "epoch": 0.31413178615830917, + "grad_norm": 0.2675791084766388, + "learning_rate": 3.42975549108993e-06, + "loss": 0.5174, + "step": 3790 + }, + { + "epoch": 0.3142146705346042, + "grad_norm": 0.26150447130203247, + "learning_rate": 3.4293410692084545e-06, + "loss": 0.5182, + "step": 3791 + }, + { + "epoch": 0.3142975549108993, + "grad_norm": 0.2853640615940094, + "learning_rate": 3.428926647326979e-06, + "loss": 0.5768, + "step": 3792 + }, + { + "epoch": 0.3143804392871944, + "grad_norm": 0.259455144405365, + "learning_rate": 3.428512225445504e-06, + "loss": 0.5207, + "step": 3793 + }, + { + "epoch": 0.31446332366348945, + "grad_norm": 0.2595706880092621, + "learning_rate": 3.4280978035640283e-06, + "loss": 0.5563, + "step": 3794 + }, + { + "epoch": 0.3145462080397845, + "grad_norm": 0.24470506608486176, + "learning_rate": 3.4276833816825535e-06, + "loss": 0.5063, + "step": 3795 + }, + { + "epoch": 0.31462909241607956, + "grad_norm": 0.27614349126815796, + "learning_rate": 3.427268959801078e-06, + "loss": 0.5802, + "step": 3796 + }, + { + "epoch": 0.3147119767923746, + "grad_norm": 0.25122934579849243, + "learning_rate": 3.4268545379196024e-06, + "loss": 0.4958, + "step": 3797 + }, + { + "epoch": 0.3147948611686697, + "grad_norm": 0.2385462373495102, + "learning_rate": 3.4264401160381272e-06, + "loss": 0.4732, + "step": 3798 + }, + { + "epoch": 0.3148777455449648, + "grad_norm": 0.25731611251831055, + "learning_rate": 3.4260256941566517e-06, + "loss": 0.5312, + "step": 3799 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 0.255347341299057, + "learning_rate": 3.4256112722751765e-06, + "loss": 0.533, + "step": 3800 + }, + { + "epoch": 0.3150435142975549, + "grad_norm": 0.2624645233154297, + "learning_rate": 3.425196850393701e-06, + "loss": 0.5602, + "step": 3801 + }, + { + "epoch": 0.31512639867385, + "grad_norm": 0.2741098701953888, + "learning_rate": 3.4247824285122254e-06, + "loss": 0.5394, + "step": 3802 + }, + { + "epoch": 0.31520928305014506, + "grad_norm": 0.2508790194988251, + "learning_rate": 3.4243680066307507e-06, + "loss": 0.5277, + "step": 3803 + }, + { + "epoch": 0.3152921674264401, + "grad_norm": 0.2571517527103424, + "learning_rate": 3.423953584749275e-06, + "loss": 0.5266, + "step": 3804 + }, + { + "epoch": 0.3153750518027352, + "grad_norm": 0.25354641675949097, + "learning_rate": 3.4235391628678e-06, + "loss": 0.5215, + "step": 3805 + }, + { + "epoch": 0.31545793617903023, + "grad_norm": 0.2648247480392456, + "learning_rate": 3.4231247409863244e-06, + "loss": 0.4881, + "step": 3806 + }, + { + "epoch": 0.31554082055532534, + "grad_norm": 0.24475178122520447, + "learning_rate": 3.422710319104849e-06, + "loss": 0.5244, + "step": 3807 + }, + { + "epoch": 0.3156237049316204, + "grad_norm": 0.2616937458515167, + "learning_rate": 3.4222958972233737e-06, + "loss": 0.5221, + "step": 3808 + }, + { + "epoch": 0.31570658930791545, + "grad_norm": 0.26383915543556213, + "learning_rate": 3.421881475341898e-06, + "loss": 0.5365, + "step": 3809 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 0.2525240480899811, + "learning_rate": 3.4214670534604225e-06, + "loss": 0.4933, + "step": 3810 + }, + { + "epoch": 0.3158723580605056, + "grad_norm": 0.2550865411758423, + "learning_rate": 3.421052631578948e-06, + "loss": 0.5414, + "step": 3811 + }, + { + "epoch": 0.3159552424368007, + "grad_norm": 0.27922701835632324, + "learning_rate": 3.4206382096974723e-06, + "loss": 0.5629, + "step": 3812 + }, + { + "epoch": 0.31603812681309573, + "grad_norm": 0.25223681330680847, + "learning_rate": 3.420223787815997e-06, + "loss": 0.5373, + "step": 3813 + }, + { + "epoch": 0.3161210111893908, + "grad_norm": 0.26813608407974243, + "learning_rate": 3.4198093659345215e-06, + "loss": 0.5369, + "step": 3814 + }, + { + "epoch": 0.31620389556568584, + "grad_norm": 0.2758404314517975, + "learning_rate": 3.419394944053046e-06, + "loss": 0.5306, + "step": 3815 + }, + { + "epoch": 0.31628677994198096, + "grad_norm": 0.2498100847005844, + "learning_rate": 3.4189805221715713e-06, + "loss": 0.5203, + "step": 3816 + }, + { + "epoch": 0.316369664318276, + "grad_norm": 0.2436426281929016, + "learning_rate": 3.4185661002900957e-06, + "loss": 0.4455, + "step": 3817 + }, + { + "epoch": 0.31645254869457107, + "grad_norm": 0.263857901096344, + "learning_rate": 3.4181516784086205e-06, + "loss": 0.5541, + "step": 3818 + }, + { + "epoch": 0.3165354330708661, + "grad_norm": 0.2303079217672348, + "learning_rate": 3.417737256527145e-06, + "loss": 0.4993, + "step": 3819 + }, + { + "epoch": 0.31661831744716123, + "grad_norm": 0.3045215904712677, + "learning_rate": 3.4173228346456694e-06, + "loss": 0.5237, + "step": 3820 + }, + { + "epoch": 0.3167012018234563, + "grad_norm": 0.29123935103416443, + "learning_rate": 3.4169084127641943e-06, + "loss": 0.5964, + "step": 3821 + }, + { + "epoch": 0.31678408619975135, + "grad_norm": 0.27974486351013184, + "learning_rate": 3.4164939908827187e-06, + "loss": 0.5305, + "step": 3822 + }, + { + "epoch": 0.3168669705760464, + "grad_norm": 0.268954873085022, + "learning_rate": 3.416079569001243e-06, + "loss": 0.5326, + "step": 3823 + }, + { + "epoch": 0.31694985495234146, + "grad_norm": 0.2764497697353363, + "learning_rate": 3.4156651471197684e-06, + "loss": 0.5433, + "step": 3824 + }, + { + "epoch": 0.31703273932863657, + "grad_norm": 0.24992816150188446, + "learning_rate": 3.415250725238293e-06, + "loss": 0.5064, + "step": 3825 + }, + { + "epoch": 0.3171156237049316, + "grad_norm": 0.24964459240436554, + "learning_rate": 3.4148363033568177e-06, + "loss": 0.5618, + "step": 3826 + }, + { + "epoch": 0.3171985080812267, + "grad_norm": 0.24827973544597626, + "learning_rate": 3.414421881475342e-06, + "loss": 0.5249, + "step": 3827 + }, + { + "epoch": 0.31728139245752174, + "grad_norm": 0.2587418854236603, + "learning_rate": 3.4140074595938666e-06, + "loss": 0.5197, + "step": 3828 + }, + { + "epoch": 0.31736427683381685, + "grad_norm": 0.28337132930755615, + "learning_rate": 3.4135930377123914e-06, + "loss": 0.5383, + "step": 3829 + }, + { + "epoch": 0.3174471612101119, + "grad_norm": 0.28459855914115906, + "learning_rate": 3.413178615830916e-06, + "loss": 0.5656, + "step": 3830 + }, + { + "epoch": 0.31753004558640696, + "grad_norm": 0.2571375072002411, + "learning_rate": 3.412764193949441e-06, + "loss": 0.5392, + "step": 3831 + }, + { + "epoch": 0.317612929962702, + "grad_norm": 0.2856287658214569, + "learning_rate": 3.4123497720679656e-06, + "loss": 0.5589, + "step": 3832 + }, + { + "epoch": 0.3176958143389971, + "grad_norm": 0.26867377758026123, + "learning_rate": 3.41193535018649e-06, + "loss": 0.5495, + "step": 3833 + }, + { + "epoch": 0.3177786987152922, + "grad_norm": 0.26073479652404785, + "learning_rate": 3.411520928305015e-06, + "loss": 0.5606, + "step": 3834 + }, + { + "epoch": 0.31786158309158724, + "grad_norm": 0.262757271528244, + "learning_rate": 3.4111065064235393e-06, + "loss": 0.5385, + "step": 3835 + }, + { + "epoch": 0.3179444674678823, + "grad_norm": 0.2717835307121277, + "learning_rate": 3.4106920845420646e-06, + "loss": 0.5898, + "step": 3836 + }, + { + "epoch": 0.31802735184417735, + "grad_norm": 0.2798512578010559, + "learning_rate": 3.410277662660589e-06, + "loss": 0.5851, + "step": 3837 + }, + { + "epoch": 0.31811023622047246, + "grad_norm": 0.26110169291496277, + "learning_rate": 3.4098632407791134e-06, + "loss": 0.5132, + "step": 3838 + }, + { + "epoch": 0.3181931205967675, + "grad_norm": 0.2460823953151703, + "learning_rate": 3.4094488188976383e-06, + "loss": 0.5085, + "step": 3839 + }, + { + "epoch": 0.3182760049730626, + "grad_norm": 0.25192487239837646, + "learning_rate": 3.4090343970161627e-06, + "loss": 0.5262, + "step": 3840 + }, + { + "epoch": 0.31835888934935763, + "grad_norm": 0.29762542247772217, + "learning_rate": 3.408619975134687e-06, + "loss": 0.5613, + "step": 3841 + }, + { + "epoch": 0.3184417737256527, + "grad_norm": 0.2843422591686249, + "learning_rate": 3.408205553253212e-06, + "loss": 0.5702, + "step": 3842 + }, + { + "epoch": 0.3185246581019478, + "grad_norm": 0.2527610957622528, + "learning_rate": 3.4077911313717364e-06, + "loss": 0.5352, + "step": 3843 + }, + { + "epoch": 0.31860754247824286, + "grad_norm": 0.24883732199668884, + "learning_rate": 3.4073767094902617e-06, + "loss": 0.4984, + "step": 3844 + }, + { + "epoch": 0.3186904268545379, + "grad_norm": 0.27329036593437195, + "learning_rate": 3.406962287608786e-06, + "loss": 0.5542, + "step": 3845 + }, + { + "epoch": 0.31877331123083297, + "grad_norm": 0.2682969570159912, + "learning_rate": 3.4065478657273106e-06, + "loss": 0.556, + "step": 3846 + }, + { + "epoch": 0.3188561956071281, + "grad_norm": 0.2591707110404968, + "learning_rate": 3.4061334438458354e-06, + "loss": 0.5156, + "step": 3847 + }, + { + "epoch": 0.31893907998342313, + "grad_norm": 0.2426518052816391, + "learning_rate": 3.40571902196436e-06, + "loss": 0.5216, + "step": 3848 + }, + { + "epoch": 0.3190219643597182, + "grad_norm": 0.27413010597229004, + "learning_rate": 3.4053046000828847e-06, + "loss": 0.5154, + "step": 3849 + }, + { + "epoch": 0.31910484873601325, + "grad_norm": 0.2501504421234131, + "learning_rate": 3.404890178201409e-06, + "loss": 0.5566, + "step": 3850 + }, + { + "epoch": 0.3191877331123083, + "grad_norm": 0.2647289037704468, + "learning_rate": 3.4044757563199336e-06, + "loss": 0.5199, + "step": 3851 + }, + { + "epoch": 0.3192706174886034, + "grad_norm": 0.26438960433006287, + "learning_rate": 3.404061334438459e-06, + "loss": 0.5177, + "step": 3852 + }, + { + "epoch": 0.31935350186489847, + "grad_norm": 0.2608979344367981, + "learning_rate": 3.4036469125569833e-06, + "loss": 0.5154, + "step": 3853 + }, + { + "epoch": 0.3194363862411935, + "grad_norm": 0.29590675234794617, + "learning_rate": 3.403232490675508e-06, + "loss": 0.4825, + "step": 3854 + }, + { + "epoch": 0.3195192706174886, + "grad_norm": 0.26486629247665405, + "learning_rate": 3.4028180687940326e-06, + "loss": 0.5845, + "step": 3855 + }, + { + "epoch": 0.3196021549937837, + "grad_norm": 0.2682054042816162, + "learning_rate": 3.402403646912557e-06, + "loss": 0.5001, + "step": 3856 + }, + { + "epoch": 0.31968503937007875, + "grad_norm": 0.26158806681632996, + "learning_rate": 3.4019892250310823e-06, + "loss": 0.5635, + "step": 3857 + }, + { + "epoch": 0.3197679237463738, + "grad_norm": 0.2595740258693695, + "learning_rate": 3.4015748031496067e-06, + "loss": 0.5193, + "step": 3858 + }, + { + "epoch": 0.31985080812266886, + "grad_norm": 0.2594318687915802, + "learning_rate": 3.401160381268131e-06, + "loss": 0.5239, + "step": 3859 + }, + { + "epoch": 0.319933692498964, + "grad_norm": 0.2631060481071472, + "learning_rate": 3.400745959386656e-06, + "loss": 0.5018, + "step": 3860 + }, + { + "epoch": 0.32001657687525903, + "grad_norm": 0.2581394612789154, + "learning_rate": 3.4003315375051804e-06, + "loss": 0.5524, + "step": 3861 + }, + { + "epoch": 0.3200994612515541, + "grad_norm": 0.2515861690044403, + "learning_rate": 3.3999171156237053e-06, + "loss": 0.4631, + "step": 3862 + }, + { + "epoch": 0.32018234562784914, + "grad_norm": 0.25171273946762085, + "learning_rate": 3.3995026937422297e-06, + "loss": 0.4821, + "step": 3863 + }, + { + "epoch": 0.3202652300041442, + "grad_norm": 0.26351359486579895, + "learning_rate": 3.399088271860754e-06, + "loss": 0.5518, + "step": 3864 + }, + { + "epoch": 0.3203481143804393, + "grad_norm": 0.2645649015903473, + "learning_rate": 3.3986738499792794e-06, + "loss": 0.5469, + "step": 3865 + }, + { + "epoch": 0.32043099875673436, + "grad_norm": 0.28918373584747314, + "learning_rate": 3.398259428097804e-06, + "loss": 0.5463, + "step": 3866 + }, + { + "epoch": 0.3205138831330294, + "grad_norm": 0.28055933117866516, + "learning_rate": 3.3978450062163287e-06, + "loss": 0.55, + "step": 3867 + }, + { + "epoch": 0.3205967675093245, + "grad_norm": 0.27645033597946167, + "learning_rate": 3.397430584334853e-06, + "loss": 0.5623, + "step": 3868 + }, + { + "epoch": 0.3206796518856196, + "grad_norm": 0.3010002374649048, + "learning_rate": 3.3970161624533776e-06, + "loss": 0.6166, + "step": 3869 + }, + { + "epoch": 0.32076253626191464, + "grad_norm": 0.2525491714477539, + "learning_rate": 3.3966017405719024e-06, + "loss": 0.5349, + "step": 3870 + }, + { + "epoch": 0.3208454206382097, + "grad_norm": 0.25244325399398804, + "learning_rate": 3.396187318690427e-06, + "loss": 0.5038, + "step": 3871 + }, + { + "epoch": 0.32092830501450476, + "grad_norm": 0.25294601917266846, + "learning_rate": 3.395772896808952e-06, + "loss": 0.5482, + "step": 3872 + }, + { + "epoch": 0.3210111893907998, + "grad_norm": 0.2732849717140198, + "learning_rate": 3.3953584749274766e-06, + "loss": 0.5268, + "step": 3873 + }, + { + "epoch": 0.3210940737670949, + "grad_norm": 0.2643308639526367, + "learning_rate": 3.394944053046001e-06, + "loss": 0.5474, + "step": 3874 + }, + { + "epoch": 0.32117695814339, + "grad_norm": 0.26631608605384827, + "learning_rate": 3.394529631164526e-06, + "loss": 0.5539, + "step": 3875 + }, + { + "epoch": 0.32125984251968503, + "grad_norm": 0.2516079545021057, + "learning_rate": 3.3941152092830503e-06, + "loss": 0.5472, + "step": 3876 + }, + { + "epoch": 0.3213427268959801, + "grad_norm": 0.2585592269897461, + "learning_rate": 3.3937007874015747e-06, + "loss": 0.5383, + "step": 3877 + }, + { + "epoch": 0.3214256112722752, + "grad_norm": 0.2721462547779083, + "learning_rate": 3.3932863655201e-06, + "loss": 0.5336, + "step": 3878 + }, + { + "epoch": 0.32150849564857026, + "grad_norm": 0.2613210082054138, + "learning_rate": 3.3928719436386244e-06, + "loss": 0.549, + "step": 3879 + }, + { + "epoch": 0.3215913800248653, + "grad_norm": 0.26434817910194397, + "learning_rate": 3.3924575217571493e-06, + "loss": 0.5246, + "step": 3880 + }, + { + "epoch": 0.32167426440116037, + "grad_norm": 0.25483638048171997, + "learning_rate": 3.3920430998756737e-06, + "loss": 0.4908, + "step": 3881 + }, + { + "epoch": 0.3217571487774554, + "grad_norm": 0.28605085611343384, + "learning_rate": 3.391628677994198e-06, + "loss": 0.5594, + "step": 3882 + }, + { + "epoch": 0.32184003315375054, + "grad_norm": 0.28096112608909607, + "learning_rate": 3.391214256112723e-06, + "loss": 0.5693, + "step": 3883 + }, + { + "epoch": 0.3219229175300456, + "grad_norm": 0.2433292269706726, + "learning_rate": 3.3907998342312474e-06, + "loss": 0.5072, + "step": 3884 + }, + { + "epoch": 0.32200580190634065, + "grad_norm": 0.2699735462665558, + "learning_rate": 3.3903854123497727e-06, + "loss": 0.5569, + "step": 3885 + }, + { + "epoch": 0.3220886862826357, + "grad_norm": 0.2739693820476532, + "learning_rate": 3.389970990468297e-06, + "loss": 0.5604, + "step": 3886 + }, + { + "epoch": 0.3221715706589308, + "grad_norm": 0.24786929786205292, + "learning_rate": 3.3895565685868216e-06, + "loss": 0.5024, + "step": 3887 + }, + { + "epoch": 0.3222544550352259, + "grad_norm": 0.2599262297153473, + "learning_rate": 3.3891421467053464e-06, + "loss": 0.5159, + "step": 3888 + }, + { + "epoch": 0.32233733941152093, + "grad_norm": 0.24777860939502716, + "learning_rate": 3.388727724823871e-06, + "loss": 0.5175, + "step": 3889 + }, + { + "epoch": 0.322420223787816, + "grad_norm": 0.2725672125816345, + "learning_rate": 3.3883133029423957e-06, + "loss": 0.5029, + "step": 3890 + }, + { + "epoch": 0.32250310816411104, + "grad_norm": 0.23997053503990173, + "learning_rate": 3.38789888106092e-06, + "loss": 0.5148, + "step": 3891 + }, + { + "epoch": 0.32258599254040615, + "grad_norm": 0.26027169823646545, + "learning_rate": 3.3874844591794446e-06, + "loss": 0.4841, + "step": 3892 + }, + { + "epoch": 0.3226688769167012, + "grad_norm": 0.25397151708602905, + "learning_rate": 3.38707003729797e-06, + "loss": 0.5271, + "step": 3893 + }, + { + "epoch": 0.32275176129299626, + "grad_norm": 0.2549907863140106, + "learning_rate": 3.3866556154164943e-06, + "loss": 0.5215, + "step": 3894 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 0.24786025285720825, + "learning_rate": 3.3862411935350187e-06, + "loss": 0.513, + "step": 3895 + }, + { + "epoch": 0.32291753004558643, + "grad_norm": 0.2454063892364502, + "learning_rate": 3.3858267716535436e-06, + "loss": 0.5164, + "step": 3896 + }, + { + "epoch": 0.3230004144218815, + "grad_norm": 0.26998502016067505, + "learning_rate": 3.385412349772068e-06, + "loss": 0.5636, + "step": 3897 + }, + { + "epoch": 0.32308329879817654, + "grad_norm": 0.28568416833877563, + "learning_rate": 3.3849979278905933e-06, + "loss": 0.5927, + "step": 3898 + }, + { + "epoch": 0.3231661831744716, + "grad_norm": 0.24796020984649658, + "learning_rate": 3.3845835060091177e-06, + "loss": 0.5061, + "step": 3899 + }, + { + "epoch": 0.32324906755076666, + "grad_norm": 0.25212186574935913, + "learning_rate": 3.384169084127642e-06, + "loss": 0.4974, + "step": 3900 + }, + { + "epoch": 0.32333195192706177, + "grad_norm": 0.25416073203086853, + "learning_rate": 3.383754662246167e-06, + "loss": 0.4953, + "step": 3901 + }, + { + "epoch": 0.3234148363033568, + "grad_norm": 0.2323729544878006, + "learning_rate": 3.3833402403646914e-06, + "loss": 0.4566, + "step": 3902 + }, + { + "epoch": 0.3234977206796519, + "grad_norm": 0.2693474590778351, + "learning_rate": 3.3829258184832163e-06, + "loss": 0.591, + "step": 3903 + }, + { + "epoch": 0.32358060505594693, + "grad_norm": 0.2540178894996643, + "learning_rate": 3.3825113966017407e-06, + "loss": 0.4637, + "step": 3904 + }, + { + "epoch": 0.32366348943224205, + "grad_norm": 0.2807297706604004, + "learning_rate": 3.382096974720265e-06, + "loss": 0.6131, + "step": 3905 + }, + { + "epoch": 0.3237463738085371, + "grad_norm": 0.24072349071502686, + "learning_rate": 3.3816825528387904e-06, + "loss": 0.5295, + "step": 3906 + }, + { + "epoch": 0.32382925818483216, + "grad_norm": 0.2734241485595703, + "learning_rate": 3.381268130957315e-06, + "loss": 0.5508, + "step": 3907 + }, + { + "epoch": 0.3239121425611272, + "grad_norm": 0.2314417064189911, + "learning_rate": 3.3808537090758393e-06, + "loss": 0.4818, + "step": 3908 + }, + { + "epoch": 0.32399502693742227, + "grad_norm": 0.2884172797203064, + "learning_rate": 3.380439287194364e-06, + "loss": 0.5622, + "step": 3909 + }, + { + "epoch": 0.3240779113137174, + "grad_norm": 0.2708169221878052, + "learning_rate": 3.3800248653128886e-06, + "loss": 0.5428, + "step": 3910 + }, + { + "epoch": 0.32416079569001244, + "grad_norm": 0.26906564831733704, + "learning_rate": 3.3796104434314135e-06, + "loss": 0.529, + "step": 3911 + }, + { + "epoch": 0.3242436800663075, + "grad_norm": 0.2552347779273987, + "learning_rate": 3.379196021549938e-06, + "loss": 0.5387, + "step": 3912 + }, + { + "epoch": 0.32432656444260255, + "grad_norm": 0.2399696707725525, + "learning_rate": 3.3787815996684627e-06, + "loss": 0.5136, + "step": 3913 + }, + { + "epoch": 0.32440944881889766, + "grad_norm": 0.27754366397857666, + "learning_rate": 3.3783671777869876e-06, + "loss": 0.5431, + "step": 3914 + }, + { + "epoch": 0.3244923331951927, + "grad_norm": 0.24912524223327637, + "learning_rate": 3.377952755905512e-06, + "loss": 0.5466, + "step": 3915 + }, + { + "epoch": 0.3245752175714878, + "grad_norm": 0.24528741836547852, + "learning_rate": 3.377538334024037e-06, + "loss": 0.4934, + "step": 3916 + }, + { + "epoch": 0.32465810194778283, + "grad_norm": 0.26939383149147034, + "learning_rate": 3.3771239121425613e-06, + "loss": 0.5421, + "step": 3917 + }, + { + "epoch": 0.3247409863240779, + "grad_norm": 0.2767173647880554, + "learning_rate": 3.3767094902610857e-06, + "loss": 0.5741, + "step": 3918 + }, + { + "epoch": 0.324823870700373, + "grad_norm": 0.2667793035507202, + "learning_rate": 3.376295068379611e-06, + "loss": 0.5049, + "step": 3919 + }, + { + "epoch": 0.32490675507666805, + "grad_norm": 0.2798005938529968, + "learning_rate": 3.3758806464981355e-06, + "loss": 0.5362, + "step": 3920 + }, + { + "epoch": 0.3249896394529631, + "grad_norm": 0.24194875359535217, + "learning_rate": 3.3754662246166603e-06, + "loss": 0.5007, + "step": 3921 + }, + { + "epoch": 0.32507252382925816, + "grad_norm": 0.25713032484054565, + "learning_rate": 3.3750518027351847e-06, + "loss": 0.537, + "step": 3922 + }, + { + "epoch": 0.3251554082055533, + "grad_norm": 0.2920835018157959, + "learning_rate": 3.374637380853709e-06, + "loss": 0.5497, + "step": 3923 + }, + { + "epoch": 0.32523829258184833, + "grad_norm": 0.2597065567970276, + "learning_rate": 3.374222958972234e-06, + "loss": 0.54, + "step": 3924 + }, + { + "epoch": 0.3253211769581434, + "grad_norm": 0.2999047636985779, + "learning_rate": 3.3738085370907585e-06, + "loss": 0.4958, + "step": 3925 + }, + { + "epoch": 0.32540406133443844, + "grad_norm": 0.2517358362674713, + "learning_rate": 3.373394115209283e-06, + "loss": 0.492, + "step": 3926 + }, + { + "epoch": 0.3254869457107335, + "grad_norm": 0.2744602560997009, + "learning_rate": 3.372979693327808e-06, + "loss": 0.5594, + "step": 3927 + }, + { + "epoch": 0.3255698300870286, + "grad_norm": 0.27405065298080444, + "learning_rate": 3.3725652714463326e-06, + "loss": 0.5747, + "step": 3928 + }, + { + "epoch": 0.32565271446332367, + "grad_norm": 0.253113329410553, + "learning_rate": 3.3721508495648575e-06, + "loss": 0.5326, + "step": 3929 + }, + { + "epoch": 0.3257355988396187, + "grad_norm": 0.24924834072589874, + "learning_rate": 3.371736427683382e-06, + "loss": 0.5116, + "step": 3930 + }, + { + "epoch": 0.3258184832159138, + "grad_norm": 0.25189536809921265, + "learning_rate": 3.3713220058019063e-06, + "loss": 0.4963, + "step": 3931 + }, + { + "epoch": 0.3259013675922089, + "grad_norm": 0.28706130385398865, + "learning_rate": 3.370907583920431e-06, + "loss": 0.5158, + "step": 3932 + }, + { + "epoch": 0.32598425196850395, + "grad_norm": 0.2691611051559448, + "learning_rate": 3.370493162038956e-06, + "loss": 0.5104, + "step": 3933 + }, + { + "epoch": 0.326067136344799, + "grad_norm": 0.2581654191017151, + "learning_rate": 3.370078740157481e-06, + "loss": 0.5018, + "step": 3934 + }, + { + "epoch": 0.32615002072109406, + "grad_norm": 0.2506330609321594, + "learning_rate": 3.3696643182760053e-06, + "loss": 0.5056, + "step": 3935 + }, + { + "epoch": 0.32623290509738917, + "grad_norm": 0.27758416533470154, + "learning_rate": 3.3692498963945298e-06, + "loss": 0.5876, + "step": 3936 + }, + { + "epoch": 0.3263157894736842, + "grad_norm": 0.2447052299976349, + "learning_rate": 3.3688354745130546e-06, + "loss": 0.5148, + "step": 3937 + }, + { + "epoch": 0.3263986738499793, + "grad_norm": 0.2545480728149414, + "learning_rate": 3.368421052631579e-06, + "loss": 0.5077, + "step": 3938 + }, + { + "epoch": 0.32648155822627434, + "grad_norm": 0.24815413355827332, + "learning_rate": 3.3680066307501043e-06, + "loss": 0.5185, + "step": 3939 + }, + { + "epoch": 0.3265644426025694, + "grad_norm": 0.2451602816581726, + "learning_rate": 3.3675922088686288e-06, + "loss": 0.5193, + "step": 3940 + }, + { + "epoch": 0.3266473269788645, + "grad_norm": 0.2379695326089859, + "learning_rate": 3.367177786987153e-06, + "loss": 0.4765, + "step": 3941 + }, + { + "epoch": 0.32673021135515956, + "grad_norm": 0.5071398019790649, + "learning_rate": 3.366763365105678e-06, + "loss": 0.5328, + "step": 3942 + }, + { + "epoch": 0.3268130957314546, + "grad_norm": 0.25903448462486267, + "learning_rate": 3.3663489432242025e-06, + "loss": 0.5371, + "step": 3943 + }, + { + "epoch": 0.3268959801077497, + "grad_norm": 0.28108328580856323, + "learning_rate": 3.365934521342727e-06, + "loss": 0.5706, + "step": 3944 + }, + { + "epoch": 0.3269788644840448, + "grad_norm": 0.2923888564109802, + "learning_rate": 3.3655200994612518e-06, + "loss": 0.5902, + "step": 3945 + }, + { + "epoch": 0.32706174886033984, + "grad_norm": 0.2435116469860077, + "learning_rate": 3.365105677579776e-06, + "loss": 0.5411, + "step": 3946 + }, + { + "epoch": 0.3271446332366349, + "grad_norm": 0.2728510797023773, + "learning_rate": 3.3646912556983015e-06, + "loss": 0.5768, + "step": 3947 + }, + { + "epoch": 0.32722751761292995, + "grad_norm": 0.25791671872138977, + "learning_rate": 3.364276833816826e-06, + "loss": 0.5164, + "step": 3948 + }, + { + "epoch": 0.327310401989225, + "grad_norm": 0.2649441957473755, + "learning_rate": 3.3638624119353503e-06, + "loss": 0.5041, + "step": 3949 + }, + { + "epoch": 0.3273932863655201, + "grad_norm": 0.27235978841781616, + "learning_rate": 3.363447990053875e-06, + "loss": 0.5426, + "step": 3950 + }, + { + "epoch": 0.3274761707418152, + "grad_norm": 0.2668672204017639, + "learning_rate": 3.3630335681723996e-06, + "loss": 0.538, + "step": 3951 + }, + { + "epoch": 0.32755905511811023, + "grad_norm": 0.2766541540622711, + "learning_rate": 3.3626191462909245e-06, + "loss": 0.5243, + "step": 3952 + }, + { + "epoch": 0.3276419394944053, + "grad_norm": 0.24921685457229614, + "learning_rate": 3.362204724409449e-06, + "loss": 0.4997, + "step": 3953 + }, + { + "epoch": 0.3277248238707004, + "grad_norm": 0.2535344660282135, + "learning_rate": 3.3617903025279738e-06, + "loss": 0.5158, + "step": 3954 + }, + { + "epoch": 0.32780770824699546, + "grad_norm": 0.27182263135910034, + "learning_rate": 3.3613758806464986e-06, + "loss": 0.5265, + "step": 3955 + }, + { + "epoch": 0.3278905926232905, + "grad_norm": 0.26011648774147034, + "learning_rate": 3.360961458765023e-06, + "loss": 0.53, + "step": 3956 + }, + { + "epoch": 0.32797347699958557, + "grad_norm": 0.27622759342193604, + "learning_rate": 3.360547036883548e-06, + "loss": 0.5477, + "step": 3957 + }, + { + "epoch": 0.3280563613758806, + "grad_norm": 0.2854398488998413, + "learning_rate": 3.3601326150020723e-06, + "loss": 0.5715, + "step": 3958 + }, + { + "epoch": 0.32813924575217573, + "grad_norm": 0.29279470443725586, + "learning_rate": 3.3597181931205968e-06, + "loss": 0.5481, + "step": 3959 + }, + { + "epoch": 0.3282221301284708, + "grad_norm": 0.26800262928009033, + "learning_rate": 3.359303771239122e-06, + "loss": 0.5039, + "step": 3960 + }, + { + "epoch": 0.32830501450476585, + "grad_norm": 0.27122947573661804, + "learning_rate": 3.3588893493576465e-06, + "loss": 0.5282, + "step": 3961 + }, + { + "epoch": 0.3283878988810609, + "grad_norm": 0.2863158881664276, + "learning_rate": 3.358474927476171e-06, + "loss": 0.6285, + "step": 3962 + }, + { + "epoch": 0.328470783257356, + "grad_norm": 0.2556924819946289, + "learning_rate": 3.3580605055946958e-06, + "loss": 0.4653, + "step": 3963 + }, + { + "epoch": 0.32855366763365107, + "grad_norm": 0.28364813327789307, + "learning_rate": 3.35764608371322e-06, + "loss": 0.5441, + "step": 3964 + }, + { + "epoch": 0.3286365520099461, + "grad_norm": 0.27955904603004456, + "learning_rate": 3.357231661831745e-06, + "loss": 0.5344, + "step": 3965 + }, + { + "epoch": 0.3287194363862412, + "grad_norm": 0.23925547301769257, + "learning_rate": 3.3568172399502695e-06, + "loss": 0.509, + "step": 3966 + }, + { + "epoch": 0.32880232076253624, + "grad_norm": 0.2759453356266022, + "learning_rate": 3.356402818068794e-06, + "loss": 0.5654, + "step": 3967 + }, + { + "epoch": 0.32888520513883135, + "grad_norm": 0.3094218671321869, + "learning_rate": 3.355988396187319e-06, + "loss": 0.4727, + "step": 3968 + }, + { + "epoch": 0.3289680895151264, + "grad_norm": 0.2781555950641632, + "learning_rate": 3.3555739743058436e-06, + "loss": 0.5202, + "step": 3969 + }, + { + "epoch": 0.32905097389142146, + "grad_norm": 0.2521318197250366, + "learning_rate": 3.3551595524243685e-06, + "loss": 0.51, + "step": 3970 + }, + { + "epoch": 0.3291338582677165, + "grad_norm": 0.25592079758644104, + "learning_rate": 3.354745130542893e-06, + "loss": 0.5102, + "step": 3971 + }, + { + "epoch": 0.32921674264401163, + "grad_norm": 0.25721824169158936, + "learning_rate": 3.3543307086614173e-06, + "loss": 0.531, + "step": 3972 + }, + { + "epoch": 0.3292996270203067, + "grad_norm": 0.2641347050666809, + "learning_rate": 3.353916286779942e-06, + "loss": 0.5361, + "step": 3973 + }, + { + "epoch": 0.32938251139660174, + "grad_norm": 0.26024481654167175, + "learning_rate": 3.353501864898467e-06, + "loss": 0.5082, + "step": 3974 + }, + { + "epoch": 0.3294653957728968, + "grad_norm": 0.2601454257965088, + "learning_rate": 3.3530874430169915e-06, + "loss": 0.5266, + "step": 3975 + }, + { + "epoch": 0.32954828014919185, + "grad_norm": 0.24107356369495392, + "learning_rate": 3.3526730211355163e-06, + "loss": 0.4993, + "step": 3976 + }, + { + "epoch": 0.32963116452548696, + "grad_norm": 0.257081001996994, + "learning_rate": 3.3522585992540408e-06, + "loss": 0.4872, + "step": 3977 + }, + { + "epoch": 0.329714048901782, + "grad_norm": 0.2440779060125351, + "learning_rate": 3.3518441773725656e-06, + "loss": 0.4988, + "step": 3978 + }, + { + "epoch": 0.3297969332780771, + "grad_norm": 0.2556155323982239, + "learning_rate": 3.35142975549109e-06, + "loss": 0.5098, + "step": 3979 + }, + { + "epoch": 0.32987981765437213, + "grad_norm": 0.2675783336162567, + "learning_rate": 3.3510153336096145e-06, + "loss": 0.5201, + "step": 3980 + }, + { + "epoch": 0.32996270203066724, + "grad_norm": 0.2702004313468933, + "learning_rate": 3.3506009117281398e-06, + "loss": 0.5645, + "step": 3981 + }, + { + "epoch": 0.3300455864069623, + "grad_norm": 0.26707544922828674, + "learning_rate": 3.350186489846664e-06, + "loss": 0.556, + "step": 3982 + }, + { + "epoch": 0.33012847078325736, + "grad_norm": 0.2596340775489807, + "learning_rate": 3.349772067965189e-06, + "loss": 0.5309, + "step": 3983 + }, + { + "epoch": 0.3302113551595524, + "grad_norm": 0.26139843463897705, + "learning_rate": 3.3493576460837135e-06, + "loss": 0.553, + "step": 3984 + }, + { + "epoch": 0.33029423953584747, + "grad_norm": 0.2622039318084717, + "learning_rate": 3.348943224202238e-06, + "loss": 0.5336, + "step": 3985 + }, + { + "epoch": 0.3303771239121426, + "grad_norm": 0.25834256410598755, + "learning_rate": 3.3485288023207628e-06, + "loss": 0.541, + "step": 3986 + }, + { + "epoch": 0.33046000828843763, + "grad_norm": 0.26075395941734314, + "learning_rate": 3.348114380439287e-06, + "loss": 0.5297, + "step": 3987 + }, + { + "epoch": 0.3305428926647327, + "grad_norm": 0.2692491114139557, + "learning_rate": 3.3476999585578125e-06, + "loss": 0.5833, + "step": 3988 + }, + { + "epoch": 0.33062577704102775, + "grad_norm": 0.2530382573604584, + "learning_rate": 3.347285536676337e-06, + "loss": 0.4805, + "step": 3989 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 0.2641041874885559, + "learning_rate": 3.3468711147948614e-06, + "loss": 0.5099, + "step": 3990 + }, + { + "epoch": 0.3307915457936179, + "grad_norm": 0.24169626832008362, + "learning_rate": 3.346456692913386e-06, + "loss": 0.5358, + "step": 3991 + }, + { + "epoch": 0.33087443016991297, + "grad_norm": 0.24419714510440826, + "learning_rate": 3.3460422710319106e-06, + "loss": 0.5178, + "step": 3992 + }, + { + "epoch": 0.330957314546208, + "grad_norm": 0.2676454186439514, + "learning_rate": 3.345627849150435e-06, + "loss": 0.5677, + "step": 3993 + }, + { + "epoch": 0.3310401989225031, + "grad_norm": 0.2506570518016815, + "learning_rate": 3.3452134272689603e-06, + "loss": 0.5147, + "step": 3994 + }, + { + "epoch": 0.3311230832987982, + "grad_norm": 0.277281790971756, + "learning_rate": 3.3447990053874848e-06, + "loss": 0.5094, + "step": 3995 + }, + { + "epoch": 0.33120596767509325, + "grad_norm": 0.25406795740127563, + "learning_rate": 3.3443845835060096e-06, + "loss": 0.5015, + "step": 3996 + }, + { + "epoch": 0.3312888520513883, + "grad_norm": 0.24733580648899078, + "learning_rate": 3.343970161624534e-06, + "loss": 0.5355, + "step": 3997 + }, + { + "epoch": 0.33137173642768336, + "grad_norm": 0.24224896728992462, + "learning_rate": 3.3435557397430585e-06, + "loss": 0.5047, + "step": 3998 + }, + { + "epoch": 0.3314546208039785, + "grad_norm": 0.26521989703178406, + "learning_rate": 3.3431413178615834e-06, + "loss": 0.5321, + "step": 3999 + }, + { + "epoch": 0.33153750518027353, + "grad_norm": 0.2778823971748352, + "learning_rate": 3.3427268959801078e-06, + "loss": 0.5495, + "step": 4000 + }, + { + "epoch": 0.3316203895565686, + "grad_norm": 0.2465781420469284, + "learning_rate": 3.342312474098633e-06, + "loss": 0.491, + "step": 4001 + }, + { + "epoch": 0.33170327393286364, + "grad_norm": 0.26754045486450195, + "learning_rate": 3.3418980522171575e-06, + "loss": 0.5071, + "step": 4002 + }, + { + "epoch": 0.3317861583091587, + "grad_norm": 0.29535096883773804, + "learning_rate": 3.341483630335682e-06, + "loss": 0.5696, + "step": 4003 + }, + { + "epoch": 0.3318690426854538, + "grad_norm": 0.2791159749031067, + "learning_rate": 3.3410692084542068e-06, + "loss": 0.5789, + "step": 4004 + }, + { + "epoch": 0.33195192706174886, + "grad_norm": 0.25642818212509155, + "learning_rate": 3.3406547865727312e-06, + "loss": 0.4798, + "step": 4005 + }, + { + "epoch": 0.3320348114380439, + "grad_norm": 0.25079965591430664, + "learning_rate": 3.340240364691256e-06, + "loss": 0.489, + "step": 4006 + }, + { + "epoch": 0.332117695814339, + "grad_norm": 0.28979721665382385, + "learning_rate": 3.3398259428097805e-06, + "loss": 0.5374, + "step": 4007 + }, + { + "epoch": 0.3322005801906341, + "grad_norm": 0.25559523701667786, + "learning_rate": 3.339411520928305e-06, + "loss": 0.5499, + "step": 4008 + }, + { + "epoch": 0.33228346456692914, + "grad_norm": 0.2373734414577484, + "learning_rate": 3.3389970990468302e-06, + "loss": 0.5319, + "step": 4009 + }, + { + "epoch": 0.3323663489432242, + "grad_norm": 0.24086759984493256, + "learning_rate": 3.3385826771653546e-06, + "loss": 0.5355, + "step": 4010 + }, + { + "epoch": 0.33244923331951926, + "grad_norm": 0.2774454951286316, + "learning_rate": 3.338168255283879e-06, + "loss": 0.549, + "step": 4011 + }, + { + "epoch": 0.33253211769581437, + "grad_norm": 0.25556716322898865, + "learning_rate": 3.337753833402404e-06, + "loss": 0.5489, + "step": 4012 + }, + { + "epoch": 0.3326150020721094, + "grad_norm": 0.271676629781723, + "learning_rate": 3.3373394115209284e-06, + "loss": 0.4649, + "step": 4013 + }, + { + "epoch": 0.3326978864484045, + "grad_norm": 0.2722279131412506, + "learning_rate": 3.3369249896394536e-06, + "loss": 0.5612, + "step": 4014 + }, + { + "epoch": 0.33278077082469953, + "grad_norm": 0.28117114305496216, + "learning_rate": 3.336510567757978e-06, + "loss": 0.4963, + "step": 4015 + }, + { + "epoch": 0.3328636552009946, + "grad_norm": 0.2761796712875366, + "learning_rate": 3.3360961458765025e-06, + "loss": 0.5884, + "step": 4016 + }, + { + "epoch": 0.3329465395772897, + "grad_norm": 0.25646257400512695, + "learning_rate": 3.3356817239950274e-06, + "loss": 0.5425, + "step": 4017 + }, + { + "epoch": 0.33302942395358476, + "grad_norm": 0.2352033108472824, + "learning_rate": 3.335267302113552e-06, + "loss": 0.5265, + "step": 4018 + }, + { + "epoch": 0.3331123083298798, + "grad_norm": 0.2802383303642273, + "learning_rate": 3.3348528802320767e-06, + "loss": 0.5894, + "step": 4019 + }, + { + "epoch": 0.33319519270617487, + "grad_norm": 0.2683698534965515, + "learning_rate": 3.334438458350601e-06, + "loss": 0.5435, + "step": 4020 + }, + { + "epoch": 0.33327807708247, + "grad_norm": 0.2648376524448395, + "learning_rate": 3.3340240364691255e-06, + "loss": 0.5149, + "step": 4021 + }, + { + "epoch": 0.33336096145876504, + "grad_norm": 0.2793973684310913, + "learning_rate": 3.333609614587651e-06, + "loss": 0.5773, + "step": 4022 + }, + { + "epoch": 0.3334438458350601, + "grad_norm": 0.24526818096637726, + "learning_rate": 3.3331951927061752e-06, + "loss": 0.5, + "step": 4023 + }, + { + "checkpoint_runtime": 538.7698 + }, + { + "epoch": 0.33352673021135515, + "grad_norm": 0.2668141722679138, + "learning_rate": 3.3327807708247e-06, + "loss": 0.5612, + "step": 4024 + }, + { + "epoch": 0.3336096145876502, + "grad_norm": 0.2487337291240692, + "learning_rate": 3.3323663489432245e-06, + "loss": 0.5075, + "step": 4025 + }, + { + "epoch": 0.3336924989639453, + "grad_norm": 0.2717536687850952, + "learning_rate": 3.331951927061749e-06, + "loss": 0.5673, + "step": 4026 + }, + { + "epoch": 0.3337753833402404, + "grad_norm": 0.26723214983940125, + "learning_rate": 3.331537505180274e-06, + "loss": 0.538, + "step": 4027 + }, + { + "epoch": 0.33385826771653543, + "grad_norm": 0.27092495560646057, + "learning_rate": 3.3311230832987982e-06, + "loss": 0.5146, + "step": 4028 + }, + { + "epoch": 0.3339411520928305, + "grad_norm": 0.28736603260040283, + "learning_rate": 3.3307086614173227e-06, + "loss": 0.5799, + "step": 4029 + }, + { + "epoch": 0.3340240364691256, + "grad_norm": 0.25537922978401184, + "learning_rate": 3.330294239535848e-06, + "loss": 0.5251, + "step": 4030 + }, + { + "epoch": 0.33410692084542065, + "grad_norm": 0.2551080584526062, + "learning_rate": 3.3298798176543724e-06, + "loss": 0.5155, + "step": 4031 + }, + { + "epoch": 0.3341898052217157, + "grad_norm": 0.2356039434671402, + "learning_rate": 3.3294653957728972e-06, + "loss": 0.4954, + "step": 4032 + }, + { + "epoch": 0.33427268959801076, + "grad_norm": 0.2607979476451874, + "learning_rate": 3.3290509738914217e-06, + "loss": 0.5241, + "step": 4033 + }, + { + "epoch": 0.3343555739743058, + "grad_norm": 0.25961121916770935, + "learning_rate": 3.328636552009946e-06, + "loss": 0.5378, + "step": 4034 + }, + { + "epoch": 0.33443845835060093, + "grad_norm": 0.3402022421360016, + "learning_rate": 3.3282221301284714e-06, + "loss": 0.5756, + "step": 4035 + }, + { + "epoch": 0.334521342726896, + "grad_norm": 0.26993080973625183, + "learning_rate": 3.327807708246996e-06, + "loss": 0.514, + "step": 4036 + }, + { + "epoch": 0.33460422710319104, + "grad_norm": 0.2483692318201065, + "learning_rate": 3.3273932863655207e-06, + "loss": 0.5273, + "step": 4037 + }, + { + "epoch": 0.3346871114794861, + "grad_norm": 0.26246485114097595, + "learning_rate": 3.326978864484045e-06, + "loss": 0.5423, + "step": 4038 + }, + { + "epoch": 0.3347699958557812, + "grad_norm": 0.25663748383522034, + "learning_rate": 3.3265644426025695e-06, + "loss": 0.5331, + "step": 4039 + }, + { + "epoch": 0.33485288023207627, + "grad_norm": 0.3026176691055298, + "learning_rate": 3.3261500207210944e-06, + "loss": 0.5259, + "step": 4040 + }, + { + "epoch": 0.3349357646083713, + "grad_norm": 0.2558353543281555, + "learning_rate": 3.325735598839619e-06, + "loss": 0.5553, + "step": 4041 + }, + { + "epoch": 0.3350186489846664, + "grad_norm": 0.26651933789253235, + "learning_rate": 3.325321176958144e-06, + "loss": 0.5722, + "step": 4042 + }, + { + "epoch": 0.33510153336096143, + "grad_norm": 0.2434639185667038, + "learning_rate": 3.3249067550766685e-06, + "loss": 0.5063, + "step": 4043 + }, + { + "epoch": 0.33518441773725655, + "grad_norm": 0.24490799009799957, + "learning_rate": 3.324492333195193e-06, + "loss": 0.4806, + "step": 4044 + }, + { + "epoch": 0.3352673021135516, + "grad_norm": 0.27598029375076294, + "learning_rate": 3.324077911313718e-06, + "loss": 0.5538, + "step": 4045 + }, + { + "epoch": 0.33535018648984666, + "grad_norm": 0.23554319143295288, + "learning_rate": 3.3236634894322422e-06, + "loss": 0.4976, + "step": 4046 + }, + { + "epoch": 0.3354330708661417, + "grad_norm": 0.24374109506607056, + "learning_rate": 3.3232490675507667e-06, + "loss": 0.5031, + "step": 4047 + }, + { + "epoch": 0.3355159552424368, + "grad_norm": 0.27225494384765625, + "learning_rate": 3.3228346456692915e-06, + "loss": 0.472, + "step": 4048 + }, + { + "epoch": 0.3355988396187319, + "grad_norm": 0.2725955545902252, + "learning_rate": 3.322420223787816e-06, + "loss": 0.5247, + "step": 4049 + }, + { + "epoch": 0.33568172399502694, + "grad_norm": 0.2435474544763565, + "learning_rate": 3.3220058019063412e-06, + "loss": 0.514, + "step": 4050 + }, + { + "epoch": 0.335764608371322, + "grad_norm": 0.24674256145954132, + "learning_rate": 3.3215913800248657e-06, + "loss": 0.5282, + "step": 4051 + }, + { + "epoch": 0.33584749274761705, + "grad_norm": 0.25716790556907654, + "learning_rate": 3.32117695814339e-06, + "loss": 0.5, + "step": 4052 + }, + { + "epoch": 0.33593037712391216, + "grad_norm": 0.2528417110443115, + "learning_rate": 3.320762536261915e-06, + "loss": 0.5328, + "step": 4053 + }, + { + "epoch": 0.3360132615002072, + "grad_norm": 0.2549654543399811, + "learning_rate": 3.3203481143804394e-06, + "loss": 0.5189, + "step": 4054 + }, + { + "epoch": 0.3360961458765023, + "grad_norm": 0.24881134927272797, + "learning_rate": 3.3199336924989647e-06, + "loss": 0.5314, + "step": 4055 + }, + { + "epoch": 0.33617903025279733, + "grad_norm": 0.23497186601161957, + "learning_rate": 3.319519270617489e-06, + "loss": 0.4927, + "step": 4056 + }, + { + "epoch": 0.33626191462909244, + "grad_norm": 0.2676345109939575, + "learning_rate": 3.3191048487360135e-06, + "loss": 0.5584, + "step": 4057 + }, + { + "epoch": 0.3363447990053875, + "grad_norm": 0.2486155480146408, + "learning_rate": 3.3186904268545384e-06, + "loss": 0.5178, + "step": 4058 + }, + { + "epoch": 0.33642768338168255, + "grad_norm": 0.259676992893219, + "learning_rate": 3.318276004973063e-06, + "loss": 0.5312, + "step": 4059 + }, + { + "epoch": 0.3365105677579776, + "grad_norm": 0.26733434200286865, + "learning_rate": 3.3178615830915872e-06, + "loss": 0.5071, + "step": 4060 + }, + { + "epoch": 0.33659345213427266, + "grad_norm": 0.274433434009552, + "learning_rate": 3.317447161210112e-06, + "loss": 0.5478, + "step": 4061 + }, + { + "epoch": 0.3366763365105678, + "grad_norm": 0.24821020662784576, + "learning_rate": 3.3170327393286365e-06, + "loss": 0.5087, + "step": 4062 + }, + { + "epoch": 0.33675922088686283, + "grad_norm": 0.265392541885376, + "learning_rate": 3.316618317447162e-06, + "loss": 0.5118, + "step": 4063 + }, + { + "epoch": 0.3368421052631579, + "grad_norm": 0.2525846064090729, + "learning_rate": 3.3162038955656862e-06, + "loss": 0.5268, + "step": 4064 + }, + { + "epoch": 0.33692498963945294, + "grad_norm": 0.2484198957681656, + "learning_rate": 3.3157894736842107e-06, + "loss": 0.4876, + "step": 4065 + }, + { + "epoch": 0.33700787401574805, + "grad_norm": 0.2879981994628906, + "learning_rate": 3.3153750518027355e-06, + "loss": 0.5436, + "step": 4066 + }, + { + "epoch": 0.3370907583920431, + "grad_norm": 0.25510576367378235, + "learning_rate": 3.31496062992126e-06, + "loss": 0.5358, + "step": 4067 + }, + { + "epoch": 0.33717364276833817, + "grad_norm": 0.2656743824481964, + "learning_rate": 3.314546208039785e-06, + "loss": 0.4814, + "step": 4068 + }, + { + "epoch": 0.3372565271446332, + "grad_norm": 0.25031954050064087, + "learning_rate": 3.3141317861583093e-06, + "loss": 0.5048, + "step": 4069 + }, + { + "epoch": 0.3373394115209283, + "grad_norm": 0.23518343269824982, + "learning_rate": 3.3137173642768337e-06, + "loss": 0.5087, + "step": 4070 + }, + { + "epoch": 0.3374222958972234, + "grad_norm": 0.26440638303756714, + "learning_rate": 3.313302942395359e-06, + "loss": 0.5376, + "step": 4071 + }, + { + "epoch": 0.33750518027351845, + "grad_norm": 0.30853402614593506, + "learning_rate": 3.3128885205138834e-06, + "loss": 0.5779, + "step": 4072 + }, + { + "epoch": 0.3375880646498135, + "grad_norm": 0.25536736845970154, + "learning_rate": 3.3124740986324082e-06, + "loss": 0.5029, + "step": 4073 + }, + { + "epoch": 0.33767094902610856, + "grad_norm": 0.26302874088287354, + "learning_rate": 3.3120596767509327e-06, + "loss": 0.5292, + "step": 4074 + }, + { + "epoch": 0.33775383340240367, + "grad_norm": 0.25434431433677673, + "learning_rate": 3.311645254869457e-06, + "loss": 0.5363, + "step": 4075 + }, + { + "epoch": 0.3378367177786987, + "grad_norm": 0.27191945910453796, + "learning_rate": 3.3112308329879824e-06, + "loss": 0.5548, + "step": 4076 + }, + { + "epoch": 0.3379196021549938, + "grad_norm": 0.2887403964996338, + "learning_rate": 3.310816411106507e-06, + "loss": 0.5385, + "step": 4077 + }, + { + "epoch": 0.33800248653128884, + "grad_norm": 0.24455378949642181, + "learning_rate": 3.3104019892250313e-06, + "loss": 0.4819, + "step": 4078 + }, + { + "epoch": 0.3380853709075839, + "grad_norm": 0.27169838547706604, + "learning_rate": 3.309987567343556e-06, + "loss": 0.5348, + "step": 4079 + }, + { + "epoch": 0.338168255283879, + "grad_norm": 0.23806306719779968, + "learning_rate": 3.3095731454620805e-06, + "loss": 0.504, + "step": 4080 + }, + { + "epoch": 0.33825113966017406, + "grad_norm": 0.24936753511428833, + "learning_rate": 3.3091587235806054e-06, + "loss": 0.4781, + "step": 4081 + }, + { + "epoch": 0.3383340240364691, + "grad_norm": 0.2510077953338623, + "learning_rate": 3.30874430169913e-06, + "loss": 0.509, + "step": 4082 + }, + { + "epoch": 0.3384169084127642, + "grad_norm": 0.2719770669937134, + "learning_rate": 3.3083298798176543e-06, + "loss": 0.5586, + "step": 4083 + }, + { + "epoch": 0.3384997927890593, + "grad_norm": 0.25622960925102234, + "learning_rate": 3.3079154579361795e-06, + "loss": 0.5437, + "step": 4084 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 0.2711159884929657, + "learning_rate": 3.307501036054704e-06, + "loss": 0.5124, + "step": 4085 + }, + { + "epoch": 0.3386655615416494, + "grad_norm": 0.263875275850296, + "learning_rate": 3.307086614173229e-06, + "loss": 0.5197, + "step": 4086 + }, + { + "epoch": 0.33874844591794445, + "grad_norm": 0.24891901016235352, + "learning_rate": 3.3066721922917533e-06, + "loss": 0.527, + "step": 4087 + }, + { + "epoch": 0.33883133029423956, + "grad_norm": 0.27670106291770935, + "learning_rate": 3.3062577704102777e-06, + "loss": 0.5655, + "step": 4088 + }, + { + "epoch": 0.3389142146705346, + "grad_norm": 0.2622976303100586, + "learning_rate": 3.3058433485288025e-06, + "loss": 0.5261, + "step": 4089 + }, + { + "epoch": 0.3389970990468297, + "grad_norm": 0.2977895736694336, + "learning_rate": 3.305428926647327e-06, + "loss": 0.5555, + "step": 4090 + }, + { + "epoch": 0.33907998342312473, + "grad_norm": 0.2707483172416687, + "learning_rate": 3.3050145047658523e-06, + "loss": 0.5414, + "step": 4091 + }, + { + "epoch": 0.3391628677994198, + "grad_norm": 0.2678997814655304, + "learning_rate": 3.3046000828843767e-06, + "loss": 0.5511, + "step": 4092 + }, + { + "epoch": 0.3392457521757149, + "grad_norm": 0.25570154190063477, + "learning_rate": 3.304185661002901e-06, + "loss": 0.5612, + "step": 4093 + }, + { + "epoch": 0.33932863655200995, + "grad_norm": 0.2671893835067749, + "learning_rate": 3.303771239121426e-06, + "loss": 0.5135, + "step": 4094 + }, + { + "epoch": 0.339411520928305, + "grad_norm": 0.26289600133895874, + "learning_rate": 3.3033568172399504e-06, + "loss": 0.558, + "step": 4095 + }, + { + "epoch": 0.33949440530460007, + "grad_norm": 0.26622188091278076, + "learning_rate": 3.302942395358475e-06, + "loss": 0.5408, + "step": 4096 + }, + { + "epoch": 0.3395772896808952, + "grad_norm": 0.26009681820869446, + "learning_rate": 3.302527973477e-06, + "loss": 0.551, + "step": 4097 + }, + { + "epoch": 0.33966017405719023, + "grad_norm": 0.24222533404827118, + "learning_rate": 3.3021135515955245e-06, + "loss": 0.5333, + "step": 4098 + }, + { + "epoch": 0.3397430584334853, + "grad_norm": 0.27473828196525574, + "learning_rate": 3.3016991297140494e-06, + "loss": 0.5087, + "step": 4099 + }, + { + "epoch": 0.33982594280978035, + "grad_norm": 0.264627605676651, + "learning_rate": 3.301284707832574e-06, + "loss": 0.541, + "step": 4100 + }, + { + "epoch": 0.3399088271860754, + "grad_norm": 0.2682136595249176, + "learning_rate": 3.3008702859510983e-06, + "loss": 0.5554, + "step": 4101 + }, + { + "epoch": 0.3399917115623705, + "grad_norm": 0.2586138844490051, + "learning_rate": 3.300455864069623e-06, + "loss": 0.5922, + "step": 4102 + }, + { + "epoch": 0.34007459593866557, + "grad_norm": 0.2548489570617676, + "learning_rate": 3.3000414421881476e-06, + "loss": 0.5125, + "step": 4103 + }, + { + "epoch": 0.3401574803149606, + "grad_norm": 0.2748565077781677, + "learning_rate": 3.299627020306673e-06, + "loss": 0.5269, + "step": 4104 + }, + { + "epoch": 0.3402403646912557, + "grad_norm": 0.264428973197937, + "learning_rate": 3.2992125984251973e-06, + "loss": 0.5641, + "step": 4105 + }, + { + "epoch": 0.3403232490675508, + "grad_norm": 0.271923691034317, + "learning_rate": 3.2987981765437217e-06, + "loss": 0.51, + "step": 4106 + }, + { + "epoch": 0.34040613344384585, + "grad_norm": 0.26217538118362427, + "learning_rate": 3.2983837546622466e-06, + "loss": 0.543, + "step": 4107 + }, + { + "epoch": 0.3404890178201409, + "grad_norm": 0.2696745693683624, + "learning_rate": 3.297969332780771e-06, + "loss": 0.5112, + "step": 4108 + }, + { + "epoch": 0.34057190219643596, + "grad_norm": 0.2626086473464966, + "learning_rate": 3.297554910899296e-06, + "loss": 0.5416, + "step": 4109 + }, + { + "epoch": 0.340654786572731, + "grad_norm": 0.2723824381828308, + "learning_rate": 3.2971404890178203e-06, + "loss": 0.5038, + "step": 4110 + }, + { + "epoch": 0.34073767094902613, + "grad_norm": 0.26529809832572937, + "learning_rate": 3.2967260671363447e-06, + "loss": 0.5533, + "step": 4111 + }, + { + "epoch": 0.3408205553253212, + "grad_norm": 0.2667498290538788, + "learning_rate": 3.29631164525487e-06, + "loss": 0.5311, + "step": 4112 + }, + { + "epoch": 0.34090343970161624, + "grad_norm": 0.2509499192237854, + "learning_rate": 3.2958972233733944e-06, + "loss": 0.5389, + "step": 4113 + }, + { + "epoch": 0.3409863240779113, + "grad_norm": 0.24481342732906342, + "learning_rate": 3.295482801491919e-06, + "loss": 0.5113, + "step": 4114 + }, + { + "epoch": 0.3410692084542064, + "grad_norm": 0.26881271600723267, + "learning_rate": 3.2950683796104437e-06, + "loss": 0.5086, + "step": 4115 + }, + { + "epoch": 0.34115209283050146, + "grad_norm": 0.26505061984062195, + "learning_rate": 3.294653957728968e-06, + "loss": 0.5801, + "step": 4116 + }, + { + "epoch": 0.3412349772067965, + "grad_norm": 0.28537866473197937, + "learning_rate": 3.2942395358474934e-06, + "loss": 0.5736, + "step": 4117 + }, + { + "epoch": 0.3413178615830916, + "grad_norm": 0.27048459649086, + "learning_rate": 3.293825113966018e-06, + "loss": 0.5481, + "step": 4118 + }, + { + "epoch": 0.34140074595938663, + "grad_norm": 0.26430055499076843, + "learning_rate": 3.2934106920845423e-06, + "loss": 0.5845, + "step": 4119 + }, + { + "epoch": 0.34148363033568174, + "grad_norm": 0.24778468906879425, + "learning_rate": 3.292996270203067e-06, + "loss": 0.5468, + "step": 4120 + }, + { + "epoch": 0.3415665147119768, + "grad_norm": 0.265174001455307, + "learning_rate": 3.2925818483215916e-06, + "loss": 0.4959, + "step": 4121 + }, + { + "epoch": 0.34164939908827185, + "grad_norm": 0.2577514052391052, + "learning_rate": 3.2921674264401164e-06, + "loss": 0.5578, + "step": 4122 + }, + { + "epoch": 0.3417322834645669, + "grad_norm": 0.25731542706489563, + "learning_rate": 3.291753004558641e-06, + "loss": 0.5562, + "step": 4123 + }, + { + "epoch": 0.341815167840862, + "grad_norm": 0.28383052349090576, + "learning_rate": 3.2913385826771653e-06, + "loss": 0.5256, + "step": 4124 + }, + { + "epoch": 0.3418980522171571, + "grad_norm": 0.24941116571426392, + "learning_rate": 3.2909241607956906e-06, + "loss": 0.5041, + "step": 4125 + }, + { + "epoch": 0.34198093659345213, + "grad_norm": 0.2680260241031647, + "learning_rate": 3.290509738914215e-06, + "loss": 0.4949, + "step": 4126 + }, + { + "epoch": 0.3420638209697472, + "grad_norm": 0.24082227051258087, + "learning_rate": 3.2900953170327394e-06, + "loss": 0.519, + "step": 4127 + }, + { + "epoch": 0.34214670534604225, + "grad_norm": 0.2625153958797455, + "learning_rate": 3.2896808951512643e-06, + "loss": 0.5414, + "step": 4128 + }, + { + "epoch": 0.34222958972233736, + "grad_norm": 0.2547418773174286, + "learning_rate": 3.2892664732697887e-06, + "loss": 0.5231, + "step": 4129 + }, + { + "epoch": 0.3423124740986324, + "grad_norm": 0.24584950506687164, + "learning_rate": 3.2888520513883136e-06, + "loss": 0.5211, + "step": 4130 + }, + { + "epoch": 0.34239535847492747, + "grad_norm": 0.26223888993263245, + "learning_rate": 3.288437629506838e-06, + "loss": 0.521, + "step": 4131 + }, + { + "epoch": 0.3424782428512225, + "grad_norm": 0.23923470079898834, + "learning_rate": 3.288023207625363e-06, + "loss": 0.5166, + "step": 4132 + }, + { + "epoch": 0.34256112722751764, + "grad_norm": 0.2512717545032501, + "learning_rate": 3.2876087857438877e-06, + "loss": 0.4932, + "step": 4133 + }, + { + "epoch": 0.3426440116038127, + "grad_norm": 0.24947261810302734, + "learning_rate": 3.287194363862412e-06, + "loss": 0.5342, + "step": 4134 + }, + { + "epoch": 0.34272689598010775, + "grad_norm": 0.2803371548652649, + "learning_rate": 3.286779941980937e-06, + "loss": 0.5611, + "step": 4135 + }, + { + "epoch": 0.3428097803564028, + "grad_norm": 0.2616669237613678, + "learning_rate": 3.2863655200994614e-06, + "loss": 0.5266, + "step": 4136 + }, + { + "epoch": 0.34289266473269786, + "grad_norm": 0.2741563022136688, + "learning_rate": 3.285951098217986e-06, + "loss": 0.5273, + "step": 4137 + }, + { + "epoch": 0.34297554910899297, + "grad_norm": 0.25301867723464966, + "learning_rate": 3.285536676336511e-06, + "loss": 0.5164, + "step": 4138 + }, + { + "epoch": 0.34305843348528803, + "grad_norm": 0.22898483276367188, + "learning_rate": 3.2851222544550356e-06, + "loss": 0.5097, + "step": 4139 + }, + { + "epoch": 0.3431413178615831, + "grad_norm": 0.2623250186443329, + "learning_rate": 3.2847078325735604e-06, + "loss": 0.4943, + "step": 4140 + }, + { + "epoch": 0.34322420223787814, + "grad_norm": 0.2587478756904602, + "learning_rate": 3.284293410692085e-06, + "loss": 0.5651, + "step": 4141 + }, + { + "epoch": 0.34330708661417325, + "grad_norm": 0.26216450333595276, + "learning_rate": 3.2838789888106093e-06, + "loss": 0.5384, + "step": 4142 + }, + { + "epoch": 0.3433899709904683, + "grad_norm": 0.2579156458377838, + "learning_rate": 3.283464566929134e-06, + "loss": 0.5294, + "step": 4143 + }, + { + "epoch": 0.34347285536676336, + "grad_norm": 0.24960407614707947, + "learning_rate": 3.2830501450476586e-06, + "loss": 0.537, + "step": 4144 + }, + { + "epoch": 0.3435557397430584, + "grad_norm": 0.25327667593955994, + "learning_rate": 3.282635723166183e-06, + "loss": 0.5186, + "step": 4145 + }, + { + "epoch": 0.3436386241193535, + "grad_norm": 0.2903306186199188, + "learning_rate": 3.2822213012847083e-06, + "loss": 0.5388, + "step": 4146 + }, + { + "epoch": 0.3437215084956486, + "grad_norm": 0.23711572587490082, + "learning_rate": 3.2818068794032327e-06, + "loss": 0.4721, + "step": 4147 + }, + { + "epoch": 0.34380439287194364, + "grad_norm": 0.26097622513771057, + "learning_rate": 3.2813924575217576e-06, + "loss": 0.5462, + "step": 4148 + }, + { + "epoch": 0.3438872772482387, + "grad_norm": 0.2906772196292877, + "learning_rate": 3.280978035640282e-06, + "loss": 0.5757, + "step": 4149 + }, + { + "epoch": 0.34397016162453375, + "grad_norm": 0.25141069293022156, + "learning_rate": 3.2805636137588064e-06, + "loss": 0.5049, + "step": 4150 + }, + { + "epoch": 0.34405304600082887, + "grad_norm": 0.25670063495635986, + "learning_rate": 3.2801491918773313e-06, + "loss": 0.5103, + "step": 4151 + }, + { + "epoch": 0.3441359303771239, + "grad_norm": 0.25458574295043945, + "learning_rate": 3.279734769995856e-06, + "loss": 0.5297, + "step": 4152 + }, + { + "epoch": 0.344218814753419, + "grad_norm": 0.2660068869590759, + "learning_rate": 3.279320348114381e-06, + "loss": 0.5165, + "step": 4153 + }, + { + "epoch": 0.34430169912971403, + "grad_norm": 0.29124921560287476, + "learning_rate": 3.2789059262329054e-06, + "loss": 0.5356, + "step": 4154 + }, + { + "epoch": 0.3443845835060091, + "grad_norm": 0.24655048549175262, + "learning_rate": 3.27849150435143e-06, + "loss": 0.4842, + "step": 4155 + }, + { + "epoch": 0.3444674678823042, + "grad_norm": 0.28527387976646423, + "learning_rate": 3.2780770824699547e-06, + "loss": 0.5313, + "step": 4156 + }, + { + "epoch": 0.34455035225859926, + "grad_norm": 0.2522020936012268, + "learning_rate": 3.277662660588479e-06, + "loss": 0.5187, + "step": 4157 + }, + { + "epoch": 0.3446332366348943, + "grad_norm": 0.2680804133415222, + "learning_rate": 3.2772482387070044e-06, + "loss": 0.5292, + "step": 4158 + }, + { + "epoch": 0.34471612101118937, + "grad_norm": 0.2347985804080963, + "learning_rate": 3.276833816825529e-06, + "loss": 0.4907, + "step": 4159 + }, + { + "epoch": 0.3447990053874845, + "grad_norm": 0.2866113483905792, + "learning_rate": 3.2764193949440533e-06, + "loss": 0.5049, + "step": 4160 + }, + { + "epoch": 0.34488188976377954, + "grad_norm": 0.265575110912323, + "learning_rate": 3.276004973062578e-06, + "loss": 0.5831, + "step": 4161 + }, + { + "epoch": 0.3449647741400746, + "grad_norm": 0.27214372158050537, + "learning_rate": 3.2755905511811026e-06, + "loss": 0.5415, + "step": 4162 + }, + { + "epoch": 0.34504765851636965, + "grad_norm": 0.2508997321128845, + "learning_rate": 3.275176129299627e-06, + "loss": 0.5349, + "step": 4163 + }, + { + "epoch": 0.34513054289266476, + "grad_norm": 0.2529197335243225, + "learning_rate": 3.274761707418152e-06, + "loss": 0.5399, + "step": 4164 + }, + { + "epoch": 0.3452134272689598, + "grad_norm": 0.2592162489891052, + "learning_rate": 3.2743472855366763e-06, + "loss": 0.5436, + "step": 4165 + }, + { + "epoch": 0.34529631164525487, + "grad_norm": 0.27730968594551086, + "learning_rate": 3.2739328636552016e-06, + "loss": 0.509, + "step": 4166 + }, + { + "epoch": 0.34537919602154993, + "grad_norm": 0.25229740142822266, + "learning_rate": 3.273518441773726e-06, + "loss": 0.4995, + "step": 4167 + }, + { + "epoch": 0.345462080397845, + "grad_norm": 0.26819512248039246, + "learning_rate": 3.2731040198922504e-06, + "loss": 0.5865, + "step": 4168 + }, + { + "epoch": 0.3455449647741401, + "grad_norm": 0.3419029414653778, + "learning_rate": 3.2726895980107753e-06, + "loss": 0.5258, + "step": 4169 + }, + { + "epoch": 0.34562784915043515, + "grad_norm": 0.25353413820266724, + "learning_rate": 3.2722751761292997e-06, + "loss": 0.5578, + "step": 4170 + }, + { + "epoch": 0.3457107335267302, + "grad_norm": 0.2728200852870941, + "learning_rate": 3.2718607542478246e-06, + "loss": 0.5611, + "step": 4171 + }, + { + "epoch": 0.34579361790302526, + "grad_norm": 0.2679423689842224, + "learning_rate": 3.271446332366349e-06, + "loss": 0.5312, + "step": 4172 + }, + { + "epoch": 0.3458765022793204, + "grad_norm": 0.24347048997879028, + "learning_rate": 3.271031910484874e-06, + "loss": 0.5165, + "step": 4173 + }, + { + "epoch": 0.34595938665561543, + "grad_norm": 0.22066465020179749, + "learning_rate": 3.2706174886033987e-06, + "loss": 0.4655, + "step": 4174 + }, + { + "epoch": 0.3460422710319105, + "grad_norm": 0.26530876755714417, + "learning_rate": 3.270203066721923e-06, + "loss": 0.5339, + "step": 4175 + }, + { + "epoch": 0.34612515540820554, + "grad_norm": 0.25706735253334045, + "learning_rate": 3.269788644840448e-06, + "loss": 0.5298, + "step": 4176 + }, + { + "epoch": 0.3462080397845006, + "grad_norm": 0.23936043679714203, + "learning_rate": 3.2693742229589724e-06, + "loss": 0.4869, + "step": 4177 + }, + { + "epoch": 0.3462909241607957, + "grad_norm": 0.28513985872268677, + "learning_rate": 3.268959801077497e-06, + "loss": 0.5615, + "step": 4178 + }, + { + "epoch": 0.34637380853709077, + "grad_norm": 0.2620239853858948, + "learning_rate": 3.268545379196022e-06, + "loss": 0.5512, + "step": 4179 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 0.2629260718822479, + "learning_rate": 3.2681309573145466e-06, + "loss": 0.5925, + "step": 4180 + }, + { + "epoch": 0.3465395772896809, + "grad_norm": 0.25440090894699097, + "learning_rate": 3.267716535433071e-06, + "loss": 0.5363, + "step": 4181 + }, + { + "epoch": 0.346622461665976, + "grad_norm": 0.23632606863975525, + "learning_rate": 3.267302113551596e-06, + "loss": 0.4944, + "step": 4182 + }, + { + "epoch": 0.34670534604227105, + "grad_norm": 0.24755018949508667, + "learning_rate": 3.2668876916701203e-06, + "loss": 0.5194, + "step": 4183 + }, + { + "epoch": 0.3467882304185661, + "grad_norm": 0.26845303177833557, + "learning_rate": 3.266473269788645e-06, + "loss": 0.5274, + "step": 4184 + }, + { + "epoch": 0.34687111479486116, + "grad_norm": 0.26527467370033264, + "learning_rate": 3.2660588479071696e-06, + "loss": 0.5381, + "step": 4185 + }, + { + "epoch": 0.3469539991711562, + "grad_norm": 0.25664055347442627, + "learning_rate": 3.265644426025694e-06, + "loss": 0.573, + "step": 4186 + }, + { + "epoch": 0.3470368835474513, + "grad_norm": 0.25691190361976624, + "learning_rate": 3.2652300041442193e-06, + "loss": 0.4815, + "step": 4187 + }, + { + "epoch": 0.3471197679237464, + "grad_norm": 0.25303635001182556, + "learning_rate": 3.2648155822627437e-06, + "loss": 0.5241, + "step": 4188 + }, + { + "epoch": 0.34720265230004144, + "grad_norm": 0.26579341292381287, + "learning_rate": 3.2644011603812686e-06, + "loss": 0.5477, + "step": 4189 + }, + { + "epoch": 0.3472855366763365, + "grad_norm": 0.24903224408626556, + "learning_rate": 3.263986738499793e-06, + "loss": 0.5196, + "step": 4190 + }, + { + "epoch": 0.3473684210526316, + "grad_norm": 0.2693108022212982, + "learning_rate": 3.2635723166183175e-06, + "loss": 0.4743, + "step": 4191 + }, + { + "epoch": 0.34745130542892666, + "grad_norm": 0.2513848841190338, + "learning_rate": 3.2631578947368423e-06, + "loss": 0.5231, + "step": 4192 + }, + { + "epoch": 0.3475341898052217, + "grad_norm": 0.2854318916797638, + "learning_rate": 3.262743472855367e-06, + "loss": 0.5553, + "step": 4193 + }, + { + "epoch": 0.34761707418151677, + "grad_norm": 0.27616849541664124, + "learning_rate": 3.262329050973892e-06, + "loss": 0.562, + "step": 4194 + }, + { + "epoch": 0.34769995855781183, + "grad_norm": 0.2515066862106323, + "learning_rate": 3.2619146290924165e-06, + "loss": 0.4747, + "step": 4195 + }, + { + "epoch": 0.34778284293410694, + "grad_norm": 0.2581586241722107, + "learning_rate": 3.261500207210941e-06, + "loss": 0.532, + "step": 4196 + }, + { + "epoch": 0.347865727310402, + "grad_norm": 0.2778302729129791, + "learning_rate": 3.2610857853294657e-06, + "loss": 0.5092, + "step": 4197 + }, + { + "epoch": 0.34794861168669705, + "grad_norm": 0.291217565536499, + "learning_rate": 3.26067136344799e-06, + "loss": 0.501, + "step": 4198 + }, + { + "epoch": 0.3480314960629921, + "grad_norm": 0.2875407040119171, + "learning_rate": 3.2602569415665146e-06, + "loss": 0.5438, + "step": 4199 + }, + { + "epoch": 0.3481143804392872, + "grad_norm": 0.27407410740852356, + "learning_rate": 3.25984251968504e-06, + "loss": 0.5195, + "step": 4200 + }, + { + "epoch": 0.3481972648155823, + "grad_norm": 0.276778906583786, + "learning_rate": 3.2594280978035643e-06, + "loss": 0.4985, + "step": 4201 + }, + { + "epoch": 0.34828014919187733, + "grad_norm": 0.25169795751571655, + "learning_rate": 3.259013675922089e-06, + "loss": 0.5343, + "step": 4202 + }, + { + "epoch": 0.3483630335681724, + "grad_norm": 0.2519007921218872, + "learning_rate": 3.2585992540406136e-06, + "loss": 0.5052, + "step": 4203 + }, + { + "epoch": 0.34844591794446744, + "grad_norm": 0.2589815557003021, + "learning_rate": 3.258184832159138e-06, + "loss": 0.5452, + "step": 4204 + }, + { + "epoch": 0.34852880232076255, + "grad_norm": 0.27631309628486633, + "learning_rate": 3.257770410277663e-06, + "loss": 0.5574, + "step": 4205 + }, + { + "epoch": 0.3486116866970576, + "grad_norm": 0.26427435874938965, + "learning_rate": 3.2573559883961873e-06, + "loss": 0.5081, + "step": 4206 + }, + { + "epoch": 0.34869457107335267, + "grad_norm": 0.25901174545288086, + "learning_rate": 3.2569415665147126e-06, + "loss": 0.4724, + "step": 4207 + }, + { + "epoch": 0.3487774554496477, + "grad_norm": 0.24706539511680603, + "learning_rate": 3.256527144633237e-06, + "loss": 0.5302, + "step": 4208 + }, + { + "epoch": 0.34886033982594283, + "grad_norm": 0.27879229187965393, + "learning_rate": 3.2561127227517615e-06, + "loss": 0.5627, + "step": 4209 + }, + { + "epoch": 0.3489432242022379, + "grad_norm": 0.26848629117012024, + "learning_rate": 3.2556983008702863e-06, + "loss": 0.565, + "step": 4210 + }, + { + "epoch": 0.34902610857853295, + "grad_norm": 0.2708277404308319, + "learning_rate": 3.2552838789888108e-06, + "loss": 0.5198, + "step": 4211 + }, + { + "epoch": 0.349108992954828, + "grad_norm": 0.27425745129585266, + "learning_rate": 3.254869457107335e-06, + "loss": 0.5906, + "step": 4212 + }, + { + "epoch": 0.34919187733112306, + "grad_norm": 0.2520531117916107, + "learning_rate": 3.2544550352258605e-06, + "loss": 0.533, + "step": 4213 + }, + { + "epoch": 0.34927476170741817, + "grad_norm": 0.23160475492477417, + "learning_rate": 3.254040613344385e-06, + "loss": 0.467, + "step": 4214 + }, + { + "epoch": 0.3493576460837132, + "grad_norm": 0.2622203230857849, + "learning_rate": 3.2536261914629098e-06, + "loss": 0.5273, + "step": 4215 + }, + { + "epoch": 0.3494405304600083, + "grad_norm": 0.23665368556976318, + "learning_rate": 3.253211769581434e-06, + "loss": 0.5435, + "step": 4216 + }, + { + "epoch": 0.34952341483630334, + "grad_norm": 0.2697499692440033, + "learning_rate": 3.2527973476999586e-06, + "loss": 0.5597, + "step": 4217 + }, + { + "epoch": 0.34960629921259845, + "grad_norm": 0.26542139053344727, + "learning_rate": 3.2523829258184835e-06, + "loss": 0.5202, + "step": 4218 + }, + { + "epoch": 0.3496891835888935, + "grad_norm": 0.26063573360443115, + "learning_rate": 3.251968503937008e-06, + "loss": 0.5107, + "step": 4219 + }, + { + "epoch": 0.34977206796518856, + "grad_norm": 0.25817951560020447, + "learning_rate": 3.251554082055533e-06, + "loss": 0.5352, + "step": 4220 + }, + { + "epoch": 0.3498549523414836, + "grad_norm": 0.2996380031108856, + "learning_rate": 3.2511396601740576e-06, + "loss": 0.4968, + "step": 4221 + }, + { + "epoch": 0.34993783671777867, + "grad_norm": 0.24038682878017426, + "learning_rate": 3.250725238292582e-06, + "loss": 0.5075, + "step": 4222 + }, + { + "epoch": 0.3500207210940738, + "grad_norm": 0.26452091336250305, + "learning_rate": 3.250310816411107e-06, + "loss": 0.5457, + "step": 4223 + }, + { + "epoch": 0.35010360547036884, + "grad_norm": 0.23775547742843628, + "learning_rate": 3.2498963945296313e-06, + "loss": 0.4952, + "step": 4224 + }, + { + "epoch": 0.3501864898466639, + "grad_norm": 0.2733628451824188, + "learning_rate": 3.249481972648156e-06, + "loss": 0.5109, + "step": 4225 + }, + { + "epoch": 0.35026937422295895, + "grad_norm": 0.25346580147743225, + "learning_rate": 3.2490675507666806e-06, + "loss": 0.4831, + "step": 4226 + }, + { + "epoch": 0.35035225859925406, + "grad_norm": 0.27653077244758606, + "learning_rate": 3.248653128885205e-06, + "loss": 0.5375, + "step": 4227 + }, + { + "epoch": 0.3504351429755491, + "grad_norm": 0.2494218796491623, + "learning_rate": 3.2482387070037303e-06, + "loss": 0.5131, + "step": 4228 + }, + { + "epoch": 0.3505180273518442, + "grad_norm": 0.2645801305770874, + "learning_rate": 3.2478242851222548e-06, + "loss": 0.5118, + "step": 4229 + }, + { + "epoch": 0.35060091172813923, + "grad_norm": 0.25578972697257996, + "learning_rate": 3.247409863240779e-06, + "loss": 0.5037, + "step": 4230 + }, + { + "epoch": 0.3506837961044343, + "grad_norm": 0.26978278160095215, + "learning_rate": 3.246995441359304e-06, + "loss": 0.5387, + "step": 4231 + }, + { + "epoch": 0.3507666804807294, + "grad_norm": 0.26576799154281616, + "learning_rate": 3.2465810194778285e-06, + "loss": 0.5032, + "step": 4232 + }, + { + "epoch": 0.35084956485702445, + "grad_norm": 0.2746669352054596, + "learning_rate": 3.2461665975963538e-06, + "loss": 0.5007, + "step": 4233 + }, + { + "epoch": 0.3509324492333195, + "grad_norm": 0.25751984119415283, + "learning_rate": 3.245752175714878e-06, + "loss": 0.4843, + "step": 4234 + }, + { + "epoch": 0.35101533360961457, + "grad_norm": 0.28138500452041626, + "learning_rate": 3.2453377538334026e-06, + "loss": 0.5536, + "step": 4235 + }, + { + "epoch": 0.3510982179859097, + "grad_norm": 0.2550415098667145, + "learning_rate": 3.2449233319519275e-06, + "loss": 0.5554, + "step": 4236 + }, + { + "epoch": 0.35118110236220473, + "grad_norm": 0.2699739933013916, + "learning_rate": 3.244508910070452e-06, + "loss": 0.5273, + "step": 4237 + }, + { + "epoch": 0.3512639867384998, + "grad_norm": 0.2963600754737854, + "learning_rate": 3.2440944881889768e-06, + "loss": 0.5521, + "step": 4238 + }, + { + "epoch": 0.35134687111479485, + "grad_norm": 0.262820303440094, + "learning_rate": 3.243680066307501e-06, + "loss": 0.5276, + "step": 4239 + }, + { + "epoch": 0.35142975549108996, + "grad_norm": 0.24638628959655762, + "learning_rate": 3.2432656444260256e-06, + "loss": 0.528, + "step": 4240 + }, + { + "epoch": 0.351512639867385, + "grad_norm": 0.2762331962585449, + "learning_rate": 3.242851222544551e-06, + "loss": 0.5455, + "step": 4241 + }, + { + "epoch": 0.35159552424368007, + "grad_norm": 0.2555232048034668, + "learning_rate": 3.2424368006630753e-06, + "loss": 0.5435, + "step": 4242 + }, + { + "epoch": 0.3516784086199751, + "grad_norm": 0.25662919878959656, + "learning_rate": 3.2420223787816e-06, + "loss": 0.5563, + "step": 4243 + }, + { + "epoch": 0.3517612929962702, + "grad_norm": 0.26137852668762207, + "learning_rate": 3.2416079569001246e-06, + "loss": 0.5622, + "step": 4244 + }, + { + "epoch": 0.3518441773725653, + "grad_norm": 0.254558801651001, + "learning_rate": 3.241193535018649e-06, + "loss": 0.493, + "step": 4245 + }, + { + "epoch": 0.35192706174886035, + "grad_norm": 0.25732406973838806, + "learning_rate": 3.240779113137174e-06, + "loss": 0.5273, + "step": 4246 + }, + { + "epoch": 0.3520099461251554, + "grad_norm": 0.26215267181396484, + "learning_rate": 3.2403646912556983e-06, + "loss": 0.5746, + "step": 4247 + }, + { + "epoch": 0.35209283050145046, + "grad_norm": 0.2914908826351166, + "learning_rate": 3.2399502693742228e-06, + "loss": 0.5731, + "step": 4248 + }, + { + "epoch": 0.35217571487774557, + "grad_norm": 0.26020458340644836, + "learning_rate": 3.239535847492748e-06, + "loss": 0.5569, + "step": 4249 + }, + { + "epoch": 0.3522585992540406, + "grad_norm": 0.27505895495414734, + "learning_rate": 3.2391214256112725e-06, + "loss": 0.5342, + "step": 4250 + }, + { + "epoch": 0.3523414836303357, + "grad_norm": 0.23428864777088165, + "learning_rate": 3.2387070037297973e-06, + "loss": 0.494, + "step": 4251 + }, + { + "epoch": 0.35242436800663074, + "grad_norm": 0.24227143824100494, + "learning_rate": 3.2382925818483218e-06, + "loss": 0.4657, + "step": 4252 + }, + { + "epoch": 0.3525072523829258, + "grad_norm": 0.24651993811130524, + "learning_rate": 3.237878159966846e-06, + "loss": 0.514, + "step": 4253 + }, + { + "epoch": 0.3525901367592209, + "grad_norm": 0.2669703960418701, + "learning_rate": 3.2374637380853715e-06, + "loss": 0.5703, + "step": 4254 + }, + { + "epoch": 0.35267302113551596, + "grad_norm": 0.2382771223783493, + "learning_rate": 3.237049316203896e-06, + "loss": 0.5169, + "step": 4255 + }, + { + "epoch": 0.352755905511811, + "grad_norm": 0.2544070780277252, + "learning_rate": 3.2366348943224208e-06, + "loss": 0.5454, + "step": 4256 + }, + { + "epoch": 0.3528387898881061, + "grad_norm": 0.2632478177547455, + "learning_rate": 3.236220472440945e-06, + "loss": 0.5654, + "step": 4257 + }, + { + "epoch": 0.3529216742644012, + "grad_norm": 0.25789615511894226, + "learning_rate": 3.2358060505594696e-06, + "loss": 0.543, + "step": 4258 + }, + { + "epoch": 0.35300455864069624, + "grad_norm": 0.28707438707351685, + "learning_rate": 3.2353916286779945e-06, + "loss": 0.5056, + "step": 4259 + }, + { + "epoch": 0.3530874430169913, + "grad_norm": 0.25860872864723206, + "learning_rate": 3.234977206796519e-06, + "loss": 0.512, + "step": 4260 + }, + { + "epoch": 0.35317032739328635, + "grad_norm": 0.2433716058731079, + "learning_rate": 3.234562784915044e-06, + "loss": 0.512, + "step": 4261 + }, + { + "epoch": 0.3532532117695814, + "grad_norm": 0.26075640320777893, + "learning_rate": 3.2341483630335686e-06, + "loss": 0.5156, + "step": 4262 + }, + { + "epoch": 0.3533360961458765, + "grad_norm": 0.2601933777332306, + "learning_rate": 3.233733941152093e-06, + "loss": 0.5026, + "step": 4263 + }, + { + "epoch": 0.3534189805221716, + "grad_norm": 0.25506895780563354, + "learning_rate": 3.233319519270618e-06, + "loss": 0.535, + "step": 4264 + }, + { + "epoch": 0.35350186489846663, + "grad_norm": 0.29387593269348145, + "learning_rate": 3.2329050973891424e-06, + "loss": 0.5583, + "step": 4265 + }, + { + "epoch": 0.3535847492747617, + "grad_norm": 0.2878546118736267, + "learning_rate": 3.2324906755076668e-06, + "loss": 0.5591, + "step": 4266 + }, + { + "epoch": 0.3536676336510568, + "grad_norm": 0.3229694068431854, + "learning_rate": 3.2320762536261916e-06, + "loss": 0.5398, + "step": 4267 + }, + { + "epoch": 0.35375051802735186, + "grad_norm": 0.26509809494018555, + "learning_rate": 3.231661831744716e-06, + "loss": 0.5895, + "step": 4268 + }, + { + "epoch": 0.3538334024036469, + "grad_norm": 0.26131337881088257, + "learning_rate": 3.2312474098632414e-06, + "loss": 0.5095, + "step": 4269 + }, + { + "epoch": 0.35391628677994197, + "grad_norm": 0.2759053111076355, + "learning_rate": 3.2308329879817658e-06, + "loss": 0.5167, + "step": 4270 + }, + { + "epoch": 0.353999171156237, + "grad_norm": 0.28475049138069153, + "learning_rate": 3.2304185661002902e-06, + "loss": 0.5566, + "step": 4271 + }, + { + "epoch": 0.35408205553253214, + "grad_norm": 0.27493414282798767, + "learning_rate": 3.230004144218815e-06, + "loss": 0.5278, + "step": 4272 + }, + { + "epoch": 0.3541649399088272, + "grad_norm": 0.26517194509506226, + "learning_rate": 3.2295897223373395e-06, + "loss": 0.5263, + "step": 4273 + }, + { + "epoch": 0.35424782428512225, + "grad_norm": 0.2537805140018463, + "learning_rate": 3.2291753004558648e-06, + "loss": 0.5287, + "step": 4274 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 0.27218496799468994, + "learning_rate": 3.228760878574389e-06, + "loss": 0.5392, + "step": 4275 + }, + { + "epoch": 0.3544135930377124, + "grad_norm": 0.25797948241233826, + "learning_rate": 3.2283464566929136e-06, + "loss": 0.5101, + "step": 4276 + }, + { + "epoch": 0.35449647741400747, + "grad_norm": 0.2635766565799713, + "learning_rate": 3.2279320348114385e-06, + "loss": 0.5424, + "step": 4277 + }, + { + "epoch": 0.3545793617903025, + "grad_norm": 0.2707141935825348, + "learning_rate": 3.227517612929963e-06, + "loss": 0.5526, + "step": 4278 + }, + { + "epoch": 0.3546622461665976, + "grad_norm": 0.24533818662166595, + "learning_rate": 3.2271031910484874e-06, + "loss": 0.5105, + "step": 4279 + }, + { + "epoch": 0.35474513054289264, + "grad_norm": 0.2548004984855652, + "learning_rate": 3.2266887691670122e-06, + "loss": 0.5103, + "step": 4280 + }, + { + "epoch": 0.35482801491918775, + "grad_norm": 0.29784372448921204, + "learning_rate": 3.2262743472855366e-06, + "loss": 0.5475, + "step": 4281 + }, + { + "epoch": 0.3549108992954828, + "grad_norm": 0.2535659968852997, + "learning_rate": 3.225859925404062e-06, + "loss": 0.5029, + "step": 4282 + }, + { + "epoch": 0.35499378367177786, + "grad_norm": 0.2520844340324402, + "learning_rate": 3.2254455035225864e-06, + "loss": 0.547, + "step": 4283 + }, + { + "epoch": 0.3550766680480729, + "grad_norm": 0.26088300347328186, + "learning_rate": 3.225031081641111e-06, + "loss": 0.5317, + "step": 4284 + }, + { + "epoch": 0.35515955242436803, + "grad_norm": 0.25544241070747375, + "learning_rate": 3.2246166597596356e-06, + "loss": 0.4969, + "step": 4285 + }, + { + "epoch": 0.3552424368006631, + "grad_norm": 0.3047521412372589, + "learning_rate": 3.22420223787816e-06, + "loss": 0.5171, + "step": 4286 + }, + { + "epoch": 0.35532532117695814, + "grad_norm": 0.26482897996902466, + "learning_rate": 3.223787815996685e-06, + "loss": 0.5497, + "step": 4287 + }, + { + "epoch": 0.3554082055532532, + "grad_norm": 0.2538937032222748, + "learning_rate": 3.2233733941152094e-06, + "loss": 0.5186, + "step": 4288 + }, + { + "epoch": 0.35549108992954825, + "grad_norm": 0.24527312815189362, + "learning_rate": 3.222958972233734e-06, + "loss": 0.5007, + "step": 4289 + }, + { + "epoch": 0.35557397430584337, + "grad_norm": 0.2611848711967468, + "learning_rate": 3.222544550352259e-06, + "loss": 0.5467, + "step": 4290 + }, + { + "epoch": 0.3556568586821384, + "grad_norm": 0.2744581401348114, + "learning_rate": 3.2221301284707835e-06, + "loss": 0.5211, + "step": 4291 + }, + { + "epoch": 0.3557397430584335, + "grad_norm": 0.262174129486084, + "learning_rate": 3.2217157065893084e-06, + "loss": 0.5778, + "step": 4292 + }, + { + "epoch": 0.35582262743472853, + "grad_norm": 0.26937130093574524, + "learning_rate": 3.221301284707833e-06, + "loss": 0.5594, + "step": 4293 + }, + { + "epoch": 0.35590551181102364, + "grad_norm": 0.2559088468551636, + "learning_rate": 3.2208868628263572e-06, + "loss": 0.5128, + "step": 4294 + }, + { + "epoch": 0.3559883961873187, + "grad_norm": 0.2556886374950409, + "learning_rate": 3.2204724409448825e-06, + "loss": 0.5089, + "step": 4295 + }, + { + "epoch": 0.35607128056361376, + "grad_norm": 0.25384238362312317, + "learning_rate": 3.220058019063407e-06, + "loss": 0.4899, + "step": 4296 + }, + { + "epoch": 0.3561541649399088, + "grad_norm": 0.269381582736969, + "learning_rate": 3.2196435971819314e-06, + "loss": 0.5716, + "step": 4297 + }, + { + "epoch": 0.35623704931620387, + "grad_norm": 0.25652503967285156, + "learning_rate": 3.2192291753004562e-06, + "loss": 0.5215, + "step": 4298 + }, + { + "epoch": 0.356319933692499, + "grad_norm": 0.25346869230270386, + "learning_rate": 3.2188147534189807e-06, + "loss": 0.5238, + "step": 4299 + }, + { + "epoch": 0.35640281806879404, + "grad_norm": 0.28985118865966797, + "learning_rate": 3.2184003315375055e-06, + "loss": 0.5627, + "step": 4300 + }, + { + "epoch": 0.3564857024450891, + "grad_norm": 0.26572316884994507, + "learning_rate": 3.21798590965603e-06, + "loss": 0.5012, + "step": 4301 + }, + { + "epoch": 0.35656858682138415, + "grad_norm": 0.23837706446647644, + "learning_rate": 3.2175714877745544e-06, + "loss": 0.5211, + "step": 4302 + }, + { + "epoch": 0.35665147119767926, + "grad_norm": 0.25944700837135315, + "learning_rate": 3.2171570658930797e-06, + "loss": 0.5598, + "step": 4303 + }, + { + "epoch": 0.3567343555739743, + "grad_norm": 0.24977558851242065, + "learning_rate": 3.216742644011604e-06, + "loss": 0.5451, + "step": 4304 + }, + { + "epoch": 0.35681723995026937, + "grad_norm": 0.26089873909950256, + "learning_rate": 3.216328222130129e-06, + "loss": 0.5083, + "step": 4305 + }, + { + "epoch": 0.3569001243265644, + "grad_norm": 0.26611506938934326, + "learning_rate": 3.2159138002486534e-06, + "loss": 0.5484, + "step": 4306 + }, + { + "epoch": 0.3569830087028595, + "grad_norm": 0.2475583255290985, + "learning_rate": 3.215499378367178e-06, + "loss": 0.528, + "step": 4307 + }, + { + "epoch": 0.3570658930791546, + "grad_norm": 0.2378983497619629, + "learning_rate": 3.2150849564857027e-06, + "loss": 0.4995, + "step": 4308 + }, + { + "epoch": 0.35714877745544965, + "grad_norm": 0.25695908069610596, + "learning_rate": 3.214670534604227e-06, + "loss": 0.5488, + "step": 4309 + }, + { + "epoch": 0.3572316618317447, + "grad_norm": 0.2650492787361145, + "learning_rate": 3.2142561127227524e-06, + "loss": 0.4908, + "step": 4310 + }, + { + "epoch": 0.35731454620803976, + "grad_norm": 0.25345519185066223, + "learning_rate": 3.213841690841277e-06, + "loss": 0.4674, + "step": 4311 + }, + { + "epoch": 0.3573974305843349, + "grad_norm": 0.27587440609931946, + "learning_rate": 3.2134272689598012e-06, + "loss": 0.5031, + "step": 4312 + }, + { + "epoch": 0.35748031496062993, + "grad_norm": 0.27788209915161133, + "learning_rate": 3.213012847078326e-06, + "loss": 0.5141, + "step": 4313 + }, + { + "epoch": 0.357563199336925, + "grad_norm": 0.26530686020851135, + "learning_rate": 3.2125984251968505e-06, + "loss": 0.5416, + "step": 4314 + }, + { + "epoch": 0.35764608371322004, + "grad_norm": 0.26278984546661377, + "learning_rate": 3.212184003315375e-06, + "loss": 0.5263, + "step": 4315 + }, + { + "epoch": 0.35772896808951515, + "grad_norm": 0.2695300877094269, + "learning_rate": 3.2117695814339002e-06, + "loss": 0.5186, + "step": 4316 + }, + { + "epoch": 0.3578118524658102, + "grad_norm": 0.2548812925815582, + "learning_rate": 3.2113551595524247e-06, + "loss": 0.5061, + "step": 4317 + }, + { + "epoch": 0.35789473684210527, + "grad_norm": 0.27660539746284485, + "learning_rate": 3.2109407376709495e-06, + "loss": 0.5378, + "step": 4318 + }, + { + "epoch": 0.3579776212184003, + "grad_norm": 0.2821427583694458, + "learning_rate": 3.210526315789474e-06, + "loss": 0.5322, + "step": 4319 + }, + { + "epoch": 0.3580605055946954, + "grad_norm": 0.2513090968132019, + "learning_rate": 3.2101118939079984e-06, + "loss": 0.5618, + "step": 4320 + }, + { + "epoch": 0.3581433899709905, + "grad_norm": 0.25751662254333496, + "learning_rate": 3.2096974720265232e-06, + "loss": 0.5015, + "step": 4321 + }, + { + "epoch": 0.35822627434728554, + "grad_norm": 0.2806255519390106, + "learning_rate": 3.2092830501450477e-06, + "loss": 0.5309, + "step": 4322 + }, + { + "epoch": 0.3583091587235806, + "grad_norm": 0.2661028802394867, + "learning_rate": 3.208868628263573e-06, + "loss": 0.5053, + "step": 4323 + }, + { + "epoch": 0.35839204309987566, + "grad_norm": 0.2561163306236267, + "learning_rate": 3.2084542063820974e-06, + "loss": 0.5311, + "step": 4324 + }, + { + "epoch": 0.35847492747617077, + "grad_norm": 0.26765114068984985, + "learning_rate": 3.208039784500622e-06, + "loss": 0.5359, + "step": 4325 + }, + { + "epoch": 0.3585578118524658, + "grad_norm": 0.26500892639160156, + "learning_rate": 3.2076253626191467e-06, + "loss": 0.5497, + "step": 4326 + }, + { + "epoch": 0.3586406962287609, + "grad_norm": 0.28996944427490234, + "learning_rate": 3.207210940737671e-06, + "loss": 0.502, + "step": 4327 + }, + { + "epoch": 0.35872358060505594, + "grad_norm": 0.29820355772972107, + "learning_rate": 3.206796518856196e-06, + "loss": 0.5331, + "step": 4328 + }, + { + "epoch": 0.358806464981351, + "grad_norm": 0.26402780413627625, + "learning_rate": 3.2063820969747204e-06, + "loss": 0.5066, + "step": 4329 + }, + { + "epoch": 0.3588893493576461, + "grad_norm": 0.2634715139865875, + "learning_rate": 3.205967675093245e-06, + "loss": 0.5336, + "step": 4330 + }, + { + "epoch": 0.35897223373394116, + "grad_norm": 0.2499314546585083, + "learning_rate": 3.20555325321177e-06, + "loss": 0.5367, + "step": 4331 + }, + { + "epoch": 0.3590551181102362, + "grad_norm": 0.23346784710884094, + "learning_rate": 3.2051388313302945e-06, + "loss": 0.4952, + "step": 4332 + }, + { + "epoch": 0.35913800248653127, + "grad_norm": 0.27352601289749146, + "learning_rate": 3.204724409448819e-06, + "loss": 0.5114, + "step": 4333 + }, + { + "epoch": 0.3592208868628264, + "grad_norm": 0.25713086128234863, + "learning_rate": 3.204309987567344e-06, + "loss": 0.5056, + "step": 4334 + }, + { + "epoch": 0.35930377123912144, + "grad_norm": 0.27379512786865234, + "learning_rate": 3.2038955656858682e-06, + "loss": 0.521, + "step": 4335 + }, + { + "epoch": 0.3593866556154165, + "grad_norm": 0.3897377550601959, + "learning_rate": 3.2034811438043935e-06, + "loss": 0.55, + "step": 4336 + }, + { + "epoch": 0.35946953999171155, + "grad_norm": 0.2662034332752228, + "learning_rate": 3.203066721922918e-06, + "loss": 0.509, + "step": 4337 + }, + { + "epoch": 0.3595524243680066, + "grad_norm": 0.2548648416996002, + "learning_rate": 3.2026523000414424e-06, + "loss": 0.5287, + "step": 4338 + }, + { + "epoch": 0.3596353087443017, + "grad_norm": 0.2532671391963959, + "learning_rate": 3.2022378781599672e-06, + "loss": 0.4958, + "step": 4339 + }, + { + "epoch": 0.3597181931205968, + "grad_norm": 0.2678300738334656, + "learning_rate": 3.2018234562784917e-06, + "loss": 0.4979, + "step": 4340 + }, + { + "epoch": 0.35980107749689183, + "grad_norm": 0.26467910408973694, + "learning_rate": 3.2014090343970165e-06, + "loss": 0.5533, + "step": 4341 + }, + { + "epoch": 0.3598839618731869, + "grad_norm": 0.25082409381866455, + "learning_rate": 3.200994612515541e-06, + "loss": 0.5459, + "step": 4342 + }, + { + "epoch": 0.359966846249482, + "grad_norm": 0.28025171160697937, + "learning_rate": 3.2005801906340654e-06, + "loss": 0.5416, + "step": 4343 + }, + { + "epoch": 0.36004973062577705, + "grad_norm": 0.27990081906318665, + "learning_rate": 3.2001657687525907e-06, + "loss": 0.576, + "step": 4344 + }, + { + "epoch": 0.3601326150020721, + "grad_norm": 0.2858120799064636, + "learning_rate": 3.199751346871115e-06, + "loss": 0.5589, + "step": 4345 + }, + { + "epoch": 0.36021549937836717, + "grad_norm": 0.24404999613761902, + "learning_rate": 3.19933692498964e-06, + "loss": 0.508, + "step": 4346 + }, + { + "epoch": 0.3602983837546622, + "grad_norm": 0.25410592555999756, + "learning_rate": 3.1989225031081644e-06, + "loss": 0.5162, + "step": 4347 + }, + { + "epoch": 0.36038126813095733, + "grad_norm": 0.25369688868522644, + "learning_rate": 3.198508081226689e-06, + "loss": 0.5208, + "step": 4348 + }, + { + "epoch": 0.3604641525072524, + "grad_norm": 0.24505138397216797, + "learning_rate": 3.1980936593452137e-06, + "loss": 0.4955, + "step": 4349 + }, + { + "epoch": 0.36054703688354744, + "grad_norm": 0.26362499594688416, + "learning_rate": 3.197679237463738e-06, + "loss": 0.4952, + "step": 4350 + }, + { + "epoch": 0.3606299212598425, + "grad_norm": 0.2550717890262604, + "learning_rate": 3.197264815582263e-06, + "loss": 0.4878, + "step": 4351 + }, + { + "epoch": 0.3607128056361376, + "grad_norm": 0.26884955167770386, + "learning_rate": 3.196850393700788e-06, + "loss": 0.5109, + "step": 4352 + }, + { + "epoch": 0.36079569001243267, + "grad_norm": 0.2787870168685913, + "learning_rate": 3.1964359718193123e-06, + "loss": 0.5698, + "step": 4353 + }, + { + "epoch": 0.3608785743887277, + "grad_norm": 0.28908079862594604, + "learning_rate": 3.196021549937837e-06, + "loss": 0.5311, + "step": 4354 + }, + { + "epoch": 0.3609614587650228, + "grad_norm": 0.2379661500453949, + "learning_rate": 3.1956071280563615e-06, + "loss": 0.5378, + "step": 4355 + }, + { + "epoch": 0.36104434314131784, + "grad_norm": 0.28336629271507263, + "learning_rate": 3.195192706174886e-06, + "loss": 0.5219, + "step": 4356 + }, + { + "epoch": 0.36112722751761295, + "grad_norm": 0.275291383266449, + "learning_rate": 3.1947782842934113e-06, + "loss": 0.5811, + "step": 4357 + }, + { + "epoch": 0.361210111893908, + "grad_norm": 0.2958420515060425, + "learning_rate": 3.1943638624119357e-06, + "loss": 0.4935, + "step": 4358 + }, + { + "epoch": 0.36129299627020306, + "grad_norm": 0.2618706524372101, + "learning_rate": 3.1939494405304605e-06, + "loss": 0.531, + "step": 4359 + }, + { + "epoch": 0.3613758806464981, + "grad_norm": 0.2592873275279999, + "learning_rate": 3.193535018648985e-06, + "loss": 0.5641, + "step": 4360 + }, + { + "epoch": 0.3614587650227932, + "grad_norm": 0.25079256296157837, + "learning_rate": 3.1931205967675094e-06, + "loss": 0.5307, + "step": 4361 + }, + { + "epoch": 0.3615416493990883, + "grad_norm": 0.2620052695274353, + "learning_rate": 3.1927061748860343e-06, + "loss": 0.4988, + "step": 4362 + }, + { + "epoch": 0.36162453377538334, + "grad_norm": 0.25854387879371643, + "learning_rate": 3.1922917530045587e-06, + "loss": 0.5409, + "step": 4363 + }, + { + "epoch": 0.3617074181516784, + "grad_norm": 0.25438040494918823, + "learning_rate": 3.191877331123083e-06, + "loss": 0.5044, + "step": 4364 + }, + { + "epoch": 0.36179030252797345, + "grad_norm": 0.27045586705207825, + "learning_rate": 3.1914629092416084e-06, + "loss": 0.5037, + "step": 4365 + }, + { + "epoch": 0.36187318690426856, + "grad_norm": 0.2805004119873047, + "learning_rate": 3.191048487360133e-06, + "loss": 0.53, + "step": 4366 + }, + { + "epoch": 0.3619560712805636, + "grad_norm": 0.26493600010871887, + "learning_rate": 3.1906340654786577e-06, + "loss": 0.5486, + "step": 4367 + }, + { + "epoch": 0.3620389556568587, + "grad_norm": 0.24093204736709595, + "learning_rate": 3.190219643597182e-06, + "loss": 0.5019, + "step": 4368 + }, + { + "epoch": 0.36212184003315373, + "grad_norm": 0.23654396831989288, + "learning_rate": 3.1898052217157066e-06, + "loss": 0.4837, + "step": 4369 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 0.2391032576560974, + "learning_rate": 3.1893907998342314e-06, + "loss": 0.5067, + "step": 4370 + }, + { + "epoch": 0.3622876087857439, + "grad_norm": 0.2512289881706238, + "learning_rate": 3.1889763779527563e-06, + "loss": 0.5069, + "step": 4371 + }, + { + "epoch": 0.36237049316203895, + "grad_norm": 0.26164332032203674, + "learning_rate": 3.188561956071281e-06, + "loss": 0.5636, + "step": 4372 + }, + { + "epoch": 0.362453377538334, + "grad_norm": 0.2932446300983429, + "learning_rate": 3.1881475341898056e-06, + "loss": 0.5699, + "step": 4373 + }, + { + "epoch": 0.36253626191462907, + "grad_norm": 0.28556203842163086, + "learning_rate": 3.18773311230833e-06, + "loss": 0.54, + "step": 4374 + }, + { + "epoch": 0.3626191462909242, + "grad_norm": 0.26350900530815125, + "learning_rate": 3.187318690426855e-06, + "loss": 0.5421, + "step": 4375 + }, + { + "epoch": 0.36270203066721923, + "grad_norm": 0.2514267861843109, + "learning_rate": 3.1869042685453793e-06, + "loss": 0.5314, + "step": 4376 + }, + { + "epoch": 0.3627849150435143, + "grad_norm": 0.23991777002811432, + "learning_rate": 3.1864898466639045e-06, + "loss": 0.5408, + "step": 4377 + }, + { + "epoch": 0.36286779941980934, + "grad_norm": 0.27222612500190735, + "learning_rate": 3.186075424782429e-06, + "loss": 0.55, + "step": 4378 + }, + { + "epoch": 0.36295068379610446, + "grad_norm": 0.2568392753601074, + "learning_rate": 3.1856610029009534e-06, + "loss": 0.5164, + "step": 4379 + }, + { + "epoch": 0.3630335681723995, + "grad_norm": 0.24057120084762573, + "learning_rate": 3.1852465810194783e-06, + "loss": 0.5035, + "step": 4380 + }, + { + "epoch": 0.36311645254869457, + "grad_norm": 0.2756480872631073, + "learning_rate": 3.1848321591380027e-06, + "loss": 0.555, + "step": 4381 + }, + { + "epoch": 0.3631993369249896, + "grad_norm": 0.250160813331604, + "learning_rate": 3.184417737256527e-06, + "loss": 0.4843, + "step": 4382 + }, + { + "epoch": 0.3632822213012847, + "grad_norm": 0.25771743059158325, + "learning_rate": 3.184003315375052e-06, + "loss": 0.5187, + "step": 4383 + }, + { + "epoch": 0.3633651056775798, + "grad_norm": 0.26376691460609436, + "learning_rate": 3.1835888934935764e-06, + "loss": 0.5311, + "step": 4384 + }, + { + "epoch": 0.36344799005387485, + "grad_norm": 0.2818605601787567, + "learning_rate": 3.1831744716121017e-06, + "loss": 0.488, + "step": 4385 + }, + { + "epoch": 0.3635308744301699, + "grad_norm": 0.25945472717285156, + "learning_rate": 3.182760049730626e-06, + "loss": 0.5088, + "step": 4386 + }, + { + "epoch": 0.36361375880646496, + "grad_norm": 0.25067901611328125, + "learning_rate": 3.1823456278491506e-06, + "loss": 0.5035, + "step": 4387 + }, + { + "epoch": 0.36369664318276007, + "grad_norm": 0.29276251792907715, + "learning_rate": 3.1819312059676754e-06, + "loss": 0.5472, + "step": 4388 + }, + { + "epoch": 0.3637795275590551, + "grad_norm": 0.2492397278547287, + "learning_rate": 3.1815167840862e-06, + "loss": 0.4682, + "step": 4389 + }, + { + "epoch": 0.3638624119353502, + "grad_norm": 0.24685727059841156, + "learning_rate": 3.1811023622047247e-06, + "loss": 0.5356, + "step": 4390 + }, + { + "epoch": 0.36394529631164524, + "grad_norm": 0.2709459960460663, + "learning_rate": 3.180687940323249e-06, + "loss": 0.5479, + "step": 4391 + }, + { + "epoch": 0.36402818068794035, + "grad_norm": 0.24930472671985626, + "learning_rate": 3.180273518441774e-06, + "loss": 0.507, + "step": 4392 + }, + { + "epoch": 0.3641110650642354, + "grad_norm": 0.2669527232646942, + "learning_rate": 3.179859096560299e-06, + "loss": 0.5527, + "step": 4393 + }, + { + "epoch": 0.36419394944053046, + "grad_norm": 0.2548328936100006, + "learning_rate": 3.1794446746788233e-06, + "loss": 0.5521, + "step": 4394 + }, + { + "epoch": 0.3642768338168255, + "grad_norm": 0.24388965964317322, + "learning_rate": 3.179030252797348e-06, + "loss": 0.5059, + "step": 4395 + }, + { + "epoch": 0.3643597181931206, + "grad_norm": 0.2511399984359741, + "learning_rate": 3.1786158309158726e-06, + "loss": 0.5605, + "step": 4396 + }, + { + "epoch": 0.3644426025694157, + "grad_norm": 0.23583073914051056, + "learning_rate": 3.178201409034397e-06, + "loss": 0.4905, + "step": 4397 + }, + { + "epoch": 0.36452548694571074, + "grad_norm": 0.26612532138824463, + "learning_rate": 3.1777869871529223e-06, + "loss": 0.5224, + "step": 4398 + }, + { + "epoch": 0.3646083713220058, + "grad_norm": 0.277444452047348, + "learning_rate": 3.1773725652714467e-06, + "loss": 0.5211, + "step": 4399 + }, + { + "epoch": 0.36469125569830085, + "grad_norm": 0.2776162326335907, + "learning_rate": 3.176958143389971e-06, + "loss": 0.5866, + "step": 4400 + }, + { + "epoch": 0.36477414007459597, + "grad_norm": 0.28187087178230286, + "learning_rate": 3.176543721508496e-06, + "loss": 0.5449, + "step": 4401 + }, + { + "epoch": 0.364857024450891, + "grad_norm": 0.2568029463291168, + "learning_rate": 3.1761292996270204e-06, + "loss": 0.5623, + "step": 4402 + }, + { + "epoch": 0.3649399088271861, + "grad_norm": 0.25029778480529785, + "learning_rate": 3.1757148777455453e-06, + "loss": 0.5289, + "step": 4403 + }, + { + "epoch": 0.36502279320348113, + "grad_norm": 0.2503984272480011, + "learning_rate": 3.1753004558640697e-06, + "loss": 0.5403, + "step": 4404 + }, + { + "epoch": 0.3651056775797762, + "grad_norm": 0.25819897651672363, + "learning_rate": 3.174886033982594e-06, + "loss": 0.5086, + "step": 4405 + }, + { + "epoch": 0.3651885619560713, + "grad_norm": 0.26152005791664124, + "learning_rate": 3.1744716121011194e-06, + "loss": 0.5007, + "step": 4406 + }, + { + "epoch": 0.36527144633236636, + "grad_norm": 0.24319732189178467, + "learning_rate": 3.174057190219644e-06, + "loss": 0.4809, + "step": 4407 + }, + { + "epoch": 0.3653543307086614, + "grad_norm": 0.24862316250801086, + "learning_rate": 3.1736427683381687e-06, + "loss": 0.5067, + "step": 4408 + }, + { + "epoch": 0.36543721508495647, + "grad_norm": 0.26848104596138, + "learning_rate": 3.173228346456693e-06, + "loss": 0.5471, + "step": 4409 + }, + { + "epoch": 0.3655200994612516, + "grad_norm": 0.2546938955783844, + "learning_rate": 3.1728139245752176e-06, + "loss": 0.5153, + "step": 4410 + }, + { + "epoch": 0.36560298383754664, + "grad_norm": 0.25505197048187256, + "learning_rate": 3.1723995026937424e-06, + "loss": 0.5208, + "step": 4411 + }, + { + "epoch": 0.3656858682138417, + "grad_norm": 0.256343275308609, + "learning_rate": 3.1719850808122673e-06, + "loss": 0.5364, + "step": 4412 + }, + { + "epoch": 0.36576875259013675, + "grad_norm": 0.2809293866157532, + "learning_rate": 3.171570658930792e-06, + "loss": 0.5635, + "step": 4413 + }, + { + "epoch": 0.3658516369664318, + "grad_norm": 0.2586604356765747, + "learning_rate": 3.1711562370493166e-06, + "loss": 0.5456, + "step": 4414 + }, + { + "epoch": 0.3659345213427269, + "grad_norm": 0.2711935341358185, + "learning_rate": 3.170741815167841e-06, + "loss": 0.541, + "step": 4415 + }, + { + "epoch": 0.36601740571902197, + "grad_norm": 0.2498967945575714, + "learning_rate": 3.170327393286366e-06, + "loss": 0.5383, + "step": 4416 + }, + { + "epoch": 0.366100290095317, + "grad_norm": 0.25912147760391235, + "learning_rate": 3.1699129714048903e-06, + "loss": 0.5052, + "step": 4417 + }, + { + "epoch": 0.3661831744716121, + "grad_norm": 0.258775919675827, + "learning_rate": 3.1694985495234147e-06, + "loss": 0.5159, + "step": 4418 + }, + { + "epoch": 0.3662660588479072, + "grad_norm": 0.2801685035228729, + "learning_rate": 3.16908412764194e-06, + "loss": 0.5267, + "step": 4419 + }, + { + "epoch": 0.36634894322420225, + "grad_norm": 0.2534509003162384, + "learning_rate": 3.1686697057604644e-06, + "loss": 0.4687, + "step": 4420 + }, + { + "epoch": 0.3664318276004973, + "grad_norm": 0.2514853775501251, + "learning_rate": 3.1682552838789893e-06, + "loss": 0.533, + "step": 4421 + }, + { + "epoch": 0.36651471197679236, + "grad_norm": 0.2636883854866028, + "learning_rate": 3.1678408619975137e-06, + "loss": 0.5162, + "step": 4422 + }, + { + "epoch": 0.3665975963530874, + "grad_norm": 0.2277776449918747, + "learning_rate": 3.167426440116038e-06, + "loss": 0.501, + "step": 4423 + }, + { + "epoch": 0.36668048072938253, + "grad_norm": 0.2990654706954956, + "learning_rate": 3.167012018234563e-06, + "loss": 0.5298, + "step": 4424 + }, + { + "epoch": 0.3667633651056776, + "grad_norm": 0.24158500134944916, + "learning_rate": 3.1665975963530874e-06, + "loss": 0.4852, + "step": 4425 + }, + { + "epoch": 0.36684624948197264, + "grad_norm": 0.25437992811203003, + "learning_rate": 3.1661831744716127e-06, + "loss": 0.5103, + "step": 4426 + }, + { + "epoch": 0.3669291338582677, + "grad_norm": 0.24211212992668152, + "learning_rate": 3.165768752590137e-06, + "loss": 0.5032, + "step": 4427 + }, + { + "epoch": 0.3670120182345628, + "grad_norm": 0.24681976437568665, + "learning_rate": 3.1653543307086616e-06, + "loss": 0.5082, + "step": 4428 + }, + { + "epoch": 0.36709490261085786, + "grad_norm": 0.28851884603500366, + "learning_rate": 3.1649399088271864e-06, + "loss": 0.5363, + "step": 4429 + }, + { + "epoch": 0.3671777869871529, + "grad_norm": 0.26948392391204834, + "learning_rate": 3.164525486945711e-06, + "loss": 0.4971, + "step": 4430 + }, + { + "epoch": 0.367260671363448, + "grad_norm": 0.2573961317539215, + "learning_rate": 3.1641110650642353e-06, + "loss": 0.5171, + "step": 4431 + }, + { + "epoch": 0.36734355573974303, + "grad_norm": 0.2722170650959015, + "learning_rate": 3.1636966431827606e-06, + "loss": 0.5457, + "step": 4432 + }, + { + "epoch": 0.36742644011603814, + "grad_norm": 0.2764262557029724, + "learning_rate": 3.163282221301285e-06, + "loss": 0.5242, + "step": 4433 + }, + { + "epoch": 0.3675093244923332, + "grad_norm": 0.24548961222171783, + "learning_rate": 3.16286779941981e-06, + "loss": 0.5233, + "step": 4434 + }, + { + "epoch": 0.36759220886862826, + "grad_norm": 0.2521306276321411, + "learning_rate": 3.1624533775383343e-06, + "loss": 0.5515, + "step": 4435 + }, + { + "epoch": 0.3676750932449233, + "grad_norm": 0.2545776665210724, + "learning_rate": 3.1620389556568587e-06, + "loss": 0.5106, + "step": 4436 + }, + { + "epoch": 0.3677579776212184, + "grad_norm": 0.29353195428848267, + "learning_rate": 3.1616245337753836e-06, + "loss": 0.5987, + "step": 4437 + }, + { + "epoch": 0.3678408619975135, + "grad_norm": 0.24504348635673523, + "learning_rate": 3.161210111893908e-06, + "loss": 0.5146, + "step": 4438 + }, + { + "epoch": 0.36792374637380854, + "grad_norm": 0.267159640789032, + "learning_rate": 3.1607956900124333e-06, + "loss": 0.5873, + "step": 4439 + }, + { + "epoch": 0.3680066307501036, + "grad_norm": 0.25004738569259644, + "learning_rate": 3.1603812681309577e-06, + "loss": 0.5333, + "step": 4440 + }, + { + "epoch": 0.36808951512639865, + "grad_norm": 0.2524699568748474, + "learning_rate": 3.159966846249482e-06, + "loss": 0.5256, + "step": 4441 + }, + { + "epoch": 0.36817239950269376, + "grad_norm": 0.24823583662509918, + "learning_rate": 3.159552424368007e-06, + "loss": 0.532, + "step": 4442 + }, + { + "epoch": 0.3682552838789888, + "grad_norm": 0.2527320683002472, + "learning_rate": 3.1591380024865314e-06, + "loss": 0.5172, + "step": 4443 + }, + { + "epoch": 0.36833816825528387, + "grad_norm": 0.2523735463619232, + "learning_rate": 3.1587235806050563e-06, + "loss": 0.4901, + "step": 4444 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 0.265667587518692, + "learning_rate": 3.1583091587235807e-06, + "loss": 0.5335, + "step": 4445 + }, + { + "epoch": 0.36850393700787404, + "grad_norm": 0.23311595618724823, + "learning_rate": 3.157894736842105e-06, + "loss": 0.4996, + "step": 4446 + }, + { + "epoch": 0.3685868213841691, + "grad_norm": 0.25594520568847656, + "learning_rate": 3.1574803149606304e-06, + "loss": 0.4826, + "step": 4447 + }, + { + "epoch": 0.36866970576046415, + "grad_norm": 0.24246081709861755, + "learning_rate": 3.157065893079155e-06, + "loss": 0.4615, + "step": 4448 + }, + { + "epoch": 0.3687525901367592, + "grad_norm": 0.26423129439353943, + "learning_rate": 3.1566514711976793e-06, + "loss": 0.5479, + "step": 4449 + }, + { + "epoch": 0.36883547451305426, + "grad_norm": 0.2555503845214844, + "learning_rate": 3.156237049316204e-06, + "loss": 0.4861, + "step": 4450 + }, + { + "epoch": 0.3689183588893494, + "grad_norm": 0.263784259557724, + "learning_rate": 3.1558226274347286e-06, + "loss": 0.5039, + "step": 4451 + }, + { + "epoch": 0.36900124326564443, + "grad_norm": 0.24443583190441132, + "learning_rate": 3.155408205553254e-06, + "loss": 0.4855, + "step": 4452 + }, + { + "epoch": 0.3690841276419395, + "grad_norm": 0.26842352747917175, + "learning_rate": 3.1549937836717783e-06, + "loss": 0.5516, + "step": 4453 + }, + { + "epoch": 0.36916701201823454, + "grad_norm": 0.2618817985057831, + "learning_rate": 3.1545793617903027e-06, + "loss": 0.5215, + "step": 4454 + }, + { + "epoch": 0.36924989639452965, + "grad_norm": 0.2509889602661133, + "learning_rate": 3.1541649399088276e-06, + "loss": 0.5423, + "step": 4455 + }, + { + "epoch": 0.3693327807708247, + "grad_norm": 0.23931679129600525, + "learning_rate": 3.153750518027352e-06, + "loss": 0.5121, + "step": 4456 + }, + { + "epoch": 0.36941566514711976, + "grad_norm": 0.31251800060272217, + "learning_rate": 3.153336096145877e-06, + "loss": 0.571, + "step": 4457 + }, + { + "epoch": 0.3694985495234148, + "grad_norm": 0.2709120810031891, + "learning_rate": 3.1529216742644013e-06, + "loss": 0.5582, + "step": 4458 + }, + { + "epoch": 0.36958143389970993, + "grad_norm": 0.2624141573905945, + "learning_rate": 3.1525072523829257e-06, + "loss": 0.507, + "step": 4459 + }, + { + "epoch": 0.369664318276005, + "grad_norm": 0.2539939284324646, + "learning_rate": 3.152092830501451e-06, + "loss": 0.5796, + "step": 4460 + }, + { + "epoch": 0.36974720265230004, + "grad_norm": 0.2538832724094391, + "learning_rate": 3.1516784086199755e-06, + "loss": 0.5285, + "step": 4461 + }, + { + "epoch": 0.3698300870285951, + "grad_norm": 0.2682187557220459, + "learning_rate": 3.1512639867385003e-06, + "loss": 0.5698, + "step": 4462 + }, + { + "epoch": 0.36991297140489016, + "grad_norm": 0.25663691759109497, + "learning_rate": 3.1508495648570247e-06, + "loss": 0.5388, + "step": 4463 + }, + { + "epoch": 0.36999585578118527, + "grad_norm": 0.27961254119873047, + "learning_rate": 3.150435142975549e-06, + "loss": 0.5612, + "step": 4464 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 0.24793323874473572, + "learning_rate": 3.150020721094074e-06, + "loss": 0.5434, + "step": 4465 + }, + { + "epoch": 0.3701616245337754, + "grad_norm": 0.24709711968898773, + "learning_rate": 3.1496062992125985e-06, + "loss": 0.5157, + "step": 4466 + }, + { + "epoch": 0.37024450891007044, + "grad_norm": 0.23144908249378204, + "learning_rate": 3.149191877331123e-06, + "loss": 0.4963, + "step": 4467 + }, + { + "epoch": 0.37032739328636555, + "grad_norm": 0.277614027261734, + "learning_rate": 3.148777455449648e-06, + "loss": 0.5641, + "step": 4468 + }, + { + "epoch": 0.3704102776626606, + "grad_norm": 0.2484825700521469, + "learning_rate": 3.1483630335681726e-06, + "loss": 0.5124, + "step": 4469 + }, + { + "epoch": 0.37049316203895566, + "grad_norm": 0.25848227739334106, + "learning_rate": 3.1479486116866975e-06, + "loss": 0.5101, + "step": 4470 + }, + { + "epoch": 0.3705760464152507, + "grad_norm": 0.2846028208732605, + "learning_rate": 3.147534189805222e-06, + "loss": 0.5457, + "step": 4471 + }, + { + "epoch": 0.37065893079154577, + "grad_norm": 0.25750160217285156, + "learning_rate": 3.1471197679237463e-06, + "loss": 0.52, + "step": 4472 + }, + { + "epoch": 0.3707418151678409, + "grad_norm": 0.2606601417064667, + "learning_rate": 3.1467053460422716e-06, + "loss": 0.5156, + "step": 4473 + }, + { + "epoch": 0.37082469954413594, + "grad_norm": 0.24447637796401978, + "learning_rate": 3.146290924160796e-06, + "loss": 0.5123, + "step": 4474 + }, + { + "epoch": 0.370907583920431, + "grad_norm": 0.2577292323112488, + "learning_rate": 3.145876502279321e-06, + "loss": 0.5429, + "step": 4475 + }, + { + "epoch": 0.37099046829672605, + "grad_norm": 0.2602491080760956, + "learning_rate": 3.1454620803978453e-06, + "loss": 0.511, + "step": 4476 + }, + { + "epoch": 0.37107335267302116, + "grad_norm": 0.2623663544654846, + "learning_rate": 3.1450476585163698e-06, + "loss": 0.561, + "step": 4477 + }, + { + "epoch": 0.3711562370493162, + "grad_norm": 0.269637793302536, + "learning_rate": 3.1446332366348946e-06, + "loss": 0.5247, + "step": 4478 + }, + { + "epoch": 0.3712391214256113, + "grad_norm": 0.27250784635543823, + "learning_rate": 3.144218814753419e-06, + "loss": 0.5244, + "step": 4479 + }, + { + "epoch": 0.37132200580190633, + "grad_norm": 0.3740835189819336, + "learning_rate": 3.1438043928719443e-06, + "loss": 0.5134, + "step": 4480 + }, + { + "epoch": 0.3714048901782014, + "grad_norm": 0.24123939871788025, + "learning_rate": 3.1433899709904687e-06, + "loss": 0.5419, + "step": 4481 + }, + { + "epoch": 0.3714877745544965, + "grad_norm": 0.25900396704673767, + "learning_rate": 3.142975549108993e-06, + "loss": 0.4978, + "step": 4482 + }, + { + "epoch": 0.37157065893079155, + "grad_norm": 0.2511722445487976, + "learning_rate": 3.142561127227518e-06, + "loss": 0.5034, + "step": 4483 + }, + { + "epoch": 0.3716535433070866, + "grad_norm": 0.2674420475959778, + "learning_rate": 3.1421467053460425e-06, + "loss": 0.512, + "step": 4484 + }, + { + "epoch": 0.37173642768338166, + "grad_norm": 0.2565935254096985, + "learning_rate": 3.141732283464567e-06, + "loss": 0.5447, + "step": 4485 + }, + { + "epoch": 0.3718193120596768, + "grad_norm": 0.26311081647872925, + "learning_rate": 3.1413178615830918e-06, + "loss": 0.531, + "step": 4486 + }, + { + "epoch": 0.37190219643597183, + "grad_norm": 0.2684539556503296, + "learning_rate": 3.140903439701616e-06, + "loss": 0.5139, + "step": 4487 + }, + { + "epoch": 0.3719850808122669, + "grad_norm": 0.2718946933746338, + "learning_rate": 3.1404890178201415e-06, + "loss": 0.5895, + "step": 4488 + }, + { + "epoch": 0.37206796518856194, + "grad_norm": 0.26770398020744324, + "learning_rate": 3.140074595938666e-06, + "loss": 0.5475, + "step": 4489 + }, + { + "epoch": 0.372150849564857, + "grad_norm": 0.24514682590961456, + "learning_rate": 3.1396601740571903e-06, + "loss": 0.5355, + "step": 4490 + }, + { + "epoch": 0.3722337339411521, + "grad_norm": 0.2523571848869324, + "learning_rate": 3.139245752175715e-06, + "loss": 0.5485, + "step": 4491 + }, + { + "epoch": 0.37231661831744717, + "grad_norm": 0.24277153611183167, + "learning_rate": 3.1388313302942396e-06, + "loss": 0.5063, + "step": 4492 + }, + { + "epoch": 0.3723995026937422, + "grad_norm": 0.27630478143692017, + "learning_rate": 3.138416908412765e-06, + "loss": 0.5518, + "step": 4493 + }, + { + "epoch": 0.3724823870700373, + "grad_norm": 0.2788238525390625, + "learning_rate": 3.1380024865312893e-06, + "loss": 0.5381, + "step": 4494 + }, + { + "epoch": 0.3725652714463324, + "grad_norm": 0.2678964138031006, + "learning_rate": 3.1375880646498138e-06, + "loss": 0.5473, + "step": 4495 + }, + { + "epoch": 0.37264815582262745, + "grad_norm": 0.26566070318222046, + "learning_rate": 3.1371736427683386e-06, + "loss": 0.5392, + "step": 4496 + }, + { + "epoch": 0.3727310401989225, + "grad_norm": 0.2711072266101837, + "learning_rate": 3.136759220886863e-06, + "loss": 0.544, + "step": 4497 + }, + { + "epoch": 0.37281392457521756, + "grad_norm": 0.25798019766807556, + "learning_rate": 3.1363447990053875e-06, + "loss": 0.5707, + "step": 4498 + }, + { + "epoch": 0.3728968089515126, + "grad_norm": 0.2683127820491791, + "learning_rate": 3.1359303771239123e-06, + "loss": 0.5433, + "step": 4499 + }, + { + "epoch": 0.3729796933278077, + "grad_norm": 0.26864001154899597, + "learning_rate": 3.1355159552424368e-06, + "loss": 0.5545, + "step": 4500 + }, + { + "epoch": 0.3730625777041028, + "grad_norm": 0.2718157470226288, + "learning_rate": 3.135101533360962e-06, + "loss": 0.5374, + "step": 4501 + }, + { + "epoch": 0.37314546208039784, + "grad_norm": 0.40637171268463135, + "learning_rate": 3.1346871114794865e-06, + "loss": 0.5262, + "step": 4502 + }, + { + "epoch": 0.3732283464566929, + "grad_norm": 0.26562464237213135, + "learning_rate": 3.134272689598011e-06, + "loss": 0.5532, + "step": 4503 + }, + { + "epoch": 0.373311230832988, + "grad_norm": 0.29200342297554016, + "learning_rate": 3.1338582677165358e-06, + "loss": 0.5714, + "step": 4504 + }, + { + "epoch": 0.37339411520928306, + "grad_norm": 0.24161994457244873, + "learning_rate": 3.13344384583506e-06, + "loss": 0.5107, + "step": 4505 + }, + { + "epoch": 0.3734769995855781, + "grad_norm": 0.2671598792076111, + "learning_rate": 3.133029423953585e-06, + "loss": 0.4864, + "step": 4506 + }, + { + "epoch": 0.3735598839618732, + "grad_norm": 0.24898195266723633, + "learning_rate": 3.1326150020721095e-06, + "loss": 0.5104, + "step": 4507 + }, + { + "epoch": 0.37364276833816823, + "grad_norm": 0.25154662132263184, + "learning_rate": 3.132200580190634e-06, + "loss": 0.5654, + "step": 4508 + }, + { + "epoch": 0.37372565271446334, + "grad_norm": 0.24460871517658234, + "learning_rate": 3.131786158309159e-06, + "loss": 0.4976, + "step": 4509 + }, + { + "epoch": 0.3738085370907584, + "grad_norm": 0.2704656720161438, + "learning_rate": 3.1313717364276836e-06, + "loss": 0.5487, + "step": 4510 + }, + { + "epoch": 0.37389142146705345, + "grad_norm": 0.27459365129470825, + "learning_rate": 3.1309573145462085e-06, + "loss": 0.5452, + "step": 4511 + }, + { + "epoch": 0.3739743058433485, + "grad_norm": 0.269766241312027, + "learning_rate": 3.130542892664733e-06, + "loss": 0.5836, + "step": 4512 + }, + { + "epoch": 0.3740571902196436, + "grad_norm": 0.2774397134780884, + "learning_rate": 3.1301284707832573e-06, + "loss": 0.5613, + "step": 4513 + }, + { + "epoch": 0.3741400745959387, + "grad_norm": 0.25932255387306213, + "learning_rate": 3.1297140489017826e-06, + "loss": 0.514, + "step": 4514 + }, + { + "epoch": 0.37422295897223373, + "grad_norm": 0.2746933102607727, + "learning_rate": 3.129299627020307e-06, + "loss": 0.5541, + "step": 4515 + }, + { + "epoch": 0.3743058433485288, + "grad_norm": 0.2735837996006012, + "learning_rate": 3.1288852051388315e-06, + "loss": 0.5353, + "step": 4516 + }, + { + "epoch": 0.37438872772482384, + "grad_norm": 0.26978251338005066, + "learning_rate": 3.1284707832573563e-06, + "loss": 0.521, + "step": 4517 + }, + { + "epoch": 0.37447161210111896, + "grad_norm": 0.2586561143398285, + "learning_rate": 3.1280563613758808e-06, + "loss": 0.5128, + "step": 4518 + }, + { + "epoch": 0.374554496477414, + "grad_norm": 0.24769484996795654, + "learning_rate": 3.1276419394944056e-06, + "loss": 0.4743, + "step": 4519 + }, + { + "epoch": 0.37463738085370907, + "grad_norm": 0.23825685679912567, + "learning_rate": 3.12722751761293e-06, + "loss": 0.5157, + "step": 4520 + }, + { + "epoch": 0.3747202652300041, + "grad_norm": 0.24727608263492584, + "learning_rate": 3.1268130957314545e-06, + "loss": 0.5514, + "step": 4521 + }, + { + "epoch": 0.37480314960629924, + "grad_norm": 0.24377861618995667, + "learning_rate": 3.1263986738499798e-06, + "loss": 0.5051, + "step": 4522 + }, + { + "epoch": 0.3748860339825943, + "grad_norm": 0.24075868725776672, + "learning_rate": 3.125984251968504e-06, + "loss": 0.54, + "step": 4523 + }, + { + "epoch": 0.37496891835888935, + "grad_norm": 0.25905972719192505, + "learning_rate": 3.125569830087029e-06, + "loss": 0.498, + "step": 4524 + }, + { + "epoch": 0.3750518027351844, + "grad_norm": 0.25788310170173645, + "learning_rate": 3.1251554082055535e-06, + "loss": 0.5171, + "step": 4525 + }, + { + "epoch": 0.37513468711147946, + "grad_norm": 0.2342282235622406, + "learning_rate": 3.124740986324078e-06, + "loss": 0.4598, + "step": 4526 + }, + { + "epoch": 0.37521757148777457, + "grad_norm": 0.2847420573234558, + "learning_rate": 3.1243265644426028e-06, + "loss": 0.5476, + "step": 4527 + }, + { + "epoch": 0.3753004558640696, + "grad_norm": 0.25537294149398804, + "learning_rate": 3.123912142561127e-06, + "loss": 0.5261, + "step": 4528 + }, + { + "epoch": 0.3753833402403647, + "grad_norm": 0.2821756601333618, + "learning_rate": 3.1234977206796525e-06, + "loss": 0.5202, + "step": 4529 + }, + { + "epoch": 0.37546622461665974, + "grad_norm": 0.2343769520521164, + "learning_rate": 3.123083298798177e-06, + "loss": 0.5281, + "step": 4530 + }, + { + "epoch": 0.37554910899295485, + "grad_norm": 0.2551538348197937, + "learning_rate": 3.1226688769167013e-06, + "loss": 0.5583, + "step": 4531 + }, + { + "epoch": 0.3756319933692499, + "grad_norm": 0.259912371635437, + "learning_rate": 3.122254455035226e-06, + "loss": 0.55, + "step": 4532 + }, + { + "epoch": 0.37571487774554496, + "grad_norm": 0.2858012318611145, + "learning_rate": 3.1218400331537506e-06, + "loss": 0.5202, + "step": 4533 + }, + { + "epoch": 0.37579776212184, + "grad_norm": 0.2490331381559372, + "learning_rate": 3.121425611272275e-06, + "loss": 0.533, + "step": 4534 + }, + { + "epoch": 0.37588064649813513, + "grad_norm": 0.2495926320552826, + "learning_rate": 3.1210111893908003e-06, + "loss": 0.5274, + "step": 4535 + }, + { + "epoch": 0.3759635308744302, + "grad_norm": 0.2861316204071045, + "learning_rate": 3.1205967675093248e-06, + "loss": 0.575, + "step": 4536 + }, + { + "epoch": 0.37604641525072524, + "grad_norm": 0.2581165134906769, + "learning_rate": 3.1201823456278496e-06, + "loss": 0.5202, + "step": 4537 + }, + { + "epoch": 0.3761292996270203, + "grad_norm": 0.2528643310070038, + "learning_rate": 3.119767923746374e-06, + "loss": 0.5092, + "step": 4538 + }, + { + "epoch": 0.37621218400331535, + "grad_norm": 0.2524365186691284, + "learning_rate": 3.1193535018648985e-06, + "loss": 0.5018, + "step": 4539 + }, + { + "epoch": 0.37629506837961046, + "grad_norm": 0.2646353244781494, + "learning_rate": 3.1189390799834234e-06, + "loss": 0.5386, + "step": 4540 + }, + { + "epoch": 0.3763779527559055, + "grad_norm": 0.23931092023849487, + "learning_rate": 3.1185246581019478e-06, + "loss": 0.5099, + "step": 4541 + }, + { + "epoch": 0.3764608371322006, + "grad_norm": 0.25717872381210327, + "learning_rate": 3.118110236220473e-06, + "loss": 0.4814, + "step": 4542 + }, + { + "epoch": 0.37654372150849563, + "grad_norm": 0.2644842863082886, + "learning_rate": 3.1176958143389975e-06, + "loss": 0.5026, + "step": 4543 + }, + { + "epoch": 0.37662660588479074, + "grad_norm": 0.2616819143295288, + "learning_rate": 3.117281392457522e-06, + "loss": 0.5007, + "step": 4544 + }, + { + "epoch": 0.3767094902610858, + "grad_norm": 0.2625613808631897, + "learning_rate": 3.1168669705760468e-06, + "loss": 0.5477, + "step": 4545 + }, + { + "epoch": 0.37679237463738086, + "grad_norm": 0.2557584047317505, + "learning_rate": 3.1164525486945712e-06, + "loss": 0.537, + "step": 4546 + }, + { + "epoch": 0.3768752590136759, + "grad_norm": 0.24952451884746552, + "learning_rate": 3.116038126813096e-06, + "loss": 0.5458, + "step": 4547 + }, + { + "epoch": 0.37695814338997097, + "grad_norm": 0.2609616816043854, + "learning_rate": 3.1156237049316205e-06, + "loss": 0.5247, + "step": 4548 + }, + { + "epoch": 0.3770410277662661, + "grad_norm": 0.25998806953430176, + "learning_rate": 3.115209283050145e-06, + "loss": 0.5306, + "step": 4549 + }, + { + "epoch": 0.37712391214256114, + "grad_norm": 0.26083406805992126, + "learning_rate": 3.1147948611686702e-06, + "loss": 0.5278, + "step": 4550 + }, + { + "epoch": 0.3772067965188562, + "grad_norm": 0.2736765444278717, + "learning_rate": 3.1143804392871946e-06, + "loss": 0.5297, + "step": 4551 + }, + { + "epoch": 0.37728968089515125, + "grad_norm": 0.26354238390922546, + "learning_rate": 3.113966017405719e-06, + "loss": 0.4783, + "step": 4552 + }, + { + "epoch": 0.37737256527144636, + "grad_norm": 0.2325506955385208, + "learning_rate": 3.113551595524244e-06, + "loss": 0.5054, + "step": 4553 + }, + { + "epoch": 0.3774554496477414, + "grad_norm": 0.24924829602241516, + "learning_rate": 3.1131371736427684e-06, + "loss": 0.4798, + "step": 4554 + }, + { + "epoch": 0.37753833402403647, + "grad_norm": 0.29066452383995056, + "learning_rate": 3.1127227517612936e-06, + "loss": 0.5476, + "step": 4555 + }, + { + "epoch": 0.3776212184003315, + "grad_norm": 0.26729753613471985, + "learning_rate": 3.112308329879818e-06, + "loss": 0.5298, + "step": 4556 + }, + { + "epoch": 0.3777041027766266, + "grad_norm": 0.26867344975471497, + "learning_rate": 3.1118939079983425e-06, + "loss": 0.5515, + "step": 4557 + }, + { + "epoch": 0.3777869871529217, + "grad_norm": 0.25943827629089355, + "learning_rate": 3.1114794861168674e-06, + "loss": 0.5405, + "step": 4558 + }, + { + "epoch": 0.37786987152921675, + "grad_norm": 0.24777548015117645, + "learning_rate": 3.111065064235392e-06, + "loss": 0.4805, + "step": 4559 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 0.27212539315223694, + "learning_rate": 3.1106506423539166e-06, + "loss": 0.5473, + "step": 4560 + }, + { + "epoch": 0.37803564028180686, + "grad_norm": 0.2526148855686188, + "learning_rate": 3.110236220472441e-06, + "loss": 0.5409, + "step": 4561 + }, + { + "epoch": 0.378118524658102, + "grad_norm": 0.2459109127521515, + "learning_rate": 3.1098217985909655e-06, + "loss": 0.5124, + "step": 4562 + }, + { + "epoch": 0.37820140903439703, + "grad_norm": 0.2757963538169861, + "learning_rate": 3.109407376709491e-06, + "loss": 0.5318, + "step": 4563 + }, + { + "epoch": 0.3782842934106921, + "grad_norm": 0.2649947702884674, + "learning_rate": 3.1089929548280152e-06, + "loss": 0.5568, + "step": 4564 + }, + { + "epoch": 0.37836717778698714, + "grad_norm": 0.2718707025051117, + "learning_rate": 3.10857853294654e-06, + "loss": 0.5316, + "step": 4565 + }, + { + "epoch": 0.3784500621632822, + "grad_norm": 0.25952601432800293, + "learning_rate": 3.1081641110650645e-06, + "loss": 0.515, + "step": 4566 + }, + { + "epoch": 0.3785329465395773, + "grad_norm": 0.2809491455554962, + "learning_rate": 3.107749689183589e-06, + "loss": 0.5796, + "step": 4567 + }, + { + "epoch": 0.37861583091587236, + "grad_norm": 0.27911272644996643, + "learning_rate": 3.107335267302114e-06, + "loss": 0.5225, + "step": 4568 + }, + { + "epoch": 0.3786987152921674, + "grad_norm": 0.269662082195282, + "learning_rate": 3.1069208454206382e-06, + "loss": 0.5673, + "step": 4569 + }, + { + "epoch": 0.3787815996684625, + "grad_norm": 0.2593768835067749, + "learning_rate": 3.106506423539163e-06, + "loss": 0.5546, + "step": 4570 + }, + { + "epoch": 0.3788644840447576, + "grad_norm": 0.2751324474811554, + "learning_rate": 3.106092001657688e-06, + "loss": 0.5731, + "step": 4571 + }, + { + "epoch": 0.37894736842105264, + "grad_norm": 0.25198549032211304, + "learning_rate": 3.1056775797762124e-06, + "loss": 0.4981, + "step": 4572 + }, + { + "epoch": 0.3790302527973477, + "grad_norm": 0.26484909653663635, + "learning_rate": 3.1052631578947372e-06, + "loss": 0.5449, + "step": 4573 + }, + { + "epoch": 0.37911313717364276, + "grad_norm": 0.252000093460083, + "learning_rate": 3.1048487360132617e-06, + "loss": 0.5385, + "step": 4574 + }, + { + "epoch": 0.3791960215499378, + "grad_norm": 0.2918587327003479, + "learning_rate": 3.104434314131786e-06, + "loss": 0.5574, + "step": 4575 + }, + { + "epoch": 0.3792789059262329, + "grad_norm": 0.25700709223747253, + "learning_rate": 3.1040198922503114e-06, + "loss": 0.5197, + "step": 4576 + }, + { + "epoch": 0.379361790302528, + "grad_norm": 0.23559728264808655, + "learning_rate": 3.103605470368836e-06, + "loss": 0.4794, + "step": 4577 + }, + { + "epoch": 0.37944467467882304, + "grad_norm": 0.2369455248117447, + "learning_rate": 3.1031910484873607e-06, + "loss": 0.499, + "step": 4578 + }, + { + "epoch": 0.3795275590551181, + "grad_norm": 0.3634714186191559, + "learning_rate": 3.102776626605885e-06, + "loss": 0.5201, + "step": 4579 + }, + { + "epoch": 0.3796104434314132, + "grad_norm": 0.2576565444469452, + "learning_rate": 3.1023622047244095e-06, + "loss": 0.5092, + "step": 4580 + }, + { + "epoch": 0.37969332780770826, + "grad_norm": 0.2701682150363922, + "learning_rate": 3.1019477828429344e-06, + "loss": 0.5005, + "step": 4581 + }, + { + "epoch": 0.3797762121840033, + "grad_norm": 0.2274976372718811, + "learning_rate": 3.101533360961459e-06, + "loss": 0.4776, + "step": 4582 + }, + { + "epoch": 0.37985909656029837, + "grad_norm": 0.25866904854774475, + "learning_rate": 3.1011189390799832e-06, + "loss": 0.5049, + "step": 4583 + }, + { + "epoch": 0.3799419809365934, + "grad_norm": 0.2658323645591736, + "learning_rate": 3.1007045171985085e-06, + "loss": 0.5565, + "step": 4584 + }, + { + "epoch": 0.38002486531288854, + "grad_norm": 0.24814671277999878, + "learning_rate": 3.100290095317033e-06, + "loss": 0.5091, + "step": 4585 + }, + { + "epoch": 0.3801077496891836, + "grad_norm": 0.2728775143623352, + "learning_rate": 3.099875673435558e-06, + "loss": 0.5856, + "step": 4586 + }, + { + "epoch": 0.38019063406547865, + "grad_norm": 0.2410493642091751, + "learning_rate": 3.0994612515540822e-06, + "loss": 0.5269, + "step": 4587 + }, + { + "epoch": 0.3802735184417737, + "grad_norm": 0.25892841815948486, + "learning_rate": 3.0990468296726067e-06, + "loss": 0.5457, + "step": 4588 + }, + { + "epoch": 0.3803564028180688, + "grad_norm": 0.23548299074172974, + "learning_rate": 3.0986324077911315e-06, + "loss": 0.4803, + "step": 4589 + }, + { + "epoch": 0.3804392871943639, + "grad_norm": 0.49501049518585205, + "learning_rate": 3.0982179859096564e-06, + "loss": 0.5159, + "step": 4590 + }, + { + "epoch": 0.38052217157065893, + "grad_norm": 0.24208678305149078, + "learning_rate": 3.0978035640281812e-06, + "loss": 0.4632, + "step": 4591 + }, + { + "epoch": 0.380605055946954, + "grad_norm": 0.2522268295288086, + "learning_rate": 3.0973891421467057e-06, + "loss": 0.4815, + "step": 4592 + }, + { + "epoch": 0.38068794032324904, + "grad_norm": 0.29444754123687744, + "learning_rate": 3.09697472026523e-06, + "loss": 0.4877, + "step": 4593 + }, + { + "epoch": 0.38077082469954415, + "grad_norm": 0.27322399616241455, + "learning_rate": 3.096560298383755e-06, + "loss": 0.514, + "step": 4594 + }, + { + "epoch": 0.3808537090758392, + "grad_norm": 0.2570771872997284, + "learning_rate": 3.0961458765022794e-06, + "loss": 0.5367, + "step": 4595 + }, + { + "epoch": 0.38093659345213426, + "grad_norm": 0.2447233945131302, + "learning_rate": 3.0957314546208047e-06, + "loss": 0.512, + "step": 4596 + }, + { + "epoch": 0.3810194778284293, + "grad_norm": 0.23900291323661804, + "learning_rate": 3.095317032739329e-06, + "loss": 0.5062, + "step": 4597 + }, + { + "epoch": 0.38110236220472443, + "grad_norm": 0.482938289642334, + "learning_rate": 3.0949026108578535e-06, + "loss": 0.5509, + "step": 4598 + }, + { + "epoch": 0.3811852465810195, + "grad_norm": 0.2732755243778229, + "learning_rate": 3.0944881889763784e-06, + "loss": 0.519, + "step": 4599 + }, + { + "epoch": 0.38126813095731454, + "grad_norm": 0.2524604797363281, + "learning_rate": 3.094073767094903e-06, + "loss": 0.4889, + "step": 4600 + }, + { + "epoch": 0.3813510153336096, + "grad_norm": 0.3082844316959381, + "learning_rate": 3.0936593452134272e-06, + "loss": 0.5472, + "step": 4601 + }, + { + "epoch": 0.38143389970990466, + "grad_norm": 0.24717216193675995, + "learning_rate": 3.093244923331952e-06, + "loss": 0.5363, + "step": 4602 + }, + { + "epoch": 0.38151678408619977, + "grad_norm": 0.24311280250549316, + "learning_rate": 3.0928305014504765e-06, + "loss": 0.5152, + "step": 4603 + }, + { + "epoch": 0.3815996684624948, + "grad_norm": 0.2440834492444992, + "learning_rate": 3.092416079569002e-06, + "loss": 0.5197, + "step": 4604 + }, + { + "epoch": 0.3816825528387899, + "grad_norm": 0.24944280087947845, + "learning_rate": 3.0920016576875262e-06, + "loss": 0.511, + "step": 4605 + }, + { + "epoch": 0.38176543721508494, + "grad_norm": 0.25964608788490295, + "learning_rate": 3.0915872358060507e-06, + "loss": 0.5627, + "step": 4606 + }, + { + "epoch": 0.38184832159138005, + "grad_norm": 0.26549291610717773, + "learning_rate": 3.0911728139245755e-06, + "loss": 0.5092, + "step": 4607 + }, + { + "epoch": 0.3819312059676751, + "grad_norm": 0.26787081360816956, + "learning_rate": 3.0907583920431e-06, + "loss": 0.5723, + "step": 4608 + }, + { + "epoch": 0.38201409034397016, + "grad_norm": 0.2543451189994812, + "learning_rate": 3.090343970161625e-06, + "loss": 0.4789, + "step": 4609 + }, + { + "epoch": 0.3820969747202652, + "grad_norm": 0.2525535225868225, + "learning_rate": 3.0899295482801492e-06, + "loss": 0.5483, + "step": 4610 + }, + { + "epoch": 0.3821798590965603, + "grad_norm": 0.2672840654850006, + "learning_rate": 3.089515126398674e-06, + "loss": 0.5219, + "step": 4611 + }, + { + "epoch": 0.3822627434728554, + "grad_norm": 0.2596135437488556, + "learning_rate": 3.089100704517199e-06, + "loss": 0.4833, + "step": 4612 + }, + { + "epoch": 0.38234562784915044, + "grad_norm": 0.24315746128559113, + "learning_rate": 3.0886862826357234e-06, + "loss": 0.5105, + "step": 4613 + }, + { + "epoch": 0.3824285122254455, + "grad_norm": 0.2623690068721771, + "learning_rate": 3.0882718607542482e-06, + "loss": 0.5187, + "step": 4614 + }, + { + "epoch": 0.38251139660174055, + "grad_norm": 0.25698021054267883, + "learning_rate": 3.0878574388727727e-06, + "loss": 0.5042, + "step": 4615 + }, + { + "epoch": 0.38259428097803566, + "grad_norm": 0.24425721168518066, + "learning_rate": 3.087443016991297e-06, + "loss": 0.4655, + "step": 4616 + }, + { + "epoch": 0.3826771653543307, + "grad_norm": 0.23964878916740417, + "learning_rate": 3.0870285951098224e-06, + "loss": 0.511, + "step": 4617 + }, + { + "epoch": 0.3827600497306258, + "grad_norm": 0.25950852036476135, + "learning_rate": 3.086614173228347e-06, + "loss": 0.5142, + "step": 4618 + }, + { + "epoch": 0.38284293410692083, + "grad_norm": 0.2558008134365082, + "learning_rate": 3.0861997513468713e-06, + "loss": 0.4816, + "step": 4619 + }, + { + "epoch": 0.38292581848321594, + "grad_norm": 0.23815609514713287, + "learning_rate": 3.085785329465396e-06, + "loss": 0.5276, + "step": 4620 + }, + { + "epoch": 0.383008702859511, + "grad_norm": 0.23774565756320953, + "learning_rate": 3.0853709075839205e-06, + "loss": 0.4799, + "step": 4621 + }, + { + "epoch": 0.38309158723580605, + "grad_norm": 0.26111727952957153, + "learning_rate": 3.0849564857024454e-06, + "loss": 0.5331, + "step": 4622 + }, + { + "epoch": 0.3831744716121011, + "grad_norm": 0.2526915669441223, + "learning_rate": 3.08454206382097e-06, + "loss": 0.4948, + "step": 4623 + }, + { + "epoch": 0.38325735598839616, + "grad_norm": 0.25235286355018616, + "learning_rate": 3.0841276419394943e-06, + "loss": 0.513, + "step": 4624 + }, + { + "epoch": 0.3833402403646913, + "grad_norm": 0.2569885551929474, + "learning_rate": 3.0837132200580195e-06, + "loss": 0.5049, + "step": 4625 + }, + { + "epoch": 0.38342312474098633, + "grad_norm": 0.24992431700229645, + "learning_rate": 3.083298798176544e-06, + "loss": 0.4894, + "step": 4626 + }, + { + "epoch": 0.3835060091172814, + "grad_norm": 0.27062761783599854, + "learning_rate": 3.082884376295069e-06, + "loss": 0.5432, + "step": 4627 + }, + { + "epoch": 0.38358889349357644, + "grad_norm": 0.2539127767086029, + "learning_rate": 3.0824699544135933e-06, + "loss": 0.5008, + "step": 4628 + }, + { + "epoch": 0.38367177786987156, + "grad_norm": 0.2388140857219696, + "learning_rate": 3.0820555325321177e-06, + "loss": 0.5214, + "step": 4629 + }, + { + "epoch": 0.3837546622461666, + "grad_norm": 0.2622075378894806, + "learning_rate": 3.0816411106506425e-06, + "loss": 0.5403, + "step": 4630 + }, + { + "epoch": 0.38383754662246167, + "grad_norm": 0.24574868381023407, + "learning_rate": 3.0812266887691674e-06, + "loss": 0.4996, + "step": 4631 + }, + { + "epoch": 0.3839204309987567, + "grad_norm": 0.27419552206993103, + "learning_rate": 3.0808122668876923e-06, + "loss": 0.5079, + "step": 4632 + }, + { + "epoch": 0.3840033153750518, + "grad_norm": 0.24126461148262024, + "learning_rate": 3.0803978450062167e-06, + "loss": 0.5172, + "step": 4633 + }, + { + "epoch": 0.3840861997513469, + "grad_norm": 0.2500907778739929, + "learning_rate": 3.079983423124741e-06, + "loss": 0.5194, + "step": 4634 + }, + { + "epoch": 0.38416908412764195, + "grad_norm": 0.26558494567871094, + "learning_rate": 3.079569001243266e-06, + "loss": 0.4755, + "step": 4635 + }, + { + "epoch": 0.384251968503937, + "grad_norm": 0.25255757570266724, + "learning_rate": 3.0791545793617904e-06, + "loss": 0.5289, + "step": 4636 + }, + { + "epoch": 0.38433485288023206, + "grad_norm": 0.2434760481119156, + "learning_rate": 3.078740157480315e-06, + "loss": 0.4945, + "step": 4637 + }, + { + "epoch": 0.38441773725652717, + "grad_norm": 0.25033822655677795, + "learning_rate": 3.07832573559884e-06, + "loss": 0.5074, + "step": 4638 + }, + { + "epoch": 0.3845006216328222, + "grad_norm": 0.23619793355464935, + "learning_rate": 3.0779113137173645e-06, + "loss": 0.5003, + "step": 4639 + }, + { + "epoch": 0.3845835060091173, + "grad_norm": 0.2662160098552704, + "learning_rate": 3.0774968918358894e-06, + "loss": 0.5278, + "step": 4640 + }, + { + "epoch": 0.38466639038541234, + "grad_norm": 0.2697000503540039, + "learning_rate": 3.077082469954414e-06, + "loss": 0.5235, + "step": 4641 + }, + { + "epoch": 0.3847492747617074, + "grad_norm": 0.25766757130622864, + "learning_rate": 3.0766680480729383e-06, + "loss": 0.5518, + "step": 4642 + }, + { + "epoch": 0.3848321591380025, + "grad_norm": 0.2606932818889618, + "learning_rate": 3.076253626191463e-06, + "loss": 0.5335, + "step": 4643 + }, + { + "epoch": 0.38491504351429756, + "grad_norm": 0.24910302460193634, + "learning_rate": 3.0758392043099876e-06, + "loss": 0.504, + "step": 4644 + }, + { + "epoch": 0.3849979278905926, + "grad_norm": 0.25171345472335815, + "learning_rate": 3.075424782428513e-06, + "loss": 0.5318, + "step": 4645 + }, + { + "epoch": 0.3850808122668877, + "grad_norm": 0.25208669900894165, + "learning_rate": 3.0750103605470373e-06, + "loss": 0.5517, + "step": 4646 + }, + { + "epoch": 0.3851636966431828, + "grad_norm": 0.26102936267852783, + "learning_rate": 3.0745959386655617e-06, + "loss": 0.5467, + "step": 4647 + }, + { + "epoch": 0.38524658101947784, + "grad_norm": 0.27869048714637756, + "learning_rate": 3.0741815167840866e-06, + "loss": 0.5358, + "step": 4648 + }, + { + "epoch": 0.3853294653957729, + "grad_norm": 0.2631678581237793, + "learning_rate": 3.073767094902611e-06, + "loss": 0.5244, + "step": 4649 + }, + { + "epoch": 0.38541234977206795, + "grad_norm": 0.24768655002117157, + "learning_rate": 3.0733526730211354e-06, + "loss": 0.4968, + "step": 4650 + }, + { + "epoch": 0.385495234148363, + "grad_norm": 0.2508442997932434, + "learning_rate": 3.0729382511396607e-06, + "loss": 0.5359, + "step": 4651 + }, + { + "epoch": 0.3855781185246581, + "grad_norm": 0.25381356477737427, + "learning_rate": 3.072523829258185e-06, + "loss": 0.5166, + "step": 4652 + }, + { + "epoch": 0.3856610029009532, + "grad_norm": 0.2810617983341217, + "learning_rate": 3.07210940737671e-06, + "loss": 0.5372, + "step": 4653 + }, + { + "epoch": 0.38574388727724823, + "grad_norm": 0.2409786731004715, + "learning_rate": 3.0716949854952344e-06, + "loss": 0.5162, + "step": 4654 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 0.24744874238967896, + "learning_rate": 3.071280563613759e-06, + "loss": 0.477, + "step": 4655 + }, + { + "epoch": 0.3859096560298384, + "grad_norm": 0.2630404531955719, + "learning_rate": 3.0708661417322837e-06, + "loss": 0.5082, + "step": 4656 + }, + { + "epoch": 0.38599254040613346, + "grad_norm": 0.2653011977672577, + "learning_rate": 3.070451719850808e-06, + "loss": 0.5258, + "step": 4657 + }, + { + "epoch": 0.3860754247824285, + "grad_norm": 0.2657395303249359, + "learning_rate": 3.0700372979693334e-06, + "loss": 0.4826, + "step": 4658 + }, + { + "epoch": 0.38615830915872357, + "grad_norm": 0.26025107502937317, + "learning_rate": 3.069622876087858e-06, + "loss": 0.5383, + "step": 4659 + }, + { + "epoch": 0.3862411935350186, + "grad_norm": 0.26165610551834106, + "learning_rate": 3.0692084542063823e-06, + "loss": 0.5204, + "step": 4660 + }, + { + "epoch": 0.38632407791131373, + "grad_norm": 0.2506769895553589, + "learning_rate": 3.068794032324907e-06, + "loss": 0.5081, + "step": 4661 + }, + { + "epoch": 0.3864069622876088, + "grad_norm": 0.2410447746515274, + "learning_rate": 3.0683796104434316e-06, + "loss": 0.4941, + "step": 4662 + }, + { + "epoch": 0.38648984666390385, + "grad_norm": 0.25073859095573425, + "learning_rate": 3.0679651885619564e-06, + "loss": 0.4997, + "step": 4663 + }, + { + "epoch": 0.3865727310401989, + "grad_norm": 0.2273397445678711, + "learning_rate": 3.067550766680481e-06, + "loss": 0.4561, + "step": 4664 + }, + { + "epoch": 0.386655615416494, + "grad_norm": 0.24208934605121613, + "learning_rate": 3.0671363447990053e-06, + "loss": 0.4948, + "step": 4665 + }, + { + "epoch": 0.38673849979278907, + "grad_norm": 0.2716154158115387, + "learning_rate": 3.0667219229175306e-06, + "loss": 0.5526, + "step": 4666 + }, + { + "epoch": 0.3868213841690841, + "grad_norm": 0.24581989645957947, + "learning_rate": 3.066307501036055e-06, + "loss": 0.4903, + "step": 4667 + }, + { + "epoch": 0.3869042685453792, + "grad_norm": 0.24856115877628326, + "learning_rate": 3.0658930791545794e-06, + "loss": 0.5311, + "step": 4668 + }, + { + "epoch": 0.38698715292167424, + "grad_norm": 0.2460024654865265, + "learning_rate": 3.0654786572731043e-06, + "loss": 0.5438, + "step": 4669 + }, + { + "epoch": 0.38707003729796935, + "grad_norm": 0.279391884803772, + "learning_rate": 3.0650642353916287e-06, + "loss": 0.5481, + "step": 4670 + }, + { + "epoch": 0.3871529216742644, + "grad_norm": 0.25801557302474976, + "learning_rate": 3.064649813510154e-06, + "loss": 0.464, + "step": 4671 + }, + { + "epoch": 0.38723580605055946, + "grad_norm": 0.2553400695323944, + "learning_rate": 3.0642353916286784e-06, + "loss": 0.5212, + "step": 4672 + }, + { + "epoch": 0.3873186904268545, + "grad_norm": 0.243752121925354, + "learning_rate": 3.063820969747203e-06, + "loss": 0.4623, + "step": 4673 + }, + { + "epoch": 0.38740157480314963, + "grad_norm": 0.2477865368127823, + "learning_rate": 3.0634065478657277e-06, + "loss": 0.5159, + "step": 4674 + }, + { + "epoch": 0.3874844591794447, + "grad_norm": 0.2722074091434479, + "learning_rate": 3.062992125984252e-06, + "loss": 0.5532, + "step": 4675 + }, + { + "epoch": 0.38756734355573974, + "grad_norm": 0.2771785259246826, + "learning_rate": 3.062577704102777e-06, + "loss": 0.5483, + "step": 4676 + }, + { + "epoch": 0.3876502279320348, + "grad_norm": 0.2733108401298523, + "learning_rate": 3.0621632822213014e-06, + "loss": 0.5831, + "step": 4677 + }, + { + "epoch": 0.38773311230832985, + "grad_norm": 0.2911473512649536, + "learning_rate": 3.061748860339826e-06, + "loss": 0.5575, + "step": 4678 + }, + { + "epoch": 0.38781599668462496, + "grad_norm": 0.2408820539712906, + "learning_rate": 3.061334438458351e-06, + "loss": 0.479, + "step": 4679 + }, + { + "epoch": 0.38789888106092, + "grad_norm": 0.25640571117401123, + "learning_rate": 3.0609200165768756e-06, + "loss": 0.5281, + "step": 4680 + }, + { + "epoch": 0.3879817654372151, + "grad_norm": 0.26550424098968506, + "learning_rate": 3.0605055946954004e-06, + "loss": 0.4906, + "step": 4681 + }, + { + "epoch": 0.38806464981351013, + "grad_norm": 0.24590250849723816, + "learning_rate": 3.060091172813925e-06, + "loss": 0.5308, + "step": 4682 + }, + { + "epoch": 0.38814753418980524, + "grad_norm": 0.25367501378059387, + "learning_rate": 3.0596767509324493e-06, + "loss": 0.5183, + "step": 4683 + }, + { + "epoch": 0.3882304185661003, + "grad_norm": 0.2709812819957733, + "learning_rate": 3.059262329050974e-06, + "loss": 0.5372, + "step": 4684 + }, + { + "epoch": 0.38831330294239536, + "grad_norm": 0.25857850909233093, + "learning_rate": 3.0588479071694986e-06, + "loss": 0.5339, + "step": 4685 + }, + { + "epoch": 0.3883961873186904, + "grad_norm": 0.2562395930290222, + "learning_rate": 3.058433485288023e-06, + "loss": 0.5687, + "step": 4686 + }, + { + "epoch": 0.3884790716949855, + "grad_norm": 0.2535606920719147, + "learning_rate": 3.0580190634065483e-06, + "loss": 0.5338, + "step": 4687 + }, + { + "epoch": 0.3885619560712806, + "grad_norm": 0.2675091028213501, + "learning_rate": 3.0576046415250727e-06, + "loss": 0.5558, + "step": 4688 + }, + { + "epoch": 0.38864484044757563, + "grad_norm": 0.22718113660812378, + "learning_rate": 3.0571902196435976e-06, + "loss": 0.4685, + "step": 4689 + }, + { + "epoch": 0.3887277248238707, + "grad_norm": 0.2552345097064972, + "learning_rate": 3.056775797762122e-06, + "loss": 0.5043, + "step": 4690 + }, + { + "epoch": 0.38881060920016575, + "grad_norm": 0.24688664078712463, + "learning_rate": 3.0563613758806464e-06, + "loss": 0.5144, + "step": 4691 + }, + { + "epoch": 0.38889349357646086, + "grad_norm": 0.26272475719451904, + "learning_rate": 3.0559469539991717e-06, + "loss": 0.49, + "step": 4692 + }, + { + "epoch": 0.3889763779527559, + "grad_norm": 0.25970765948295593, + "learning_rate": 3.055532532117696e-06, + "loss": 0.5268, + "step": 4693 + }, + { + "epoch": 0.38905926232905097, + "grad_norm": 0.2621018588542938, + "learning_rate": 3.055118110236221e-06, + "loss": 0.5596, + "step": 4694 + }, + { + "epoch": 0.389142146705346, + "grad_norm": 0.24186640977859497, + "learning_rate": 3.0547036883547454e-06, + "loss": 0.5099, + "step": 4695 + }, + { + "epoch": 0.38922503108164114, + "grad_norm": 0.2720845937728882, + "learning_rate": 3.05428926647327e-06, + "loss": 0.5339, + "step": 4696 + }, + { + "epoch": 0.3893079154579362, + "grad_norm": 0.2856716513633728, + "learning_rate": 3.0538748445917947e-06, + "loss": 0.535, + "step": 4697 + }, + { + "epoch": 0.38939079983423125, + "grad_norm": 0.27772772312164307, + "learning_rate": 3.053460422710319e-06, + "loss": 0.5518, + "step": 4698 + }, + { + "epoch": 0.3894736842105263, + "grad_norm": 0.25690171122550964, + "learning_rate": 3.0530460008288444e-06, + "loss": 0.4957, + "step": 4699 + }, + { + "epoch": 0.38955656858682136, + "grad_norm": 0.2619157135486603, + "learning_rate": 3.052631578947369e-06, + "loss": 0.4805, + "step": 4700 + }, + { + "epoch": 0.3896394529631165, + "grad_norm": 0.27451273798942566, + "learning_rate": 3.0522171570658933e-06, + "loss": 0.5255, + "step": 4701 + }, + { + "epoch": 0.38972233733941153, + "grad_norm": 0.2758919894695282, + "learning_rate": 3.051802735184418e-06, + "loss": 0.5583, + "step": 4702 + }, + { + "epoch": 0.3898052217157066, + "grad_norm": 0.23795291781425476, + "learning_rate": 3.0513883133029426e-06, + "loss": 0.5126, + "step": 4703 + }, + { + "epoch": 0.38988810609200164, + "grad_norm": 0.28131794929504395, + "learning_rate": 3.050973891421467e-06, + "loss": 0.5654, + "step": 4704 + }, + { + "epoch": 0.38997099046829675, + "grad_norm": 0.2550713121891022, + "learning_rate": 3.050559469539992e-06, + "loss": 0.4948, + "step": 4705 + }, + { + "epoch": 0.3900538748445918, + "grad_norm": 0.25033456087112427, + "learning_rate": 3.0501450476585163e-06, + "loss": 0.4995, + "step": 4706 + }, + { + "epoch": 0.39013675922088686, + "grad_norm": 0.28303441405296326, + "learning_rate": 3.0497306257770416e-06, + "loss": 0.5288, + "step": 4707 + }, + { + "epoch": 0.3902196435971819, + "grad_norm": 0.2539862394332886, + "learning_rate": 3.049316203895566e-06, + "loss": 0.4998, + "step": 4708 + }, + { + "epoch": 0.390302527973477, + "grad_norm": 0.2665258049964905, + "learning_rate": 3.0489017820140904e-06, + "loss": 0.5968, + "step": 4709 + }, + { + "epoch": 0.3903854123497721, + "grad_norm": 0.28557583689689636, + "learning_rate": 3.0484873601326153e-06, + "loss": 0.5675, + "step": 4710 + }, + { + "epoch": 0.39046829672606714, + "grad_norm": 0.2571103274822235, + "learning_rate": 3.0480729382511397e-06, + "loss": 0.5097, + "step": 4711 + }, + { + "epoch": 0.3905511811023622, + "grad_norm": 0.2470082938671112, + "learning_rate": 3.047658516369665e-06, + "loss": 0.4837, + "step": 4712 + }, + { + "epoch": 0.39063406547865726, + "grad_norm": 0.29212307929992676, + "learning_rate": 3.0472440944881894e-06, + "loss": 0.5505, + "step": 4713 + }, + { + "epoch": 0.39071694985495237, + "grad_norm": 0.25990521907806396, + "learning_rate": 3.046829672606714e-06, + "loss": 0.5146, + "step": 4714 + }, + { + "epoch": 0.3907998342312474, + "grad_norm": 0.24910524487495422, + "learning_rate": 3.0464152507252387e-06, + "loss": 0.5469, + "step": 4715 + }, + { + "epoch": 0.3908827186075425, + "grad_norm": 0.2880533039569855, + "learning_rate": 3.046000828843763e-06, + "loss": 0.5645, + "step": 4716 + }, + { + "epoch": 0.39096560298383753, + "grad_norm": 0.2687453031539917, + "learning_rate": 3.045586406962288e-06, + "loss": 0.4952, + "step": 4717 + }, + { + "epoch": 0.3910484873601326, + "grad_norm": 0.2403765767812729, + "learning_rate": 3.0451719850808124e-06, + "loss": 0.5328, + "step": 4718 + }, + { + "epoch": 0.3911313717364277, + "grad_norm": 0.2733615040779114, + "learning_rate": 3.044757563199337e-06, + "loss": 0.5245, + "step": 4719 + }, + { + "epoch": 0.39121425611272276, + "grad_norm": 0.26025888323783875, + "learning_rate": 3.044343141317862e-06, + "loss": 0.4906, + "step": 4720 + }, + { + "epoch": 0.3912971404890178, + "grad_norm": 0.2553655803203583, + "learning_rate": 3.0439287194363866e-06, + "loss": 0.5444, + "step": 4721 + }, + { + "epoch": 0.39138002486531287, + "grad_norm": 0.2811269760131836, + "learning_rate": 3.043514297554911e-06, + "loss": 0.5728, + "step": 4722 + }, + { + "epoch": 0.391462909241608, + "grad_norm": 0.256359338760376, + "learning_rate": 3.043099875673436e-06, + "loss": 0.5189, + "step": 4723 + }, + { + "epoch": 0.39154579361790304, + "grad_norm": 0.2626519501209259, + "learning_rate": 3.0426854537919603e-06, + "loss": 0.4817, + "step": 4724 + }, + { + "epoch": 0.3916286779941981, + "grad_norm": 0.26654666662216187, + "learning_rate": 3.042271031910485e-06, + "loss": 0.5158, + "step": 4725 + }, + { + "epoch": 0.39171156237049315, + "grad_norm": 0.2731899321079254, + "learning_rate": 3.0418566100290096e-06, + "loss": 0.5182, + "step": 4726 + }, + { + "epoch": 0.3917944467467882, + "grad_norm": 0.26410338282585144, + "learning_rate": 3.041442188147534e-06, + "loss": 0.5168, + "step": 4727 + }, + { + "epoch": 0.3918773311230833, + "grad_norm": 0.2543181777000427, + "learning_rate": 3.0410277662660593e-06, + "loss": 0.536, + "step": 4728 + }, + { + "epoch": 0.3919602154993784, + "grad_norm": 0.24694564938545227, + "learning_rate": 3.0406133443845837e-06, + "loss": 0.5012, + "step": 4729 + }, + { + "epoch": 0.39204309987567343, + "grad_norm": 0.2415819615125656, + "learning_rate": 3.0401989225031086e-06, + "loss": 0.4795, + "step": 4730 + }, + { + "epoch": 0.3921259842519685, + "grad_norm": 0.2660196125507355, + "learning_rate": 3.039784500621633e-06, + "loss": 0.5528, + "step": 4731 + }, + { + "epoch": 0.3922088686282636, + "grad_norm": 0.26324400305747986, + "learning_rate": 3.0393700787401575e-06, + "loss": 0.5607, + "step": 4732 + }, + { + "epoch": 0.39229175300455865, + "grad_norm": 0.25594890117645264, + "learning_rate": 3.0389556568586827e-06, + "loss": 0.4976, + "step": 4733 + }, + { + "epoch": 0.3923746373808537, + "grad_norm": 0.23732627928256989, + "learning_rate": 3.038541234977207e-06, + "loss": 0.5443, + "step": 4734 + }, + { + "epoch": 0.39245752175714876, + "grad_norm": 0.2276262491941452, + "learning_rate": 3.0381268130957316e-06, + "loss": 0.4711, + "step": 4735 + }, + { + "epoch": 0.3925404061334438, + "grad_norm": 0.24897727370262146, + "learning_rate": 3.0377123912142565e-06, + "loss": 0.5201, + "step": 4736 + }, + { + "epoch": 0.39262329050973893, + "grad_norm": 0.2705540955066681, + "learning_rate": 3.037297969332781e-06, + "loss": 0.5755, + "step": 4737 + }, + { + "epoch": 0.392706174886034, + "grad_norm": 0.24753688275814056, + "learning_rate": 3.0368835474513057e-06, + "loss": 0.496, + "step": 4738 + }, + { + "epoch": 0.39278905926232904, + "grad_norm": 0.23240144550800323, + "learning_rate": 3.03646912556983e-06, + "loss": 0.4552, + "step": 4739 + }, + { + "epoch": 0.3928719436386241, + "grad_norm": 0.2585722506046295, + "learning_rate": 3.0360547036883546e-06, + "loss": 0.5209, + "step": 4740 + }, + { + "epoch": 0.3929548280149192, + "grad_norm": 0.27161461114883423, + "learning_rate": 3.03564028180688e-06, + "loss": 0.5278, + "step": 4741 + }, + { + "epoch": 0.39303771239121427, + "grad_norm": 0.2632003724575043, + "learning_rate": 3.0352258599254043e-06, + "loss": 0.5386, + "step": 4742 + }, + { + "epoch": 0.3931205967675093, + "grad_norm": 0.24082247912883759, + "learning_rate": 3.034811438043929e-06, + "loss": 0.5378, + "step": 4743 + }, + { + "epoch": 0.3932034811438044, + "grad_norm": 0.2744540870189667, + "learning_rate": 3.0343970161624536e-06, + "loss": 0.5799, + "step": 4744 + }, + { + "epoch": 0.39328636552009943, + "grad_norm": 0.24748241901397705, + "learning_rate": 3.033982594280978e-06, + "loss": 0.516, + "step": 4745 + }, + { + "epoch": 0.39336924989639455, + "grad_norm": 0.28285834193229675, + "learning_rate": 3.033568172399503e-06, + "loss": 0.5378, + "step": 4746 + }, + { + "epoch": 0.3934521342726896, + "grad_norm": 0.24257081747055054, + "learning_rate": 3.0331537505180273e-06, + "loss": 0.4943, + "step": 4747 + }, + { + "epoch": 0.39353501864898466, + "grad_norm": 0.2400343120098114, + "learning_rate": 3.0327393286365526e-06, + "loss": 0.5204, + "step": 4748 + }, + { + "epoch": 0.3936179030252797, + "grad_norm": 0.26041096448898315, + "learning_rate": 3.032324906755077e-06, + "loss": 0.5609, + "step": 4749 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 0.2379569709300995, + "learning_rate": 3.0319104848736015e-06, + "loss": 0.4924, + "step": 4750 + }, + { + "epoch": 0.3937836717778699, + "grad_norm": 0.25557148456573486, + "learning_rate": 3.0314960629921263e-06, + "loss": 0.5274, + "step": 4751 + }, + { + "epoch": 0.39386655615416494, + "grad_norm": 0.2512422204017639, + "learning_rate": 3.0310816411106508e-06, + "loss": 0.5049, + "step": 4752 + }, + { + "epoch": 0.39394944053046, + "grad_norm": 0.27175992727279663, + "learning_rate": 3.030667219229175e-06, + "loss": 0.5523, + "step": 4753 + }, + { + "epoch": 0.39403232490675505, + "grad_norm": 0.24837632477283478, + "learning_rate": 3.0302527973477005e-06, + "loss": 0.5142, + "step": 4754 + }, + { + "epoch": 0.39411520928305016, + "grad_norm": 0.2709707021713257, + "learning_rate": 3.029838375466225e-06, + "loss": 0.5122, + "step": 4755 + }, + { + "epoch": 0.3941980936593452, + "grad_norm": 0.29032275080680847, + "learning_rate": 3.0294239535847497e-06, + "loss": 0.529, + "step": 4756 + }, + { + "epoch": 0.3942809780356403, + "grad_norm": 0.25926247239112854, + "learning_rate": 3.029009531703274e-06, + "loss": 0.5384, + "step": 4757 + }, + { + "epoch": 0.39436386241193533, + "grad_norm": 0.2473958283662796, + "learning_rate": 3.0285951098217986e-06, + "loss": 0.4839, + "step": 4758 + }, + { + "epoch": 0.39444674678823044, + "grad_norm": 0.27929309010505676, + "learning_rate": 3.0281806879403235e-06, + "loss": 0.5883, + "step": 4759 + }, + { + "epoch": 0.3945296311645255, + "grad_norm": 0.2471829354763031, + "learning_rate": 3.027766266058848e-06, + "loss": 0.4932, + "step": 4760 + }, + { + "epoch": 0.39461251554082055, + "grad_norm": 0.24754919111728668, + "learning_rate": 3.027351844177373e-06, + "loss": 0.5327, + "step": 4761 + }, + { + "epoch": 0.3946953999171156, + "grad_norm": 0.2789086401462555, + "learning_rate": 3.0269374222958976e-06, + "loss": 0.5171, + "step": 4762 + }, + { + "epoch": 0.3947782842934107, + "grad_norm": 0.2802376449108124, + "learning_rate": 3.026523000414422e-06, + "loss": 0.5553, + "step": 4763 + }, + { + "epoch": 0.3948611686697058, + "grad_norm": 0.2519262135028839, + "learning_rate": 3.026108578532947e-06, + "loss": 0.516, + "step": 4764 + }, + { + "epoch": 0.39494405304600083, + "grad_norm": 0.25732359290122986, + "learning_rate": 3.0256941566514713e-06, + "loss": 0.4852, + "step": 4765 + }, + { + "epoch": 0.3950269374222959, + "grad_norm": 0.25794628262519836, + "learning_rate": 3.025279734769996e-06, + "loss": 0.5196, + "step": 4766 + }, + { + "epoch": 0.39510982179859094, + "grad_norm": 0.2880038022994995, + "learning_rate": 3.0248653128885206e-06, + "loss": 0.5669, + "step": 4767 + }, + { + "epoch": 0.39519270617488605, + "grad_norm": 0.25644567608833313, + "learning_rate": 3.024450891007045e-06, + "loss": 0.4967, + "step": 4768 + }, + { + "epoch": 0.3952755905511811, + "grad_norm": 0.2535068988800049, + "learning_rate": 3.0240364691255703e-06, + "loss": 0.5172, + "step": 4769 + }, + { + "epoch": 0.39535847492747617, + "grad_norm": 0.26775923371315, + "learning_rate": 3.0236220472440948e-06, + "loss": 0.574, + "step": 4770 + }, + { + "epoch": 0.3954413593037712, + "grad_norm": 0.29602575302124023, + "learning_rate": 3.023207625362619e-06, + "loss": 0.5109, + "step": 4771 + }, + { + "epoch": 0.39552424368006633, + "grad_norm": 0.24854572117328644, + "learning_rate": 3.022793203481144e-06, + "loss": 0.5015, + "step": 4772 + }, + { + "epoch": 0.3956071280563614, + "grad_norm": 0.2622416913509369, + "learning_rate": 3.0223787815996685e-06, + "loss": 0.527, + "step": 4773 + }, + { + "epoch": 0.39569001243265645, + "grad_norm": 0.25200459361076355, + "learning_rate": 3.0219643597181938e-06, + "loss": 0.5303, + "step": 4774 + }, + { + "epoch": 0.3957728968089515, + "grad_norm": 0.24779796600341797, + "learning_rate": 3.021549937836718e-06, + "loss": 0.5181, + "step": 4775 + }, + { + "epoch": 0.39585578118524656, + "grad_norm": 0.2683712840080261, + "learning_rate": 3.0211355159552426e-06, + "loss": 0.5149, + "step": 4776 + }, + { + "epoch": 0.39593866556154167, + "grad_norm": 0.2517310380935669, + "learning_rate": 3.0207210940737675e-06, + "loss": 0.5374, + "step": 4777 + }, + { + "epoch": 0.3960215499378367, + "grad_norm": 0.2521214783191681, + "learning_rate": 3.020306672192292e-06, + "loss": 0.5397, + "step": 4778 + }, + { + "epoch": 0.3961044343141318, + "grad_norm": 0.24985240399837494, + "learning_rate": 3.0198922503108168e-06, + "loss": 0.4944, + "step": 4779 + }, + { + "epoch": 0.39618731869042684, + "grad_norm": 0.23953130841255188, + "learning_rate": 3.019477828429341e-06, + "loss": 0.5236, + "step": 4780 + }, + { + "epoch": 0.39627020306672195, + "grad_norm": 0.2594156861305237, + "learning_rate": 3.0190634065478656e-06, + "loss": 0.546, + "step": 4781 + }, + { + "epoch": 0.396353087443017, + "grad_norm": 0.26049429178237915, + "learning_rate": 3.018648984666391e-06, + "loss": 0.5247, + "step": 4782 + }, + { + "epoch": 0.39643597181931206, + "grad_norm": 0.26313143968582153, + "learning_rate": 3.0182345627849153e-06, + "loss": 0.5207, + "step": 4783 + }, + { + "epoch": 0.3965188561956071, + "grad_norm": 0.2513478696346283, + "learning_rate": 3.01782014090344e-06, + "loss": 0.5128, + "step": 4784 + }, + { + "epoch": 0.3966017405719022, + "grad_norm": 0.24686910212039948, + "learning_rate": 3.0174057190219646e-06, + "loss": 0.4683, + "step": 4785 + }, + { + "epoch": 0.3966846249481973, + "grad_norm": 0.2674047350883484, + "learning_rate": 3.016991297140489e-06, + "loss": 0.5302, + "step": 4786 + }, + { + "epoch": 0.39676750932449234, + "grad_norm": 0.23889605700969696, + "learning_rate": 3.016576875259014e-06, + "loss": 0.4604, + "step": 4787 + }, + { + "epoch": 0.3968503937007874, + "grad_norm": 0.26108258962631226, + "learning_rate": 3.0161624533775383e-06, + "loss": 0.5407, + "step": 4788 + }, + { + "epoch": 0.39693327807708245, + "grad_norm": 0.23729915916919708, + "learning_rate": 3.015748031496063e-06, + "loss": 0.4803, + "step": 4789 + }, + { + "epoch": 0.39701616245337756, + "grad_norm": 0.2533389627933502, + "learning_rate": 3.015333609614588e-06, + "loss": 0.5497, + "step": 4790 + }, + { + "epoch": 0.3970990468296726, + "grad_norm": 0.253523051738739, + "learning_rate": 3.0149191877331125e-06, + "loss": 0.52, + "step": 4791 + }, + { + "epoch": 0.3971819312059677, + "grad_norm": 0.26345184445381165, + "learning_rate": 3.0145047658516373e-06, + "loss": 0.5417, + "step": 4792 + }, + { + "epoch": 0.39726481558226273, + "grad_norm": 0.25879353284835815, + "learning_rate": 3.0140903439701618e-06, + "loss": 0.5086, + "step": 4793 + }, + { + "epoch": 0.3973476999585578, + "grad_norm": 0.2643851935863495, + "learning_rate": 3.013675922088686e-06, + "loss": 0.5259, + "step": 4794 + }, + { + "epoch": 0.3974305843348529, + "grad_norm": 0.2743457555770874, + "learning_rate": 3.0132615002072115e-06, + "loss": 0.533, + "step": 4795 + }, + { + "epoch": 0.39751346871114795, + "grad_norm": 0.2579635679721832, + "learning_rate": 3.012847078325736e-06, + "loss": 0.5242, + "step": 4796 + }, + { + "epoch": 0.397596353087443, + "grad_norm": 0.2557319402694702, + "learning_rate": 3.0124326564442608e-06, + "loss": 0.5135, + "step": 4797 + }, + { + "epoch": 0.39767923746373807, + "grad_norm": 0.2556900978088379, + "learning_rate": 3.012018234562785e-06, + "loss": 0.5118, + "step": 4798 + }, + { + "epoch": 0.3977621218400332, + "grad_norm": 0.26178744435310364, + "learning_rate": 3.0116038126813096e-06, + "loss": 0.5453, + "step": 4799 + }, + { + "epoch": 0.39784500621632823, + "grad_norm": 0.2508390545845032, + "learning_rate": 3.0111893907998345e-06, + "loss": 0.5035, + "step": 4800 + }, + { + "epoch": 0.3979278905926233, + "grad_norm": 0.26729151606559753, + "learning_rate": 3.010774968918359e-06, + "loss": 0.5078, + "step": 4801 + }, + { + "epoch": 0.39801077496891835, + "grad_norm": 0.2597721517086029, + "learning_rate": 3.0103605470368834e-06, + "loss": 0.5266, + "step": 4802 + }, + { + "epoch": 0.3980936593452134, + "grad_norm": 0.2300974428653717, + "learning_rate": 3.0099461251554086e-06, + "loss": 0.4731, + "step": 4803 + }, + { + "epoch": 0.3981765437215085, + "grad_norm": 0.24744825065135956, + "learning_rate": 3.009531703273933e-06, + "loss": 0.4724, + "step": 4804 + }, + { + "epoch": 0.39825942809780357, + "grad_norm": 0.24575874209403992, + "learning_rate": 3.009117281392458e-06, + "loss": 0.5146, + "step": 4805 + }, + { + "epoch": 0.3983423124740986, + "grad_norm": 0.24327732622623444, + "learning_rate": 3.0087028595109823e-06, + "loss": 0.5068, + "step": 4806 + }, + { + "epoch": 0.3984251968503937, + "grad_norm": 0.23466403782367706, + "learning_rate": 3.0082884376295068e-06, + "loss": 0.4701, + "step": 4807 + }, + { + "epoch": 0.3985080812266888, + "grad_norm": 0.24918293952941895, + "learning_rate": 3.0078740157480316e-06, + "loss": 0.4797, + "step": 4808 + }, + { + "epoch": 0.39859096560298385, + "grad_norm": 0.24187195301055908, + "learning_rate": 3.0074595938665565e-06, + "loss": 0.5013, + "step": 4809 + }, + { + "epoch": 0.3986738499792789, + "grad_norm": 0.24905522167682648, + "learning_rate": 3.0070451719850813e-06, + "loss": 0.4667, + "step": 4810 + }, + { + "epoch": 0.39875673435557396, + "grad_norm": 0.26336726546287537, + "learning_rate": 3.0066307501036058e-06, + "loss": 0.5572, + "step": 4811 + }, + { + "epoch": 0.398839618731869, + "grad_norm": 0.23085060715675354, + "learning_rate": 3.00621632822213e-06, + "loss": 0.491, + "step": 4812 + }, + { + "epoch": 0.39892250310816413, + "grad_norm": 0.27075520157814026, + "learning_rate": 3.005801906340655e-06, + "loss": 0.5464, + "step": 4813 + }, + { + "epoch": 0.3990053874844592, + "grad_norm": 0.27155858278274536, + "learning_rate": 3.0053874844591795e-06, + "loss": 0.539, + "step": 4814 + }, + { + "epoch": 0.39908827186075424, + "grad_norm": 0.277916818857193, + "learning_rate": 3.0049730625777048e-06, + "loss": 0.5443, + "step": 4815 + }, + { + "epoch": 0.3991711562370493, + "grad_norm": 0.25960108637809753, + "learning_rate": 3.004558640696229e-06, + "loss": 0.5071, + "step": 4816 + }, + { + "epoch": 0.3992540406133444, + "grad_norm": 0.25640520453453064, + "learning_rate": 3.0041442188147536e-06, + "loss": 0.562, + "step": 4817 + }, + { + "epoch": 0.39933692498963946, + "grad_norm": 0.2540709674358368, + "learning_rate": 3.0037297969332785e-06, + "loss": 0.5296, + "step": 4818 + }, + { + "epoch": 0.3994198093659345, + "grad_norm": 0.23236526548862457, + "learning_rate": 3.003315375051803e-06, + "loss": 0.4643, + "step": 4819 + }, + { + "epoch": 0.3995026937422296, + "grad_norm": 0.27223071455955505, + "learning_rate": 3.0029009531703274e-06, + "loss": 0.5295, + "step": 4820 + }, + { + "epoch": 0.39958557811852463, + "grad_norm": 0.251467227935791, + "learning_rate": 3.0024865312888522e-06, + "loss": 0.5281, + "step": 4821 + }, + { + "epoch": 0.39966846249481974, + "grad_norm": 0.23605981469154358, + "learning_rate": 3.0020721094073766e-06, + "loss": 0.5, + "step": 4822 + }, + { + "epoch": 0.3997513468711148, + "grad_norm": 0.2395111322402954, + "learning_rate": 3.001657687525902e-06, + "loss": 0.4971, + "step": 4823 + }, + { + "epoch": 0.39983423124740985, + "grad_norm": 0.27352094650268555, + "learning_rate": 3.0012432656444264e-06, + "loss": 0.5952, + "step": 4824 + }, + { + "epoch": 0.3999171156237049, + "grad_norm": 0.2532286047935486, + "learning_rate": 3.000828843762951e-06, + "loss": 0.4985, + "step": 4825 + }, + { + "epoch": 0.4, + "grad_norm": 0.2768342196941376, + "learning_rate": 3.0004144218814756e-06, + "loss": 0.5492, + "step": 4826 + }, + { + "epoch": 0.4000828843762951, + "grad_norm": 0.2659139633178711, + "learning_rate": 3e-06, + "loss": 0.5977, + "step": 4827 + }, + { + "epoch": 0.40016576875259013, + "grad_norm": 0.24578259885311127, + "learning_rate": 2.999585578118525e-06, + "loss": 0.4908, + "step": 4828 + }, + { + "epoch": 0.4002486531288852, + "grad_norm": 0.26190805435180664, + "learning_rate": 2.9991711562370494e-06, + "loss": 0.5554, + "step": 4829 + }, + { + "epoch": 0.40033153750518025, + "grad_norm": 0.28514522314071655, + "learning_rate": 2.9987567343555742e-06, + "loss": 0.5983, + "step": 4830 + }, + { + "epoch": 0.40041442188147536, + "grad_norm": 0.2602531313896179, + "learning_rate": 2.998342312474099e-06, + "loss": 0.5096, + "step": 4831 + }, + { + "epoch": 0.4004973062577704, + "grad_norm": 0.24198058247566223, + "learning_rate": 2.9979278905926235e-06, + "loss": 0.4777, + "step": 4832 + }, + { + "epoch": 0.40058019063406547, + "grad_norm": 0.292356014251709, + "learning_rate": 2.9975134687111484e-06, + "loss": 0.5421, + "step": 4833 + }, + { + "epoch": 0.4006630750103605, + "grad_norm": 0.2772829830646515, + "learning_rate": 2.997099046829673e-06, + "loss": 0.5334, + "step": 4834 + }, + { + "epoch": 0.40074595938665564, + "grad_norm": 0.2472977340221405, + "learning_rate": 2.9966846249481972e-06, + "loss": 0.5314, + "step": 4835 + }, + { + "epoch": 0.4008288437629507, + "grad_norm": 0.35984113812446594, + "learning_rate": 2.9962702030667225e-06, + "loss": 0.5079, + "step": 4836 + }, + { + "epoch": 0.40091172813924575, + "grad_norm": 0.2547149956226349, + "learning_rate": 2.995855781185247e-06, + "loss": 0.4842, + "step": 4837 + }, + { + "epoch": 0.4009946125155408, + "grad_norm": 0.26149728894233704, + "learning_rate": 2.9954413593037714e-06, + "loss": 0.5311, + "step": 4838 + }, + { + "epoch": 0.4010774968918359, + "grad_norm": 0.2652897834777832, + "learning_rate": 2.9950269374222962e-06, + "loss": 0.5425, + "step": 4839 + }, + { + "epoch": 0.40116038126813097, + "grad_norm": 0.24491502344608307, + "learning_rate": 2.9946125155408207e-06, + "loss": 0.5195, + "step": 4840 + }, + { + "epoch": 0.40124326564442603, + "grad_norm": 0.25960293412208557, + "learning_rate": 2.9941980936593455e-06, + "loss": 0.5609, + "step": 4841 + }, + { + "epoch": 0.4013261500207211, + "grad_norm": 0.2522276043891907, + "learning_rate": 2.99378367177787e-06, + "loss": 0.5077, + "step": 4842 + }, + { + "epoch": 0.40140903439701614, + "grad_norm": 0.27546370029449463, + "learning_rate": 2.9933692498963944e-06, + "loss": 0.5641, + "step": 4843 + }, + { + "epoch": 0.40149191877331125, + "grad_norm": 0.24766869843006134, + "learning_rate": 2.9929548280149197e-06, + "loss": 0.5201, + "step": 4844 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 0.24587060511112213, + "learning_rate": 2.992540406133444e-06, + "loss": 0.5137, + "step": 4845 + }, + { + "epoch": 0.40165768752590136, + "grad_norm": 0.2551528513431549, + "learning_rate": 2.992125984251969e-06, + "loss": 0.4807, + "step": 4846 + }, + { + "epoch": 0.4017405719021964, + "grad_norm": 0.25921785831451416, + "learning_rate": 2.9917115623704934e-06, + "loss": 0.5036, + "step": 4847 + }, + { + "epoch": 0.40182345627849153, + "grad_norm": 0.2747209072113037, + "learning_rate": 2.991297140489018e-06, + "loss": 0.528, + "step": 4848 + }, + { + "epoch": 0.4019063406547866, + "grad_norm": 0.29480892419815063, + "learning_rate": 2.9908827186075427e-06, + "loss": 0.5341, + "step": 4849 + }, + { + "epoch": 0.40198922503108164, + "grad_norm": 0.2709771692752838, + "learning_rate": 2.9904682967260675e-06, + "loss": 0.5718, + "step": 4850 + }, + { + "epoch": 0.4020721094073767, + "grad_norm": 0.2430618554353714, + "learning_rate": 2.9900538748445924e-06, + "loss": 0.4948, + "step": 4851 + }, + { + "epoch": 0.40215499378367175, + "grad_norm": 0.2458696812391281, + "learning_rate": 2.989639452963117e-06, + "loss": 0.4656, + "step": 4852 + }, + { + "epoch": 0.40223787815996687, + "grad_norm": 0.25445255637168884, + "learning_rate": 2.9892250310816412e-06, + "loss": 0.5373, + "step": 4853 + }, + { + "epoch": 0.4023207625362619, + "grad_norm": 0.2590615749359131, + "learning_rate": 2.988810609200166e-06, + "loss": 0.5336, + "step": 4854 + }, + { + "epoch": 0.402403646912557, + "grad_norm": 0.25814446806907654, + "learning_rate": 2.9883961873186905e-06, + "loss": 0.4982, + "step": 4855 + }, + { + "epoch": 0.40248653128885203, + "grad_norm": 0.25506430864334106, + "learning_rate": 2.987981765437215e-06, + "loss": 0.5193, + "step": 4856 + }, + { + "epoch": 0.40256941566514715, + "grad_norm": 0.2573833465576172, + "learning_rate": 2.9875673435557402e-06, + "loss": 0.542, + "step": 4857 + }, + { + "epoch": 0.4026523000414422, + "grad_norm": 0.23374703526496887, + "learning_rate": 2.9871529216742647e-06, + "loss": 0.467, + "step": 4858 + }, + { + "epoch": 0.40273518441773726, + "grad_norm": 0.2757858633995056, + "learning_rate": 2.9867384997927895e-06, + "loss": 0.5494, + "step": 4859 + }, + { + "epoch": 0.4028180687940323, + "grad_norm": 0.2438145875930786, + "learning_rate": 2.986324077911314e-06, + "loss": 0.5226, + "step": 4860 + }, + { + "epoch": 0.40290095317032737, + "grad_norm": 0.25373145937919617, + "learning_rate": 2.9859096560298384e-06, + "loss": 0.5599, + "step": 4861 + }, + { + "epoch": 0.4029838375466225, + "grad_norm": 0.2720401883125305, + "learning_rate": 2.9854952341483632e-06, + "loss": 0.529, + "step": 4862 + }, + { + "epoch": 0.40306672192291754, + "grad_norm": 0.248639777302742, + "learning_rate": 2.9850808122668877e-06, + "loss": 0.5192, + "step": 4863 + }, + { + "epoch": 0.4031496062992126, + "grad_norm": 0.2638450860977173, + "learning_rate": 2.984666390385413e-06, + "loss": 0.4973, + "step": 4864 + }, + { + "epoch": 0.40323249067550765, + "grad_norm": 0.24709035456180573, + "learning_rate": 2.9842519685039374e-06, + "loss": 0.4993, + "step": 4865 + }, + { + "epoch": 0.40331537505180276, + "grad_norm": 0.238994300365448, + "learning_rate": 2.983837546622462e-06, + "loss": 0.4879, + "step": 4866 + }, + { + "epoch": 0.4033982594280978, + "grad_norm": 0.27854517102241516, + "learning_rate": 2.9834231247409867e-06, + "loss": 0.5206, + "step": 4867 + }, + { + "epoch": 0.40348114380439287, + "grad_norm": 0.2754071354866028, + "learning_rate": 2.983008702859511e-06, + "loss": 0.4969, + "step": 4868 + }, + { + "epoch": 0.40356402818068793, + "grad_norm": 0.2888338565826416, + "learning_rate": 2.982594280978036e-06, + "loss": 0.5257, + "step": 4869 + }, + { + "epoch": 0.403646912556983, + "grad_norm": 0.266605406999588, + "learning_rate": 2.982179859096561e-06, + "loss": 0.5576, + "step": 4870 + }, + { + "epoch": 0.4037297969332781, + "grad_norm": 0.25818949937820435, + "learning_rate": 2.9817654372150852e-06, + "loss": 0.5488, + "step": 4871 + }, + { + "epoch": 0.40381268130957315, + "grad_norm": 0.2542945444583893, + "learning_rate": 2.98135101533361e-06, + "loss": 0.48, + "step": 4872 + }, + { + "epoch": 0.4038955656858682, + "grad_norm": 0.2723189890384674, + "learning_rate": 2.9809365934521345e-06, + "loss": 0.499, + "step": 4873 + }, + { + "epoch": 0.40397845006216326, + "grad_norm": 0.25790154933929443, + "learning_rate": 2.980522171570659e-06, + "loss": 0.5469, + "step": 4874 + }, + { + "epoch": 0.4040613344384584, + "grad_norm": 0.2816120982170105, + "learning_rate": 2.980107749689184e-06, + "loss": 0.5629, + "step": 4875 + }, + { + "epoch": 0.40414421881475343, + "grad_norm": 0.26743945479393005, + "learning_rate": 2.9796933278077082e-06, + "loss": 0.5447, + "step": 4876 + }, + { + "epoch": 0.4042271031910485, + "grad_norm": 0.2587490975856781, + "learning_rate": 2.9792789059262335e-06, + "loss": 0.4917, + "step": 4877 + }, + { + "epoch": 0.40430998756734354, + "grad_norm": 0.2677125036716461, + "learning_rate": 2.978864484044758e-06, + "loss": 0.4968, + "step": 4878 + }, + { + "epoch": 0.4043928719436386, + "grad_norm": 0.2574442923069, + "learning_rate": 2.9784500621632824e-06, + "loss": 0.5429, + "step": 4879 + }, + { + "epoch": 0.4044757563199337, + "grad_norm": 0.25471827387809753, + "learning_rate": 2.9780356402818072e-06, + "loss": 0.5352, + "step": 4880 + }, + { + "epoch": 0.40455864069622877, + "grad_norm": 0.259956955909729, + "learning_rate": 2.9776212184003317e-06, + "loss": 0.531, + "step": 4881 + }, + { + "epoch": 0.4046415250725238, + "grad_norm": 0.27281853556632996, + "learning_rate": 2.9772067965188565e-06, + "loss": 0.4959, + "step": 4882 + }, + { + "epoch": 0.4047244094488189, + "grad_norm": 0.2715339660644531, + "learning_rate": 2.976792374637381e-06, + "loss": 0.5143, + "step": 4883 + }, + { + "epoch": 0.404807293825114, + "grad_norm": 0.2712486684322357, + "learning_rate": 2.9763779527559054e-06, + "loss": 0.567, + "step": 4884 + }, + { + "epoch": 0.40489017820140905, + "grad_norm": 0.27260005474090576, + "learning_rate": 2.9759635308744307e-06, + "loss": 0.5285, + "step": 4885 + }, + { + "epoch": 0.4049730625777041, + "grad_norm": 0.2673589885234833, + "learning_rate": 2.975549108992955e-06, + "loss": 0.5077, + "step": 4886 + }, + { + "epoch": 0.40505594695399916, + "grad_norm": 0.2769814729690552, + "learning_rate": 2.9751346871114795e-06, + "loss": 0.5599, + "step": 4887 + }, + { + "epoch": 0.4051388313302942, + "grad_norm": 0.2629331946372986, + "learning_rate": 2.9747202652300044e-06, + "loss": 0.5175, + "step": 4888 + }, + { + "epoch": 0.4052217157065893, + "grad_norm": 0.26450368762016296, + "learning_rate": 2.974305843348529e-06, + "loss": 0.518, + "step": 4889 + }, + { + "epoch": 0.4053046000828844, + "grad_norm": 0.2444290965795517, + "learning_rate": 2.973891421467054e-06, + "loss": 0.5184, + "step": 4890 + }, + { + "epoch": 0.40538748445917944, + "grad_norm": 0.25875478982925415, + "learning_rate": 2.9734769995855785e-06, + "loss": 0.5844, + "step": 4891 + }, + { + "epoch": 0.4054703688354745, + "grad_norm": 0.2565544545650482, + "learning_rate": 2.973062577704103e-06, + "loss": 0.5426, + "step": 4892 + }, + { + "epoch": 0.4055532532117696, + "grad_norm": 0.3166554570198059, + "learning_rate": 2.972648155822628e-06, + "loss": 0.5593, + "step": 4893 + }, + { + "epoch": 0.40563613758806466, + "grad_norm": 0.2728252410888672, + "learning_rate": 2.9722337339411523e-06, + "loss": 0.4731, + "step": 4894 + }, + { + "epoch": 0.4057190219643597, + "grad_norm": 0.2269505113363266, + "learning_rate": 2.971819312059677e-06, + "loss": 0.5051, + "step": 4895 + }, + { + "epoch": 0.40580190634065477, + "grad_norm": 0.2706831097602844, + "learning_rate": 2.9714048901782015e-06, + "loss": 0.5506, + "step": 4896 + }, + { + "epoch": 0.40588479071694983, + "grad_norm": 0.25724685192108154, + "learning_rate": 2.970990468296726e-06, + "loss": 0.5645, + "step": 4897 + }, + { + "epoch": 0.40596767509324494, + "grad_norm": 0.2517000436782837, + "learning_rate": 2.9705760464152513e-06, + "loss": 0.5328, + "step": 4898 + }, + { + "epoch": 0.40605055946954, + "grad_norm": 0.255593478679657, + "learning_rate": 2.9701616245337757e-06, + "loss": 0.5414, + "step": 4899 + }, + { + "epoch": 0.40613344384583505, + "grad_norm": 0.2584759593009949, + "learning_rate": 2.9697472026523005e-06, + "loss": 0.5269, + "step": 4900 + }, + { + "epoch": 0.4062163282221301, + "grad_norm": 0.26033762097358704, + "learning_rate": 2.969332780770825e-06, + "loss": 0.5209, + "step": 4901 + }, + { + "epoch": 0.4062992125984252, + "grad_norm": 0.2352954000234604, + "learning_rate": 2.9689183588893494e-06, + "loss": 0.4849, + "step": 4902 + }, + { + "epoch": 0.4063820969747203, + "grad_norm": 0.24179396033287048, + "learning_rate": 2.9685039370078743e-06, + "loss": 0.5031, + "step": 4903 + }, + { + "epoch": 0.40646498135101533, + "grad_norm": 0.2478998303413391, + "learning_rate": 2.9680895151263987e-06, + "loss": 0.5229, + "step": 4904 + }, + { + "epoch": 0.4065478657273104, + "grad_norm": 0.2416868805885315, + "learning_rate": 2.967675093244923e-06, + "loss": 0.5015, + "step": 4905 + }, + { + "epoch": 0.40663075010360544, + "grad_norm": 0.25402772426605225, + "learning_rate": 2.9672606713634484e-06, + "loss": 0.5812, + "step": 4906 + }, + { + "epoch": 0.40671363447990055, + "grad_norm": 0.25727126002311707, + "learning_rate": 2.966846249481973e-06, + "loss": 0.5458, + "step": 4907 + }, + { + "epoch": 0.4067965188561956, + "grad_norm": 0.25249263644218445, + "learning_rate": 2.9664318276004977e-06, + "loss": 0.5329, + "step": 4908 + }, + { + "epoch": 0.40687940323249067, + "grad_norm": 0.25001317262649536, + "learning_rate": 2.966017405719022e-06, + "loss": 0.4903, + "step": 4909 + }, + { + "epoch": 0.4069622876087857, + "grad_norm": 0.252037912607193, + "learning_rate": 2.9656029838375465e-06, + "loss": 0.541, + "step": 4910 + }, + { + "epoch": 0.40704517198508083, + "grad_norm": 0.27182888984680176, + "learning_rate": 2.965188561956072e-06, + "loss": 0.557, + "step": 4911 + }, + { + "epoch": 0.4071280563613759, + "grad_norm": 0.2712077498435974, + "learning_rate": 2.9647741400745963e-06, + "loss": 0.5509, + "step": 4912 + }, + { + "epoch": 0.40721094073767095, + "grad_norm": 0.2548147141933441, + "learning_rate": 2.964359718193121e-06, + "loss": 0.4954, + "step": 4913 + }, + { + "epoch": 0.407293825113966, + "grad_norm": 0.24016264081001282, + "learning_rate": 2.9639452963116455e-06, + "loss": 0.4731, + "step": 4914 + }, + { + "epoch": 0.4073767094902611, + "grad_norm": 0.25127214193344116, + "learning_rate": 2.96353087443017e-06, + "loss": 0.5551, + "step": 4915 + }, + { + "epoch": 0.40745959386655617, + "grad_norm": 0.2459174245595932, + "learning_rate": 2.963116452548695e-06, + "loss": 0.5228, + "step": 4916 + }, + { + "epoch": 0.4075424782428512, + "grad_norm": 0.243338480591774, + "learning_rate": 2.9627020306672193e-06, + "loss": 0.4933, + "step": 4917 + }, + { + "epoch": 0.4076253626191463, + "grad_norm": 0.24294346570968628, + "learning_rate": 2.9622876087857445e-06, + "loss": 0.5289, + "step": 4918 + }, + { + "epoch": 0.40770824699544134, + "grad_norm": 0.26000234484672546, + "learning_rate": 2.961873186904269e-06, + "loss": 0.5456, + "step": 4919 + }, + { + "epoch": 0.40779113137173645, + "grad_norm": 0.263433039188385, + "learning_rate": 2.9614587650227934e-06, + "loss": 0.5377, + "step": 4920 + }, + { + "epoch": 0.4078740157480315, + "grad_norm": 0.2690461575984955, + "learning_rate": 2.9610443431413183e-06, + "loss": 0.5594, + "step": 4921 + }, + { + "epoch": 0.40795690012432656, + "grad_norm": 0.29023832082748413, + "learning_rate": 2.9606299212598427e-06, + "loss": 0.5446, + "step": 4922 + }, + { + "epoch": 0.4080397845006216, + "grad_norm": 0.24239572882652283, + "learning_rate": 2.960215499378367e-06, + "loss": 0.5291, + "step": 4923 + }, + { + "epoch": 0.4081226688769167, + "grad_norm": 0.2788420021533966, + "learning_rate": 2.959801077496892e-06, + "loss": 0.5636, + "step": 4924 + }, + { + "epoch": 0.4082055532532118, + "grad_norm": 0.25694069266319275, + "learning_rate": 2.9593866556154164e-06, + "loss": 0.5277, + "step": 4925 + }, + { + "epoch": 0.40828843762950684, + "grad_norm": 0.25424110889434814, + "learning_rate": 2.9589722337339417e-06, + "loss": 0.5211, + "step": 4926 + }, + { + "epoch": 0.4083713220058019, + "grad_norm": 0.24508875608444214, + "learning_rate": 2.958557811852466e-06, + "loss": 0.5072, + "step": 4927 + }, + { + "epoch": 0.40845420638209695, + "grad_norm": 0.2361994832754135, + "learning_rate": 2.9581433899709906e-06, + "loss": 0.4926, + "step": 4928 + }, + { + "epoch": 0.40853709075839206, + "grad_norm": 0.2447447031736374, + "learning_rate": 2.9577289680895154e-06, + "loss": 0.4754, + "step": 4929 + }, + { + "epoch": 0.4086199751346871, + "grad_norm": 0.24740296602249146, + "learning_rate": 2.95731454620804e-06, + "loss": 0.5341, + "step": 4930 + }, + { + "epoch": 0.4087028595109822, + "grad_norm": 0.2539823055267334, + "learning_rate": 2.956900124326565e-06, + "loss": 0.515, + "step": 4931 + }, + { + "epoch": 0.40878574388727723, + "grad_norm": 0.24572648108005524, + "learning_rate": 2.9564857024450896e-06, + "loss": 0.5463, + "step": 4932 + }, + { + "epoch": 0.40886862826357234, + "grad_norm": 0.2609131932258606, + "learning_rate": 2.956071280563614e-06, + "loss": 0.5295, + "step": 4933 + }, + { + "epoch": 0.4089515126398674, + "grad_norm": 0.2604348957538605, + "learning_rate": 2.955656858682139e-06, + "loss": 0.4955, + "step": 4934 + }, + { + "epoch": 0.40903439701616245, + "grad_norm": 0.26468929648399353, + "learning_rate": 2.9552424368006633e-06, + "loss": 0.501, + "step": 4935 + }, + { + "epoch": 0.4091172813924575, + "grad_norm": 0.25447168946266174, + "learning_rate": 2.954828014919188e-06, + "loss": 0.4934, + "step": 4936 + }, + { + "epoch": 0.40920016576875257, + "grad_norm": 0.26291072368621826, + "learning_rate": 2.9544135930377126e-06, + "loss": 0.5432, + "step": 4937 + }, + { + "epoch": 0.4092830501450477, + "grad_norm": 0.2434622347354889, + "learning_rate": 2.953999171156237e-06, + "loss": 0.4738, + "step": 4938 + }, + { + "epoch": 0.40936593452134273, + "grad_norm": 0.2440808266401291, + "learning_rate": 2.9535847492747623e-06, + "loss": 0.5192, + "step": 4939 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 0.24782626330852509, + "learning_rate": 2.9531703273932867e-06, + "loss": 0.5186, + "step": 4940 + }, + { + "epoch": 0.40953170327393285, + "grad_norm": 0.25640374422073364, + "learning_rate": 2.952755905511811e-06, + "loss": 0.5154, + "step": 4941 + }, + { + "epoch": 0.40961458765022796, + "grad_norm": 0.23874159157276154, + "learning_rate": 2.952341483630336e-06, + "loss": 0.5186, + "step": 4942 + }, + { + "epoch": 0.409697472026523, + "grad_norm": 0.25458747148513794, + "learning_rate": 2.9519270617488604e-06, + "loss": 0.5606, + "step": 4943 + }, + { + "epoch": 0.40978035640281807, + "grad_norm": 0.26823166012763977, + "learning_rate": 2.9515126398673853e-06, + "loss": 0.5984, + "step": 4944 + }, + { + "epoch": 0.4098632407791131, + "grad_norm": 0.27262768149375916, + "learning_rate": 2.9510982179859097e-06, + "loss": 0.5391, + "step": 4945 + }, + { + "epoch": 0.4099461251554082, + "grad_norm": 0.25767460465431213, + "learning_rate": 2.950683796104434e-06, + "loss": 0.5321, + "step": 4946 + }, + { + "epoch": 0.4100290095317033, + "grad_norm": 0.2769319415092468, + "learning_rate": 2.9502693742229594e-06, + "loss": 0.5546, + "step": 4947 + }, + { + "epoch": 0.41011189390799835, + "grad_norm": 0.25028929114341736, + "learning_rate": 2.949854952341484e-06, + "loss": 0.5226, + "step": 4948 + }, + { + "epoch": 0.4101947782842934, + "grad_norm": 0.26034876704216003, + "learning_rate": 2.9494405304600087e-06, + "loss": 0.5228, + "step": 4949 + }, + { + "epoch": 0.41027766266058846, + "grad_norm": 0.24665643274784088, + "learning_rate": 2.949026108578533e-06, + "loss": 0.5145, + "step": 4950 + }, + { + "epoch": 0.41036054703688357, + "grad_norm": 0.25472575426101685, + "learning_rate": 2.9486116866970576e-06, + "loss": 0.528, + "step": 4951 + }, + { + "epoch": 0.4104434314131786, + "grad_norm": 0.2666727602481842, + "learning_rate": 2.948197264815583e-06, + "loss": 0.5358, + "step": 4952 + }, + { + "epoch": 0.4105263157894737, + "grad_norm": 0.2697432041168213, + "learning_rate": 2.9477828429341073e-06, + "loss": 0.5442, + "step": 4953 + }, + { + "epoch": 0.41060920016576874, + "grad_norm": 0.27392852306365967, + "learning_rate": 2.9473684210526317e-06, + "loss": 0.5428, + "step": 4954 + }, + { + "epoch": 0.4106920845420638, + "grad_norm": 0.26656824350357056, + "learning_rate": 2.9469539991711566e-06, + "loss": 0.5274, + "step": 4955 + }, + { + "epoch": 0.4107749689183589, + "grad_norm": 0.266789048910141, + "learning_rate": 2.946539577289681e-06, + "loss": 0.5426, + "step": 4956 + }, + { + "epoch": 0.41085785329465396, + "grad_norm": 0.2634730935096741, + "learning_rate": 2.946125155408206e-06, + "loss": 0.5056, + "step": 4957 + }, + { + "epoch": 0.410940737670949, + "grad_norm": 0.265895813703537, + "learning_rate": 2.9457107335267303e-06, + "loss": 0.5111, + "step": 4958 + }, + { + "epoch": 0.4110236220472441, + "grad_norm": 0.252230167388916, + "learning_rate": 2.9452963116452547e-06, + "loss": 0.4728, + "step": 4959 + }, + { + "epoch": 0.4111065064235392, + "grad_norm": 0.26063039898872375, + "learning_rate": 2.94488188976378e-06, + "loss": 0.532, + "step": 4960 + }, + { + "epoch": 0.41118939079983424, + "grad_norm": 0.2858526408672333, + "learning_rate": 2.9444674678823044e-06, + "loss": 0.5868, + "step": 4961 + }, + { + "epoch": 0.4112722751761293, + "grad_norm": 0.2645666301250458, + "learning_rate": 2.9440530460008293e-06, + "loss": 0.5306, + "step": 4962 + }, + { + "epoch": 0.41135515955242435, + "grad_norm": 0.2594447135925293, + "learning_rate": 2.9436386241193537e-06, + "loss": 0.5512, + "step": 4963 + }, + { + "epoch": 0.4114380439287194, + "grad_norm": 0.23119932413101196, + "learning_rate": 2.943224202237878e-06, + "loss": 0.4801, + "step": 4964 + }, + { + "epoch": 0.4115209283050145, + "grad_norm": 0.2506735026836395, + "learning_rate": 2.942809780356403e-06, + "loss": 0.5037, + "step": 4965 + }, + { + "epoch": 0.4116038126813096, + "grad_norm": 0.2675521671772003, + "learning_rate": 2.9423953584749274e-06, + "loss": 0.539, + "step": 4966 + }, + { + "epoch": 0.41168669705760463, + "grad_norm": 0.2681416869163513, + "learning_rate": 2.9419809365934527e-06, + "loss": 0.5605, + "step": 4967 + }, + { + "epoch": 0.4117695814338997, + "grad_norm": 0.25146716833114624, + "learning_rate": 2.941566514711977e-06, + "loss": 0.5247, + "step": 4968 + }, + { + "epoch": 0.4118524658101948, + "grad_norm": 0.2506044805049896, + "learning_rate": 2.9411520928305016e-06, + "loss": 0.5311, + "step": 4969 + }, + { + "epoch": 0.41193535018648986, + "grad_norm": 0.24961569905281067, + "learning_rate": 2.9407376709490264e-06, + "loss": 0.5111, + "step": 4970 + }, + { + "epoch": 0.4120182345627849, + "grad_norm": 0.24840208888053894, + "learning_rate": 2.940323249067551e-06, + "loss": 0.4982, + "step": 4971 + }, + { + "epoch": 0.41210111893907997, + "grad_norm": 0.2625523507595062, + "learning_rate": 2.9399088271860753e-06, + "loss": 0.5122, + "step": 4972 + }, + { + "epoch": 0.412184003315375, + "grad_norm": 0.24971280992031097, + "learning_rate": 2.9394944053046006e-06, + "loss": 0.5179, + "step": 4973 + }, + { + "epoch": 0.41226688769167014, + "grad_norm": 0.25132226943969727, + "learning_rate": 2.939079983423125e-06, + "loss": 0.5208, + "step": 4974 + }, + { + "epoch": 0.4123497720679652, + "grad_norm": 0.25553518533706665, + "learning_rate": 2.93866556154165e-06, + "loss": 0.5019, + "step": 4975 + }, + { + "epoch": 0.41243265644426025, + "grad_norm": 0.24509674310684204, + "learning_rate": 2.9382511396601743e-06, + "loss": 0.5007, + "step": 4976 + }, + { + "epoch": 0.4125155408205553, + "grad_norm": 0.23406681418418884, + "learning_rate": 2.9378367177786987e-06, + "loss": 0.5113, + "step": 4977 + }, + { + "epoch": 0.4125984251968504, + "grad_norm": 0.26357725262641907, + "learning_rate": 2.9374222958972236e-06, + "loss": 0.504, + "step": 4978 + }, + { + "epoch": 0.41268130957314547, + "grad_norm": 0.2646195888519287, + "learning_rate": 2.937007874015748e-06, + "loss": 0.4855, + "step": 4979 + }, + { + "epoch": 0.4127641939494405, + "grad_norm": 0.2630488872528076, + "learning_rate": 2.9365934521342733e-06, + "loss": 0.5033, + "step": 4980 + }, + { + "epoch": 0.4128470783257356, + "grad_norm": 0.2610452175140381, + "learning_rate": 2.9361790302527977e-06, + "loss": 0.4805, + "step": 4981 + }, + { + "epoch": 0.41292996270203064, + "grad_norm": 0.24615177512168884, + "learning_rate": 2.935764608371322e-06, + "loss": 0.5507, + "step": 4982 + }, + { + "epoch": 0.41301284707832575, + "grad_norm": 0.2518768310546875, + "learning_rate": 2.935350186489847e-06, + "loss": 0.504, + "step": 4983 + }, + { + "epoch": 0.4130957314546208, + "grad_norm": 0.25878646969795227, + "learning_rate": 2.9349357646083714e-06, + "loss": 0.531, + "step": 4984 + }, + { + "epoch": 0.41317861583091586, + "grad_norm": 0.25684013962745667, + "learning_rate": 2.9345213427268963e-06, + "loss": 0.4917, + "step": 4985 + }, + { + "epoch": 0.4132615002072109, + "grad_norm": 0.2518225610256195, + "learning_rate": 2.9341069208454207e-06, + "loss": 0.5154, + "step": 4986 + }, + { + "epoch": 0.41334438458350603, + "grad_norm": 0.24584074318408966, + "learning_rate": 2.933692498963945e-06, + "loss": 0.5703, + "step": 4987 + }, + { + "epoch": 0.4134272689598011, + "grad_norm": 0.24787208437919617, + "learning_rate": 2.9332780770824704e-06, + "loss": 0.4625, + "step": 4988 + }, + { + "epoch": 0.41351015333609614, + "grad_norm": 0.24846825003623962, + "learning_rate": 2.932863655200995e-06, + "loss": 0.5134, + "step": 4989 + }, + { + "epoch": 0.4135930377123912, + "grad_norm": 0.2510080337524414, + "learning_rate": 2.9324492333195193e-06, + "loss": 0.5406, + "step": 4990 + }, + { + "epoch": 0.4136759220886863, + "grad_norm": 0.2734019458293915, + "learning_rate": 2.932034811438044e-06, + "loss": 0.5658, + "step": 4991 + }, + { + "epoch": 0.41375880646498137, + "grad_norm": 0.26595374941825867, + "learning_rate": 2.9316203895565686e-06, + "loss": 0.4812, + "step": 4992 + }, + { + "epoch": 0.4138416908412764, + "grad_norm": 0.2626924514770508, + "learning_rate": 2.931205967675094e-06, + "loss": 0.5349, + "step": 4993 + }, + { + "epoch": 0.4139245752175715, + "grad_norm": 0.25561973452568054, + "learning_rate": 2.9307915457936183e-06, + "loss": 0.5222, + "step": 4994 + }, + { + "epoch": 0.41400745959386653, + "grad_norm": 0.2545754313468933, + "learning_rate": 2.9303771239121427e-06, + "loss": 0.5063, + "step": 4995 + }, + { + "epoch": 0.41409034397016165, + "grad_norm": 0.24929966032505035, + "learning_rate": 2.9299627020306676e-06, + "loss": 0.5614, + "step": 4996 + }, + { + "epoch": 0.4141732283464567, + "grad_norm": 0.24484783411026, + "learning_rate": 2.929548280149192e-06, + "loss": 0.4918, + "step": 4997 + }, + { + "epoch": 0.41425611272275176, + "grad_norm": 0.26942792534828186, + "learning_rate": 2.929133858267717e-06, + "loss": 0.553, + "step": 4998 + }, + { + "epoch": 0.4143389970990468, + "grad_norm": 0.24062731862068176, + "learning_rate": 2.9287194363862413e-06, + "loss": 0.5411, + "step": 4999 + }, + { + "epoch": 0.4144218814753419, + "grad_norm": 0.27081722021102905, + "learning_rate": 2.9283050145047657e-06, + "loss": 0.515, + "step": 5000 + }, + { + "epoch": 0.414504765851637, + "grad_norm": 0.25544750690460205, + "learning_rate": 2.927890592623291e-06, + "loss": 0.5297, + "step": 5001 + }, + { + "epoch": 0.41458765022793204, + "grad_norm": 0.2662317156791687, + "learning_rate": 2.9274761707418155e-06, + "loss": 0.5651, + "step": 5002 + }, + { + "epoch": 0.4146705346042271, + "grad_norm": 0.25782889127731323, + "learning_rate": 2.9270617488603403e-06, + "loss": 0.5558, + "step": 5003 + }, + { + "epoch": 0.41475341898052215, + "grad_norm": 0.2667877972126007, + "learning_rate": 2.9266473269788647e-06, + "loss": 0.524, + "step": 5004 + }, + { + "epoch": 0.41483630335681726, + "grad_norm": 0.2662474811077118, + "learning_rate": 2.926232905097389e-06, + "loss": 0.5078, + "step": 5005 + }, + { + "epoch": 0.4149191877331123, + "grad_norm": 0.24992790818214417, + "learning_rate": 2.925818483215914e-06, + "loss": 0.5336, + "step": 5006 + }, + { + "epoch": 0.41500207210940737, + "grad_norm": 0.2614598572254181, + "learning_rate": 2.9254040613344385e-06, + "loss": 0.5433, + "step": 5007 + }, + { + "epoch": 0.4150849564857024, + "grad_norm": 0.24745707213878632, + "learning_rate": 2.9249896394529633e-06, + "loss": 0.5033, + "step": 5008 + }, + { + "epoch": 0.41516784086199754, + "grad_norm": 0.2511752247810364, + "learning_rate": 2.924575217571488e-06, + "loss": 0.4948, + "step": 5009 + }, + { + "epoch": 0.4152507252382926, + "grad_norm": 0.26516491174697876, + "learning_rate": 2.9241607956900126e-06, + "loss": 0.494, + "step": 5010 + }, + { + "epoch": 0.41533360961458765, + "grad_norm": 0.25768303871154785, + "learning_rate": 2.9237463738085375e-06, + "loss": 0.5368, + "step": 5011 + }, + { + "epoch": 0.4154164939908827, + "grad_norm": 0.255367249250412, + "learning_rate": 2.923331951927062e-06, + "loss": 0.5331, + "step": 5012 + }, + { + "epoch": 0.41549937836717776, + "grad_norm": 0.25258317589759827, + "learning_rate": 2.9229175300455863e-06, + "loss": 0.4776, + "step": 5013 + }, + { + "epoch": 0.4155822627434729, + "grad_norm": 0.2753553092479706, + "learning_rate": 2.9225031081641116e-06, + "loss": 0.5239, + "step": 5014 + }, + { + "epoch": 0.41566514711976793, + "grad_norm": 0.2406659722328186, + "learning_rate": 2.922088686282636e-06, + "loss": 0.4987, + "step": 5015 + }, + { + "epoch": 0.415748031496063, + "grad_norm": 0.26624226570129395, + "learning_rate": 2.921674264401161e-06, + "loss": 0.5465, + "step": 5016 + }, + { + "epoch": 0.41583091587235804, + "grad_norm": 0.2725326418876648, + "learning_rate": 2.9212598425196853e-06, + "loss": 0.511, + "step": 5017 + }, + { + "epoch": 0.41591380024865315, + "grad_norm": 0.2581869661808014, + "learning_rate": 2.9208454206382097e-06, + "loss": 0.502, + "step": 5018 + }, + { + "epoch": 0.4159966846249482, + "grad_norm": 0.27246615290641785, + "learning_rate": 2.9204309987567346e-06, + "loss": 0.4702, + "step": 5019 + }, + { + "epoch": 0.41607956900124327, + "grad_norm": 0.2690960764884949, + "learning_rate": 2.920016576875259e-06, + "loss": 0.5502, + "step": 5020 + }, + { + "epoch": 0.4161624533775383, + "grad_norm": 0.2362431138753891, + "learning_rate": 2.9196021549937843e-06, + "loss": 0.4792, + "step": 5021 + }, + { + "epoch": 0.4162453377538334, + "grad_norm": 0.23588323593139648, + "learning_rate": 2.9191877331123087e-06, + "loss": 0.4687, + "step": 5022 + }, + { + "epoch": 0.4163282221301285, + "grad_norm": 0.2506137490272522, + "learning_rate": 2.918773311230833e-06, + "loss": 0.5273, + "step": 5023 + }, + { + "epoch": 0.41641110650642355, + "grad_norm": 0.24542836844921112, + "learning_rate": 2.918358889349358e-06, + "loss": 0.4775, + "step": 5024 + }, + { + "epoch": 0.4164939908827186, + "grad_norm": 0.25093743205070496, + "learning_rate": 2.9179444674678825e-06, + "loss": 0.5081, + "step": 5025 + }, + { + "epoch": 0.41657687525901366, + "grad_norm": 0.2603587806224823, + "learning_rate": 2.917530045586407e-06, + "loss": 0.5137, + "step": 5026 + }, + { + "epoch": 0.41665975963530877, + "grad_norm": 0.2647421956062317, + "learning_rate": 2.9171156237049318e-06, + "loss": 0.527, + "step": 5027 + }, + { + "epoch": 0.4167426440116038, + "grad_norm": 0.26106226444244385, + "learning_rate": 2.9167012018234566e-06, + "loss": 0.499, + "step": 5028 + }, + { + "epoch": 0.4168255283878989, + "grad_norm": 0.25934097170829773, + "learning_rate": 2.9162867799419815e-06, + "loss": 0.5286, + "step": 5029 + }, + { + "epoch": 0.41690841276419394, + "grad_norm": 0.2540843188762665, + "learning_rate": 2.915872358060506e-06, + "loss": 0.5439, + "step": 5030 + }, + { + "epoch": 0.416991297140489, + "grad_norm": 0.45225203037261963, + "learning_rate": 2.9154579361790303e-06, + "loss": 0.5099, + "step": 5031 + }, + { + "epoch": 0.4170741815167841, + "grad_norm": 0.2748815417289734, + "learning_rate": 2.915043514297555e-06, + "loss": 0.551, + "step": 5032 + }, + { + "epoch": 0.41715706589307916, + "grad_norm": 0.24205376207828522, + "learning_rate": 2.9146290924160796e-06, + "loss": 0.5121, + "step": 5033 + }, + { + "epoch": 0.4172399502693742, + "grad_norm": 0.2542150914669037, + "learning_rate": 2.914214670534605e-06, + "loss": 0.4913, + "step": 5034 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 0.23880890011787415, + "learning_rate": 2.9138002486531293e-06, + "loss": 0.501, + "step": 5035 + }, + { + "epoch": 0.4174057190219644, + "grad_norm": 0.24568955600261688, + "learning_rate": 2.9133858267716538e-06, + "loss": 0.504, + "step": 5036 + }, + { + "epoch": 0.41748860339825944, + "grad_norm": 0.2600177526473999, + "learning_rate": 2.9129714048901786e-06, + "loss": 0.5491, + "step": 5037 + }, + { + "epoch": 0.4175714877745545, + "grad_norm": 0.2567923069000244, + "learning_rate": 2.912556983008703e-06, + "loss": 0.5248, + "step": 5038 + }, + { + "epoch": 0.41765437215084955, + "grad_norm": 0.2566099762916565, + "learning_rate": 2.9121425611272275e-06, + "loss": 0.5339, + "step": 5039 + }, + { + "epoch": 0.4177372565271446, + "grad_norm": 0.23807400465011597, + "learning_rate": 2.9117281392457523e-06, + "loss": 0.5266, + "step": 5040 + }, + { + "epoch": 0.4178201409034397, + "grad_norm": 0.2575412392616272, + "learning_rate": 2.9113137173642768e-06, + "loss": 0.5336, + "step": 5041 + }, + { + "epoch": 0.4179030252797348, + "grad_norm": 0.2583102583885193, + "learning_rate": 2.910899295482802e-06, + "loss": 0.5372, + "step": 5042 + }, + { + "epoch": 0.41798590965602983, + "grad_norm": 0.2510637640953064, + "learning_rate": 2.9104848736013265e-06, + "loss": 0.5367, + "step": 5043 + }, + { + "epoch": 0.4180687940323249, + "grad_norm": 0.28144553303718567, + "learning_rate": 2.910070451719851e-06, + "loss": 0.5433, + "step": 5044 + }, + { + "epoch": 0.41815167840862, + "grad_norm": 0.2551608383655548, + "learning_rate": 2.9096560298383758e-06, + "loss": 0.4949, + "step": 5045 + }, + { + "epoch": 0.41823456278491505, + "grad_norm": 0.2584105134010315, + "learning_rate": 2.9092416079569e-06, + "loss": 0.5087, + "step": 5046 + }, + { + "epoch": 0.4183174471612101, + "grad_norm": 0.27657613158226013, + "learning_rate": 2.908827186075425e-06, + "loss": 0.5014, + "step": 5047 + }, + { + "epoch": 0.41840033153750517, + "grad_norm": 0.26344338059425354, + "learning_rate": 2.9084127641939495e-06, + "loss": 0.4958, + "step": 5048 + }, + { + "epoch": 0.4184832159138002, + "grad_norm": 0.24705170094966888, + "learning_rate": 2.9079983423124743e-06, + "loss": 0.5307, + "step": 5049 + }, + { + "epoch": 0.41856610029009533, + "grad_norm": 0.2604838013648987, + "learning_rate": 2.907583920430999e-06, + "loss": 0.5185, + "step": 5050 + }, + { + "epoch": 0.4186489846663904, + "grad_norm": 0.2756589949131012, + "learning_rate": 2.9071694985495236e-06, + "loss": 0.5255, + "step": 5051 + }, + { + "epoch": 0.41873186904268545, + "grad_norm": 0.2628152668476105, + "learning_rate": 2.9067550766680485e-06, + "loss": 0.5289, + "step": 5052 + }, + { + "epoch": 0.4188147534189805, + "grad_norm": 0.2546418011188507, + "learning_rate": 2.906340654786573e-06, + "loss": 0.5085, + "step": 5053 + }, + { + "epoch": 0.4188976377952756, + "grad_norm": 0.28383028507232666, + "learning_rate": 2.9059262329050973e-06, + "loss": 0.5294, + "step": 5054 + }, + { + "epoch": 0.41898052217157067, + "grad_norm": 0.2700231969356537, + "learning_rate": 2.9055118110236226e-06, + "loss": 0.4969, + "step": 5055 + }, + { + "epoch": 0.4190634065478657, + "grad_norm": 0.2451544553041458, + "learning_rate": 2.905097389142147e-06, + "loss": 0.5243, + "step": 5056 + }, + { + "epoch": 0.4191462909241608, + "grad_norm": 0.2553543150424957, + "learning_rate": 2.9046829672606715e-06, + "loss": 0.5134, + "step": 5057 + }, + { + "epoch": 0.41922917530045584, + "grad_norm": 0.2661764621734619, + "learning_rate": 2.9042685453791963e-06, + "loss": 0.5156, + "step": 5058 + }, + { + "epoch": 0.41931205967675095, + "grad_norm": 0.26039814949035645, + "learning_rate": 2.9038541234977208e-06, + "loss": 0.5121, + "step": 5059 + }, + { + "epoch": 0.419394944053046, + "grad_norm": 0.26363271474838257, + "learning_rate": 2.9034397016162456e-06, + "loss": 0.5342, + "step": 5060 + }, + { + "epoch": 0.41947782842934106, + "grad_norm": 0.2502456307411194, + "learning_rate": 2.90302527973477e-06, + "loss": 0.5024, + "step": 5061 + }, + { + "epoch": 0.4195607128056361, + "grad_norm": 0.25918808579444885, + "learning_rate": 2.9026108578532945e-06, + "loss": 0.5388, + "step": 5062 + }, + { + "epoch": 0.4196435971819312, + "grad_norm": 0.24835895001888275, + "learning_rate": 2.9021964359718198e-06, + "loss": 0.5007, + "step": 5063 + }, + { + "epoch": 0.4197264815582263, + "grad_norm": 0.2398509681224823, + "learning_rate": 2.901782014090344e-06, + "loss": 0.4683, + "step": 5064 + }, + { + "epoch": 0.41980936593452134, + "grad_norm": 0.25060155987739563, + "learning_rate": 2.901367592208869e-06, + "loss": 0.5243, + "step": 5065 + }, + { + "epoch": 0.4198922503108164, + "grad_norm": 0.27608540654182434, + "learning_rate": 2.9009531703273935e-06, + "loss": 0.4937, + "step": 5066 + }, + { + "epoch": 0.4199751346871115, + "grad_norm": 0.24253778159618378, + "learning_rate": 2.900538748445918e-06, + "loss": 0.474, + "step": 5067 + }, + { + "epoch": 0.42005801906340656, + "grad_norm": 0.25196337699890137, + "learning_rate": 2.9001243265644428e-06, + "loss": 0.5627, + "step": 5068 + }, + { + "epoch": 0.4201409034397016, + "grad_norm": 0.23633849620819092, + "learning_rate": 2.8997099046829676e-06, + "loss": 0.5052, + "step": 5069 + }, + { + "epoch": 0.4202237878159967, + "grad_norm": 0.24936647713184357, + "learning_rate": 2.8992954828014925e-06, + "loss": 0.5153, + "step": 5070 + }, + { + "epoch": 0.42030667219229173, + "grad_norm": 0.24509726464748383, + "learning_rate": 2.898881060920017e-06, + "loss": 0.5024, + "step": 5071 + }, + { + "epoch": 0.42038955656858684, + "grad_norm": 0.2694768011569977, + "learning_rate": 2.8984666390385413e-06, + "loss": 0.5214, + "step": 5072 + }, + { + "epoch": 0.4204724409448819, + "grad_norm": 0.24644364416599274, + "learning_rate": 2.898052217157066e-06, + "loss": 0.493, + "step": 5073 + }, + { + "epoch": 0.42055532532117695, + "grad_norm": 0.2819516360759735, + "learning_rate": 2.8976377952755906e-06, + "loss": 0.4943, + "step": 5074 + }, + { + "epoch": 0.420638209697472, + "grad_norm": 0.23194396495819092, + "learning_rate": 2.897223373394115e-06, + "loss": 0.4941, + "step": 5075 + }, + { + "epoch": 0.4207210940737671, + "grad_norm": 0.2611530125141144, + "learning_rate": 2.8968089515126403e-06, + "loss": 0.511, + "step": 5076 + }, + { + "epoch": 0.4208039784500622, + "grad_norm": 0.2582188844680786, + "learning_rate": 2.8963945296311648e-06, + "loss": 0.5228, + "step": 5077 + }, + { + "epoch": 0.42088686282635723, + "grad_norm": 0.2546788156032562, + "learning_rate": 2.8959801077496896e-06, + "loss": 0.4937, + "step": 5078 + }, + { + "epoch": 0.4209697472026523, + "grad_norm": 0.25850844383239746, + "learning_rate": 2.895565685868214e-06, + "loss": 0.5399, + "step": 5079 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 0.25153589248657227, + "learning_rate": 2.8951512639867385e-06, + "loss": 0.5498, + "step": 5080 + }, + { + "epoch": 0.42113551595524246, + "grad_norm": 0.25599774718284607, + "learning_rate": 2.8947368421052634e-06, + "loss": 0.5239, + "step": 5081 + }, + { + "epoch": 0.4212184003315375, + "grad_norm": 0.24726103246212006, + "learning_rate": 2.8943224202237878e-06, + "loss": 0.4933, + "step": 5082 + }, + { + "epoch": 0.42130128470783257, + "grad_norm": 0.2592427134513855, + "learning_rate": 2.893907998342313e-06, + "loss": 0.5081, + "step": 5083 + }, + { + "epoch": 0.4213841690841276, + "grad_norm": 0.26935988664627075, + "learning_rate": 2.8934935764608375e-06, + "loss": 0.5276, + "step": 5084 + }, + { + "epoch": 0.42146705346042274, + "grad_norm": 0.2632657289505005, + "learning_rate": 2.893079154579362e-06, + "loss": 0.5032, + "step": 5085 + }, + { + "epoch": 0.4215499378367178, + "grad_norm": 0.2515704035758972, + "learning_rate": 2.8926647326978868e-06, + "loss": 0.5228, + "step": 5086 + }, + { + "epoch": 0.42163282221301285, + "grad_norm": 0.25346317887306213, + "learning_rate": 2.892250310816411e-06, + "loss": 0.5154, + "step": 5087 + }, + { + "epoch": 0.4217157065893079, + "grad_norm": 0.26209551095962524, + "learning_rate": 2.891835888934936e-06, + "loss": 0.5129, + "step": 5088 + }, + { + "epoch": 0.42179859096560296, + "grad_norm": 0.2657172381877899, + "learning_rate": 2.891421467053461e-06, + "loss": 0.5167, + "step": 5089 + }, + { + "epoch": 0.42188147534189807, + "grad_norm": 0.27805912494659424, + "learning_rate": 2.8910070451719854e-06, + "loss": 0.5463, + "step": 5090 + }, + { + "epoch": 0.4219643597181931, + "grad_norm": 0.2330651581287384, + "learning_rate": 2.89059262329051e-06, + "loss": 0.4624, + "step": 5091 + }, + { + "epoch": 0.4220472440944882, + "grad_norm": 0.28149813413619995, + "learning_rate": 2.8901782014090346e-06, + "loss": 0.5201, + "step": 5092 + }, + { + "epoch": 0.42213012847078324, + "grad_norm": 0.290612131357193, + "learning_rate": 2.889763779527559e-06, + "loss": 0.5845, + "step": 5093 + }, + { + "epoch": 0.42221301284707835, + "grad_norm": 0.2523595988750458, + "learning_rate": 2.889349357646084e-06, + "loss": 0.5194, + "step": 5094 + }, + { + "epoch": 0.4222958972233734, + "grad_norm": 0.25151658058166504, + "learning_rate": 2.8889349357646084e-06, + "loss": 0.527, + "step": 5095 + }, + { + "epoch": 0.42237878159966846, + "grad_norm": 0.25810274481773376, + "learning_rate": 2.8885205138831336e-06, + "loss": 0.5229, + "step": 5096 + }, + { + "epoch": 0.4224616659759635, + "grad_norm": 0.2767379879951477, + "learning_rate": 2.888106092001658e-06, + "loss": 0.575, + "step": 5097 + }, + { + "epoch": 0.4225445503522586, + "grad_norm": 0.2696339786052704, + "learning_rate": 2.8876916701201825e-06, + "loss": 0.5475, + "step": 5098 + }, + { + "epoch": 0.4226274347285537, + "grad_norm": 0.25393548607826233, + "learning_rate": 2.8872772482387074e-06, + "loss": 0.5082, + "step": 5099 + }, + { + "epoch": 0.42271031910484874, + "grad_norm": 0.27789953351020813, + "learning_rate": 2.886862826357232e-06, + "loss": 0.5354, + "step": 5100 + }, + { + "epoch": 0.4227932034811438, + "grad_norm": 0.23126012086868286, + "learning_rate": 2.8864484044757566e-06, + "loss": 0.4594, + "step": 5101 + }, + { + "epoch": 0.42287608785743885, + "grad_norm": 0.2550865113735199, + "learning_rate": 2.886033982594281e-06, + "loss": 0.5381, + "step": 5102 + }, + { + "epoch": 0.42295897223373397, + "grad_norm": 0.24243856966495514, + "learning_rate": 2.8856195607128055e-06, + "loss": 0.5164, + "step": 5103 + }, + { + "epoch": 0.423041856610029, + "grad_norm": 0.2529657781124115, + "learning_rate": 2.8852051388313308e-06, + "loss": 0.5314, + "step": 5104 + }, + { + "epoch": 0.4231247409863241, + "grad_norm": 0.23253823816776276, + "learning_rate": 2.8847907169498552e-06, + "loss": 0.491, + "step": 5105 + }, + { + "epoch": 0.42320762536261913, + "grad_norm": 0.25677579641342163, + "learning_rate": 2.8843762950683797e-06, + "loss": 0.5313, + "step": 5106 + }, + { + "epoch": 0.4232905097389142, + "grad_norm": 0.22751082479953766, + "learning_rate": 2.8839618731869045e-06, + "loss": 0.4789, + "step": 5107 + }, + { + "epoch": 0.4233733941152093, + "grad_norm": 0.2653357982635498, + "learning_rate": 2.883547451305429e-06, + "loss": 0.5138, + "step": 5108 + }, + { + "epoch": 0.42345627849150436, + "grad_norm": 0.25560301542282104, + "learning_rate": 2.8831330294239542e-06, + "loss": 0.5298, + "step": 5109 + }, + { + "epoch": 0.4235391628677994, + "grad_norm": 0.31059733033180237, + "learning_rate": 2.8827186075424786e-06, + "loss": 0.5814, + "step": 5110 + }, + { + "epoch": 0.42362204724409447, + "grad_norm": 0.2409559190273285, + "learning_rate": 2.882304185661003e-06, + "loss": 0.4811, + "step": 5111 + }, + { + "epoch": 0.4237049316203896, + "grad_norm": 0.24284924566745758, + "learning_rate": 2.881889763779528e-06, + "loss": 0.5366, + "step": 5112 + }, + { + "epoch": 0.42378781599668464, + "grad_norm": 0.2325243502855301, + "learning_rate": 2.8814753418980524e-06, + "loss": 0.486, + "step": 5113 + }, + { + "epoch": 0.4238707003729797, + "grad_norm": 0.26954254508018494, + "learning_rate": 2.8810609200165772e-06, + "loss": 0.522, + "step": 5114 + }, + { + "epoch": 0.42395358474927475, + "grad_norm": 0.25340160727500916, + "learning_rate": 2.8806464981351017e-06, + "loss": 0.5103, + "step": 5115 + }, + { + "epoch": 0.4240364691255698, + "grad_norm": 0.24098297953605652, + "learning_rate": 2.880232076253626e-06, + "loss": 0.5429, + "step": 5116 + }, + { + "epoch": 0.4241193535018649, + "grad_norm": 0.25702184438705444, + "learning_rate": 2.8798176543721514e-06, + "loss": 0.5403, + "step": 5117 + }, + { + "epoch": 0.42420223787815997, + "grad_norm": 0.2381190061569214, + "learning_rate": 2.879403232490676e-06, + "loss": 0.5139, + "step": 5118 + }, + { + "epoch": 0.424285122254455, + "grad_norm": 0.26775383949279785, + "learning_rate": 2.8789888106092007e-06, + "loss": 0.5255, + "step": 5119 + }, + { + "epoch": 0.4243680066307501, + "grad_norm": 0.2490263432264328, + "learning_rate": 2.878574388727725e-06, + "loss": 0.5044, + "step": 5120 + }, + { + "epoch": 0.4244508910070452, + "grad_norm": 0.24072401225566864, + "learning_rate": 2.8781599668462495e-06, + "loss": 0.5405, + "step": 5121 + }, + { + "epoch": 0.42453377538334025, + "grad_norm": 0.2485877424478531, + "learning_rate": 2.8777455449647744e-06, + "loss": 0.494, + "step": 5122 + }, + { + "epoch": 0.4246166597596353, + "grad_norm": 0.24746328592300415, + "learning_rate": 2.877331123083299e-06, + "loss": 0.5376, + "step": 5123 + }, + { + "epoch": 0.42469954413593036, + "grad_norm": 0.2506711781024933, + "learning_rate": 2.8769167012018232e-06, + "loss": 0.5609, + "step": 5124 + }, + { + "epoch": 0.4247824285122254, + "grad_norm": 0.2692551016807556, + "learning_rate": 2.8765022793203485e-06, + "loss": 0.5355, + "step": 5125 + }, + { + "epoch": 0.42486531288852053, + "grad_norm": 0.2747463583946228, + "learning_rate": 2.876087857438873e-06, + "loss": 0.5448, + "step": 5126 + }, + { + "epoch": 0.4249481972648156, + "grad_norm": 0.2534034550189972, + "learning_rate": 2.875673435557398e-06, + "loss": 0.4923, + "step": 5127 + }, + { + "epoch": 0.42503108164111064, + "grad_norm": 0.24879693984985352, + "learning_rate": 2.8752590136759222e-06, + "loss": 0.5223, + "step": 5128 + }, + { + "epoch": 0.4251139660174057, + "grad_norm": 0.24395544826984406, + "learning_rate": 2.8748445917944467e-06, + "loss": 0.4818, + "step": 5129 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 0.2522708773612976, + "learning_rate": 2.874430169912972e-06, + "loss": 0.5324, + "step": 5130 + }, + { + "epoch": 0.42527973476999587, + "grad_norm": 0.286657452583313, + "learning_rate": 2.8740157480314964e-06, + "loss": 0.587, + "step": 5131 + }, + { + "epoch": 0.4253626191462909, + "grad_norm": 0.28010645508766174, + "learning_rate": 2.8736013261500212e-06, + "loss": 0.5189, + "step": 5132 + }, + { + "epoch": 0.425445503522586, + "grad_norm": 0.2602810859680176, + "learning_rate": 2.8731869042685457e-06, + "loss": 0.5348, + "step": 5133 + }, + { + "epoch": 0.4255283878988811, + "grad_norm": 0.264132559299469, + "learning_rate": 2.87277248238707e-06, + "loss": 0.5089, + "step": 5134 + }, + { + "epoch": 0.42561127227517614, + "grad_norm": 0.2609390616416931, + "learning_rate": 2.872358060505595e-06, + "loss": 0.5002, + "step": 5135 + }, + { + "epoch": 0.4256941566514712, + "grad_norm": 0.2785664498806, + "learning_rate": 2.8719436386241194e-06, + "loss": 0.519, + "step": 5136 + }, + { + "epoch": 0.42577704102776626, + "grad_norm": 0.2511146664619446, + "learning_rate": 2.8715292167426447e-06, + "loss": 0.4971, + "step": 5137 + }, + { + "epoch": 0.4258599254040613, + "grad_norm": 0.24168995022773743, + "learning_rate": 2.871114794861169e-06, + "loss": 0.5072, + "step": 5138 + }, + { + "epoch": 0.4259428097803564, + "grad_norm": 0.2527907192707062, + "learning_rate": 2.8707003729796935e-06, + "loss": 0.5616, + "step": 5139 + }, + { + "epoch": 0.4260256941566515, + "grad_norm": 0.24739032983779907, + "learning_rate": 2.8702859510982184e-06, + "loss": 0.5007, + "step": 5140 + }, + { + "epoch": 0.42610857853294654, + "grad_norm": 0.25174733996391296, + "learning_rate": 2.869871529216743e-06, + "loss": 0.5124, + "step": 5141 + }, + { + "epoch": 0.4261914629092416, + "grad_norm": 0.22921964526176453, + "learning_rate": 2.8694571073352672e-06, + "loss": 0.4791, + "step": 5142 + }, + { + "epoch": 0.4262743472855367, + "grad_norm": 0.26318979263305664, + "learning_rate": 2.869042685453792e-06, + "loss": 0.5155, + "step": 5143 + }, + { + "epoch": 0.42635723166183176, + "grad_norm": 0.24979138374328613, + "learning_rate": 2.8686282635723165e-06, + "loss": 0.5183, + "step": 5144 + }, + { + "epoch": 0.4264401160381268, + "grad_norm": 0.2511877119541168, + "learning_rate": 2.868213841690842e-06, + "loss": 0.5464, + "step": 5145 + }, + { + "epoch": 0.42652300041442187, + "grad_norm": 0.24551165103912354, + "learning_rate": 2.8677994198093662e-06, + "loss": 0.4973, + "step": 5146 + }, + { + "epoch": 0.4266058847907169, + "grad_norm": 0.2570388913154602, + "learning_rate": 2.8673849979278907e-06, + "loss": 0.5115, + "step": 5147 + }, + { + "epoch": 0.42668876916701204, + "grad_norm": 0.24681341648101807, + "learning_rate": 2.8669705760464155e-06, + "loss": 0.5378, + "step": 5148 + }, + { + "epoch": 0.4267716535433071, + "grad_norm": 0.25318312644958496, + "learning_rate": 2.86655615416494e-06, + "loss": 0.5392, + "step": 5149 + }, + { + "epoch": 0.42685453791960215, + "grad_norm": 0.23577818274497986, + "learning_rate": 2.8661417322834652e-06, + "loss": 0.4986, + "step": 5150 + }, + { + "epoch": 0.4269374222958972, + "grad_norm": 0.2565193474292755, + "learning_rate": 2.8657273104019897e-06, + "loss": 0.5515, + "step": 5151 + }, + { + "epoch": 0.4270203066721923, + "grad_norm": 0.24810096621513367, + "learning_rate": 2.865312888520514e-06, + "loss": 0.5274, + "step": 5152 + }, + { + "epoch": 0.4271031910484874, + "grad_norm": 0.2618880569934845, + "learning_rate": 2.864898466639039e-06, + "loss": 0.5292, + "step": 5153 + }, + { + "epoch": 0.42718607542478243, + "grad_norm": 0.26635733246803284, + "learning_rate": 2.8644840447575634e-06, + "loss": 0.5063, + "step": 5154 + }, + { + "epoch": 0.4272689598010775, + "grad_norm": 0.22996778786182404, + "learning_rate": 2.8640696228760882e-06, + "loss": 0.4923, + "step": 5155 + }, + { + "epoch": 0.42735184417737254, + "grad_norm": 0.24530339241027832, + "learning_rate": 2.8636552009946127e-06, + "loss": 0.5054, + "step": 5156 + }, + { + "epoch": 0.42743472855366765, + "grad_norm": 0.24659645557403564, + "learning_rate": 2.863240779113137e-06, + "loss": 0.5147, + "step": 5157 + }, + { + "epoch": 0.4275176129299627, + "grad_norm": 0.2702854871749878, + "learning_rate": 2.8628263572316624e-06, + "loss": 0.5472, + "step": 5158 + }, + { + "epoch": 0.42760049730625777, + "grad_norm": 0.26752224564552307, + "learning_rate": 2.862411935350187e-06, + "loss": 0.5469, + "step": 5159 + }, + { + "epoch": 0.4276833816825528, + "grad_norm": 0.25127285718917847, + "learning_rate": 2.8619975134687112e-06, + "loss": 0.5309, + "step": 5160 + }, + { + "epoch": 0.42776626605884793, + "grad_norm": 0.243705615401268, + "learning_rate": 2.861583091587236e-06, + "loss": 0.5529, + "step": 5161 + }, + { + "epoch": 0.427849150435143, + "grad_norm": 0.26306483149528503, + "learning_rate": 2.8611686697057605e-06, + "loss": 0.5598, + "step": 5162 + }, + { + "epoch": 0.42793203481143804, + "grad_norm": 0.27829232811927795, + "learning_rate": 2.8607542478242854e-06, + "loss": 0.4904, + "step": 5163 + }, + { + "epoch": 0.4280149191877331, + "grad_norm": 0.26011258363723755, + "learning_rate": 2.86033982594281e-06, + "loss": 0.5139, + "step": 5164 + }, + { + "epoch": 0.42809780356402816, + "grad_norm": 0.2796385884284973, + "learning_rate": 2.8599254040613343e-06, + "loss": 0.5184, + "step": 5165 + }, + { + "epoch": 0.42818068794032327, + "grad_norm": 0.2857528626918793, + "learning_rate": 2.8595109821798595e-06, + "loss": 0.5398, + "step": 5166 + }, + { + "epoch": 0.4282635723166183, + "grad_norm": 0.22451595962047577, + "learning_rate": 2.859096560298384e-06, + "loss": 0.4645, + "step": 5167 + }, + { + "epoch": 0.4283464566929134, + "grad_norm": 0.28474822640419006, + "learning_rate": 2.858682138416909e-06, + "loss": 0.5795, + "step": 5168 + }, + { + "epoch": 0.42842934106920844, + "grad_norm": 0.2574058771133423, + "learning_rate": 2.8582677165354333e-06, + "loss": 0.5078, + "step": 5169 + }, + { + "epoch": 0.42851222544550355, + "grad_norm": 0.2534273862838745, + "learning_rate": 2.8578532946539577e-06, + "loss": 0.5447, + "step": 5170 + }, + { + "epoch": 0.4285951098217986, + "grad_norm": 0.24811983108520508, + "learning_rate": 2.857438872772483e-06, + "loss": 0.5353, + "step": 5171 + }, + { + "epoch": 0.42867799419809366, + "grad_norm": 0.23818939924240112, + "learning_rate": 2.8570244508910074e-06, + "loss": 0.5069, + "step": 5172 + }, + { + "epoch": 0.4287608785743887, + "grad_norm": 0.2518976926803589, + "learning_rate": 2.856610029009532e-06, + "loss": 0.5411, + "step": 5173 + }, + { + "epoch": 0.42884376295068377, + "grad_norm": 0.2543903887271881, + "learning_rate": 2.8561956071280567e-06, + "loss": 0.5288, + "step": 5174 + }, + { + "epoch": 0.4289266473269789, + "grad_norm": 0.2446458339691162, + "learning_rate": 2.855781185246581e-06, + "loss": 0.5327, + "step": 5175 + }, + { + "epoch": 0.42900953170327394, + "grad_norm": 0.25038087368011475, + "learning_rate": 2.855366763365106e-06, + "loss": 0.5289, + "step": 5176 + }, + { + "epoch": 0.429092416079569, + "grad_norm": 0.2704561948776245, + "learning_rate": 2.8549523414836304e-06, + "loss": 0.5259, + "step": 5177 + }, + { + "epoch": 0.42917530045586405, + "grad_norm": 0.2569301724433899, + "learning_rate": 2.854537919602155e-06, + "loss": 0.5209, + "step": 5178 + }, + { + "epoch": 0.42925818483215916, + "grad_norm": 0.23300209641456604, + "learning_rate": 2.85412349772068e-06, + "loss": 0.4943, + "step": 5179 + }, + { + "epoch": 0.4293410692084542, + "grad_norm": 0.2447858601808548, + "learning_rate": 2.8537090758392045e-06, + "loss": 0.5131, + "step": 5180 + }, + { + "epoch": 0.4294239535847493, + "grad_norm": 0.32370519638061523, + "learning_rate": 2.8532946539577294e-06, + "loss": 0.5302, + "step": 5181 + }, + { + "epoch": 0.42950683796104433, + "grad_norm": 0.26094532012939453, + "learning_rate": 2.852880232076254e-06, + "loss": 0.5357, + "step": 5182 + }, + { + "epoch": 0.4295897223373394, + "grad_norm": 0.26513925194740295, + "learning_rate": 2.8524658101947783e-06, + "loss": 0.4942, + "step": 5183 + }, + { + "epoch": 0.4296726067136345, + "grad_norm": 0.24182002246379852, + "learning_rate": 2.852051388313303e-06, + "loss": 0.5256, + "step": 5184 + }, + { + "epoch": 0.42975549108992955, + "grad_norm": 0.23416733741760254, + "learning_rate": 2.8516369664318276e-06, + "loss": 0.5316, + "step": 5185 + }, + { + "epoch": 0.4298383754662246, + "grad_norm": 0.2858116626739502, + "learning_rate": 2.851222544550353e-06, + "loss": 0.5889, + "step": 5186 + }, + { + "epoch": 0.42992125984251967, + "grad_norm": 0.26866042613983154, + "learning_rate": 2.8508081226688773e-06, + "loss": 0.5297, + "step": 5187 + }, + { + "epoch": 0.4300041442188148, + "grad_norm": 0.23610661923885345, + "learning_rate": 2.8503937007874017e-06, + "loss": 0.5338, + "step": 5188 + }, + { + "epoch": 0.43008702859510983, + "grad_norm": 0.25178593397140503, + "learning_rate": 2.8499792789059265e-06, + "loss": 0.5294, + "step": 5189 + }, + { + "epoch": 0.4301699129714049, + "grad_norm": 0.24090343713760376, + "learning_rate": 2.849564857024451e-06, + "loss": 0.4915, + "step": 5190 + }, + { + "epoch": 0.43025279734769994, + "grad_norm": 0.2836218476295471, + "learning_rate": 2.8491504351429754e-06, + "loss": 0.5703, + "step": 5191 + }, + { + "epoch": 0.430335681723995, + "grad_norm": 0.2588792145252228, + "learning_rate": 2.8487360132615007e-06, + "loss": 0.5025, + "step": 5192 + }, + { + "epoch": 0.4304185661002901, + "grad_norm": 0.260219007730484, + "learning_rate": 2.848321591380025e-06, + "loss": 0.56, + "step": 5193 + }, + { + "epoch": 0.43050145047658517, + "grad_norm": 0.27483734488487244, + "learning_rate": 2.84790716949855e-06, + "loss": 0.5699, + "step": 5194 + }, + { + "epoch": 0.4305843348528802, + "grad_norm": 0.24816079437732697, + "learning_rate": 2.8474927476170744e-06, + "loss": 0.488, + "step": 5195 + }, + { + "epoch": 0.4306672192291753, + "grad_norm": 0.2508851885795593, + "learning_rate": 2.847078325735599e-06, + "loss": 0.537, + "step": 5196 + }, + { + "epoch": 0.4307501036054704, + "grad_norm": 0.2656272053718567, + "learning_rate": 2.8466639038541237e-06, + "loss": 0.5444, + "step": 5197 + }, + { + "epoch": 0.43083298798176545, + "grad_norm": 0.22687986493110657, + "learning_rate": 2.846249481972648e-06, + "loss": 0.427, + "step": 5198 + }, + { + "epoch": 0.4309158723580605, + "grad_norm": 0.23955410718917847, + "learning_rate": 2.8458350600911734e-06, + "loss": 0.5095, + "step": 5199 + }, + { + "epoch": 0.43099875673435556, + "grad_norm": 0.2716139256954193, + "learning_rate": 2.845420638209698e-06, + "loss": 0.5236, + "step": 5200 + }, + { + "epoch": 0.4310816411106506, + "grad_norm": 0.2780202031135559, + "learning_rate": 2.8450062163282223e-06, + "loss": 0.5239, + "step": 5201 + }, + { + "epoch": 0.4311645254869457, + "grad_norm": 0.2852136492729187, + "learning_rate": 2.844591794446747e-06, + "loss": 0.5456, + "step": 5202 + }, + { + "epoch": 0.4312474098632408, + "grad_norm": 0.27588772773742676, + "learning_rate": 2.8441773725652716e-06, + "loss": 0.5184, + "step": 5203 + }, + { + "epoch": 0.43133029423953584, + "grad_norm": 0.2455872893333435, + "learning_rate": 2.8437629506837964e-06, + "loss": 0.5071, + "step": 5204 + }, + { + "epoch": 0.4314131786158309, + "grad_norm": 0.25020894408226013, + "learning_rate": 2.843348528802321e-06, + "loss": 0.4925, + "step": 5205 + }, + { + "epoch": 0.431496062992126, + "grad_norm": 0.21416687965393066, + "learning_rate": 2.8429341069208453e-06, + "loss": 0.4232, + "step": 5206 + }, + { + "epoch": 0.43157894736842106, + "grad_norm": 0.2606964409351349, + "learning_rate": 2.8425196850393706e-06, + "loss": 0.489, + "step": 5207 + }, + { + "epoch": 0.4316618317447161, + "grad_norm": 0.24258488416671753, + "learning_rate": 2.842105263157895e-06, + "loss": 0.5182, + "step": 5208 + }, + { + "epoch": 0.4317447161210112, + "grad_norm": 0.24004611372947693, + "learning_rate": 2.8416908412764194e-06, + "loss": 0.4873, + "step": 5209 + }, + { + "epoch": 0.4318276004973063, + "grad_norm": 0.2737581729888916, + "learning_rate": 2.8412764193949443e-06, + "loss": 0.5356, + "step": 5210 + }, + { + "epoch": 0.43191048487360134, + "grad_norm": 0.23726293444633484, + "learning_rate": 2.8408619975134687e-06, + "loss": 0.4823, + "step": 5211 + }, + { + "epoch": 0.4319933692498964, + "grad_norm": 0.24587492644786835, + "learning_rate": 2.840447575631994e-06, + "loss": 0.5455, + "step": 5212 + }, + { + "epoch": 0.43207625362619145, + "grad_norm": 0.24350252747535706, + "learning_rate": 2.8400331537505184e-06, + "loss": 0.5161, + "step": 5213 + }, + { + "epoch": 0.4321591380024865, + "grad_norm": 0.2576490044593811, + "learning_rate": 2.839618731869043e-06, + "loss": 0.5386, + "step": 5214 + }, + { + "epoch": 0.4322420223787816, + "grad_norm": 0.2501968443393707, + "learning_rate": 2.8392043099875677e-06, + "loss": 0.5397, + "step": 5215 + }, + { + "epoch": 0.4323249067550767, + "grad_norm": 0.24447616934776306, + "learning_rate": 2.838789888106092e-06, + "loss": 0.5093, + "step": 5216 + }, + { + "epoch": 0.43240779113137173, + "grad_norm": 0.24189409613609314, + "learning_rate": 2.838375466224617e-06, + "loss": 0.488, + "step": 5217 + }, + { + "epoch": 0.4324906755076668, + "grad_norm": 0.23336447775363922, + "learning_rate": 2.8379610443431414e-06, + "loss": 0.4874, + "step": 5218 + }, + { + "epoch": 0.4325735598839619, + "grad_norm": 0.2607559263706207, + "learning_rate": 2.837546622461666e-06, + "loss": 0.5262, + "step": 5219 + }, + { + "epoch": 0.43265644426025696, + "grad_norm": 0.2593734860420227, + "learning_rate": 2.837132200580191e-06, + "loss": 0.5303, + "step": 5220 + }, + { + "epoch": 0.432739328636552, + "grad_norm": 0.2491120547056198, + "learning_rate": 2.8367177786987156e-06, + "loss": 0.5368, + "step": 5221 + }, + { + "epoch": 0.43282221301284707, + "grad_norm": 0.2524268925189972, + "learning_rate": 2.8363033568172404e-06, + "loss": 0.5527, + "step": 5222 + }, + { + "epoch": 0.4329050973891421, + "grad_norm": 0.260125994682312, + "learning_rate": 2.835888934935765e-06, + "loss": 0.5106, + "step": 5223 + }, + { + "epoch": 0.43298798176543724, + "grad_norm": 0.25186148285865784, + "learning_rate": 2.8354745130542893e-06, + "loss": 0.5164, + "step": 5224 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 0.24863828718662262, + "learning_rate": 2.835060091172814e-06, + "loss": 0.5166, + "step": 5225 + }, + { + "epoch": 0.43315375051802735, + "grad_norm": 0.29166486859321594, + "learning_rate": 2.8346456692913386e-06, + "loss": 0.5545, + "step": 5226 + }, + { + "epoch": 0.4332366348943224, + "grad_norm": 0.2459433525800705, + "learning_rate": 2.8342312474098634e-06, + "loss": 0.4917, + "step": 5227 + }, + { + "epoch": 0.4333195192706175, + "grad_norm": 0.2561839520931244, + "learning_rate": 2.8338168255283883e-06, + "loss": 0.4959, + "step": 5228 + }, + { + "epoch": 0.43340240364691257, + "grad_norm": 0.2617505192756653, + "learning_rate": 2.8334024036469127e-06, + "loss": 0.5401, + "step": 5229 + }, + { + "epoch": 0.4334852880232076, + "grad_norm": 0.26690909266471863, + "learning_rate": 2.8329879817654376e-06, + "loss": 0.5279, + "step": 5230 + }, + { + "epoch": 0.4335681723995027, + "grad_norm": 0.23328612744808197, + "learning_rate": 2.832573559883962e-06, + "loss": 0.4859, + "step": 5231 + }, + { + "epoch": 0.43365105677579774, + "grad_norm": 0.23995989561080933, + "learning_rate": 2.8321591380024864e-06, + "loss": 0.5012, + "step": 5232 + }, + { + "epoch": 0.43373394115209285, + "grad_norm": 0.25016462802886963, + "learning_rate": 2.8317447161210117e-06, + "loss": 0.5564, + "step": 5233 + }, + { + "epoch": 0.4338168255283879, + "grad_norm": 0.2667665183544159, + "learning_rate": 2.831330294239536e-06, + "loss": 0.5266, + "step": 5234 + }, + { + "epoch": 0.43389970990468296, + "grad_norm": 0.2574343979358673, + "learning_rate": 2.830915872358061e-06, + "loss": 0.5144, + "step": 5235 + }, + { + "epoch": 0.433982594280978, + "grad_norm": 0.2572869062423706, + "learning_rate": 2.8305014504765854e-06, + "loss": 0.4874, + "step": 5236 + }, + { + "epoch": 0.43406547865727313, + "grad_norm": 0.24457313120365143, + "learning_rate": 2.83008702859511e-06, + "loss": 0.5088, + "step": 5237 + }, + { + "epoch": 0.4341483630335682, + "grad_norm": 0.24505515396595, + "learning_rate": 2.8296726067136347e-06, + "loss": 0.5094, + "step": 5238 + }, + { + "epoch": 0.43423124740986324, + "grad_norm": 0.2541201114654541, + "learning_rate": 2.829258184832159e-06, + "loss": 0.5123, + "step": 5239 + }, + { + "epoch": 0.4343141317861583, + "grad_norm": 0.27587270736694336, + "learning_rate": 2.8288437629506844e-06, + "loss": 0.5115, + "step": 5240 + }, + { + "epoch": 0.43439701616245335, + "grad_norm": 0.24641378223896027, + "learning_rate": 2.828429341069209e-06, + "loss": 0.4882, + "step": 5241 + }, + { + "epoch": 0.43447990053874846, + "grad_norm": 0.2696695625782013, + "learning_rate": 2.8280149191877333e-06, + "loss": 0.6089, + "step": 5242 + }, + { + "epoch": 0.4345627849150435, + "grad_norm": 0.26147153973579407, + "learning_rate": 2.827600497306258e-06, + "loss": 0.498, + "step": 5243 + }, + { + "epoch": 0.4346456692913386, + "grad_norm": 0.23661236464977264, + "learning_rate": 2.8271860754247826e-06, + "loss": 0.5461, + "step": 5244 + }, + { + "epoch": 0.43472855366763363, + "grad_norm": 0.2544079124927521, + "learning_rate": 2.826771653543307e-06, + "loss": 0.51, + "step": 5245 + }, + { + "epoch": 0.43481143804392874, + "grad_norm": 0.2562776505947113, + "learning_rate": 2.826357231661832e-06, + "loss": 0.5255, + "step": 5246 + }, + { + "epoch": 0.4348943224202238, + "grad_norm": 0.2469797283411026, + "learning_rate": 2.8259428097803567e-06, + "loss": 0.4846, + "step": 5247 + }, + { + "epoch": 0.43497720679651886, + "grad_norm": 0.23846235871315002, + "learning_rate": 2.8255283878988816e-06, + "loss": 0.4902, + "step": 5248 + }, + { + "epoch": 0.4350600911728139, + "grad_norm": 0.23190627992153168, + "learning_rate": 2.825113966017406e-06, + "loss": 0.4975, + "step": 5249 + }, + { + "epoch": 0.43514297554910897, + "grad_norm": 0.24921630322933197, + "learning_rate": 2.8246995441359304e-06, + "loss": 0.4651, + "step": 5250 + }, + { + "epoch": 0.4352258599254041, + "grad_norm": 0.2520318925380707, + "learning_rate": 2.8242851222544553e-06, + "loss": 0.5291, + "step": 5251 + }, + { + "epoch": 0.43530874430169914, + "grad_norm": 0.2578127086162567, + "learning_rate": 2.8238707003729797e-06, + "loss": 0.5404, + "step": 5252 + }, + { + "epoch": 0.4353916286779942, + "grad_norm": 0.24403531849384308, + "learning_rate": 2.823456278491505e-06, + "loss": 0.4744, + "step": 5253 + }, + { + "epoch": 0.43547451305428925, + "grad_norm": 0.2665512263774872, + "learning_rate": 2.8230418566100294e-06, + "loss": 0.5284, + "step": 5254 + }, + { + "epoch": 0.43555739743058436, + "grad_norm": 0.2598167955875397, + "learning_rate": 2.822627434728554e-06, + "loss": 0.515, + "step": 5255 + }, + { + "epoch": 0.4356402818068794, + "grad_norm": 0.2378029227256775, + "learning_rate": 2.8222130128470787e-06, + "loss": 0.4797, + "step": 5256 + }, + { + "epoch": 0.43572316618317447, + "grad_norm": 0.25250139832496643, + "learning_rate": 2.821798590965603e-06, + "loss": 0.4835, + "step": 5257 + }, + { + "epoch": 0.4358060505594695, + "grad_norm": 0.28557088971138, + "learning_rate": 2.8213841690841276e-06, + "loss": 0.5799, + "step": 5258 + }, + { + "epoch": 0.4358889349357646, + "grad_norm": 0.2414311170578003, + "learning_rate": 2.8209697472026524e-06, + "loss": 0.5541, + "step": 5259 + }, + { + "epoch": 0.4359718193120597, + "grad_norm": 0.2598526179790497, + "learning_rate": 2.820555325321177e-06, + "loss": 0.5026, + "step": 5260 + }, + { + "epoch": 0.43605470368835475, + "grad_norm": 0.2608531415462494, + "learning_rate": 2.820140903439702e-06, + "loss": 0.5031, + "step": 5261 + }, + { + "epoch": 0.4361375880646498, + "grad_norm": 0.26646652817726135, + "learning_rate": 2.8197264815582266e-06, + "loss": 0.4874, + "step": 5262 + }, + { + "epoch": 0.43622047244094486, + "grad_norm": 0.26478347182273865, + "learning_rate": 2.819312059676751e-06, + "loss": 0.5485, + "step": 5263 + }, + { + "epoch": 0.43630335681724, + "grad_norm": 0.2556110918521881, + "learning_rate": 2.818897637795276e-06, + "loss": 0.4669, + "step": 5264 + }, + { + "epoch": 0.43638624119353503, + "grad_norm": 0.2529478073120117, + "learning_rate": 2.8184832159138003e-06, + "loss": 0.5095, + "step": 5265 + }, + { + "epoch": 0.4364691255698301, + "grad_norm": 0.27578556537628174, + "learning_rate": 2.818068794032325e-06, + "loss": 0.5287, + "step": 5266 + }, + { + "epoch": 0.43655200994612514, + "grad_norm": 0.2566321790218353, + "learning_rate": 2.8176543721508496e-06, + "loss": 0.5146, + "step": 5267 + }, + { + "epoch": 0.4366348943224202, + "grad_norm": 0.2488023340702057, + "learning_rate": 2.8172399502693744e-06, + "loss": 0.5151, + "step": 5268 + }, + { + "epoch": 0.4367177786987153, + "grad_norm": 0.26405540108680725, + "learning_rate": 2.8168255283878993e-06, + "loss": 0.5397, + "step": 5269 + }, + { + "epoch": 0.43680066307501036, + "grad_norm": 0.25462543964385986, + "learning_rate": 2.8164111065064237e-06, + "loss": 0.5483, + "step": 5270 + }, + { + "epoch": 0.4368835474513054, + "grad_norm": 0.2762688994407654, + "learning_rate": 2.8159966846249486e-06, + "loss": 0.5331, + "step": 5271 + }, + { + "epoch": 0.4369664318276005, + "grad_norm": 0.2610211968421936, + "learning_rate": 2.815582262743473e-06, + "loss": 0.5109, + "step": 5272 + }, + { + "epoch": 0.4370493162038956, + "grad_norm": 0.26340895891189575, + "learning_rate": 2.8151678408619975e-06, + "loss": 0.4684, + "step": 5273 + }, + { + "epoch": 0.43713220058019064, + "grad_norm": 0.2525016963481903, + "learning_rate": 2.8147534189805227e-06, + "loss": 0.4732, + "step": 5274 + }, + { + "epoch": 0.4372150849564857, + "grad_norm": 0.26172879338264465, + "learning_rate": 2.814338997099047e-06, + "loss": 0.5473, + "step": 5275 + }, + { + "epoch": 0.43729796933278076, + "grad_norm": 0.2737736701965332, + "learning_rate": 2.8139245752175716e-06, + "loss": 0.529, + "step": 5276 + }, + { + "epoch": 0.4373808537090758, + "grad_norm": 0.2516099512577057, + "learning_rate": 2.8135101533360965e-06, + "loss": 0.5146, + "step": 5277 + }, + { + "epoch": 0.4374637380853709, + "grad_norm": 0.22976182401180267, + "learning_rate": 2.813095731454621e-06, + "loss": 0.4646, + "step": 5278 + }, + { + "epoch": 0.437546622461666, + "grad_norm": 0.250583678483963, + "learning_rate": 2.8126813095731457e-06, + "loss": 0.5135, + "step": 5279 + }, + { + "epoch": 0.43762950683796104, + "grad_norm": 0.25816264748573303, + "learning_rate": 2.81226688769167e-06, + "loss": 0.5236, + "step": 5280 + }, + { + "epoch": 0.4377123912142561, + "grad_norm": 0.25602301955223083, + "learning_rate": 2.8118524658101946e-06, + "loss": 0.5019, + "step": 5281 + }, + { + "epoch": 0.4377952755905512, + "grad_norm": 0.2638482451438904, + "learning_rate": 2.81143804392872e-06, + "loss": 0.5256, + "step": 5282 + }, + { + "epoch": 0.43787815996684626, + "grad_norm": 0.2708737254142761, + "learning_rate": 2.8110236220472443e-06, + "loss": 0.576, + "step": 5283 + }, + { + "epoch": 0.4379610443431413, + "grad_norm": 0.26806768774986267, + "learning_rate": 2.810609200165769e-06, + "loss": 0.5502, + "step": 5284 + }, + { + "epoch": 0.43804392871943637, + "grad_norm": 0.250840961933136, + "learning_rate": 2.8101947782842936e-06, + "loss": 0.5497, + "step": 5285 + }, + { + "epoch": 0.4381268130957315, + "grad_norm": 0.24293595552444458, + "learning_rate": 2.809780356402818e-06, + "loss": 0.5139, + "step": 5286 + }, + { + "epoch": 0.43820969747202654, + "grad_norm": 0.2537114918231964, + "learning_rate": 2.809365934521343e-06, + "loss": 0.5405, + "step": 5287 + }, + { + "epoch": 0.4382925818483216, + "grad_norm": 0.2500779330730438, + "learning_rate": 2.8089515126398677e-06, + "loss": 0.475, + "step": 5288 + }, + { + "epoch": 0.43837546622461665, + "grad_norm": 0.2503072917461395, + "learning_rate": 2.8085370907583926e-06, + "loss": 0.5255, + "step": 5289 + }, + { + "epoch": 0.4384583506009117, + "grad_norm": 0.2497822791337967, + "learning_rate": 2.808122668876917e-06, + "loss": 0.4795, + "step": 5290 + }, + { + "epoch": 0.4385412349772068, + "grad_norm": 0.24481791257858276, + "learning_rate": 2.8077082469954415e-06, + "loss": 0.5242, + "step": 5291 + }, + { + "epoch": 0.4386241193535019, + "grad_norm": 0.2714848816394806, + "learning_rate": 2.8072938251139663e-06, + "loss": 0.5029, + "step": 5292 + }, + { + "epoch": 0.43870700372979693, + "grad_norm": 0.25696444511413574, + "learning_rate": 2.8068794032324907e-06, + "loss": 0.4877, + "step": 5293 + }, + { + "epoch": 0.438789888106092, + "grad_norm": 0.2760876417160034, + "learning_rate": 2.806464981351015e-06, + "loss": 0.5579, + "step": 5294 + }, + { + "epoch": 0.4388727724823871, + "grad_norm": 0.2579100430011749, + "learning_rate": 2.8060505594695405e-06, + "loss": 0.5271, + "step": 5295 + }, + { + "epoch": 0.43895565685868215, + "grad_norm": 0.24207231402397156, + "learning_rate": 2.805636137588065e-06, + "loss": 0.4919, + "step": 5296 + }, + { + "epoch": 0.4390385412349772, + "grad_norm": 0.23749218881130219, + "learning_rate": 2.8052217157065897e-06, + "loss": 0.4832, + "step": 5297 + }, + { + "epoch": 0.43912142561127226, + "grad_norm": 0.2804615795612335, + "learning_rate": 2.804807293825114e-06, + "loss": 0.5656, + "step": 5298 + }, + { + "epoch": 0.4392043099875673, + "grad_norm": 0.23041434586048126, + "learning_rate": 2.8043928719436386e-06, + "loss": 0.496, + "step": 5299 + }, + { + "epoch": 0.43928719436386243, + "grad_norm": 0.2514626085758209, + "learning_rate": 2.8039784500621635e-06, + "loss": 0.5429, + "step": 5300 + }, + { + "epoch": 0.4393700787401575, + "grad_norm": 0.277759850025177, + "learning_rate": 2.803564028180688e-06, + "loss": 0.5739, + "step": 5301 + }, + { + "epoch": 0.43945296311645254, + "grad_norm": 0.2679968774318695, + "learning_rate": 2.803149606299213e-06, + "loss": 0.524, + "step": 5302 + }, + { + "epoch": 0.4395358474927476, + "grad_norm": 0.2514494061470032, + "learning_rate": 2.8027351844177376e-06, + "loss": 0.4918, + "step": 5303 + }, + { + "epoch": 0.4396187318690427, + "grad_norm": 0.3018876016139984, + "learning_rate": 2.802320762536262e-06, + "loss": 0.5642, + "step": 5304 + }, + { + "epoch": 0.43970161624533777, + "grad_norm": 0.26885026693344116, + "learning_rate": 2.801906340654787e-06, + "loss": 0.5408, + "step": 5305 + }, + { + "epoch": 0.4397845006216328, + "grad_norm": 0.258227676153183, + "learning_rate": 2.8014919187733113e-06, + "loss": 0.5389, + "step": 5306 + }, + { + "epoch": 0.4398673849979279, + "grad_norm": 0.2549477219581604, + "learning_rate": 2.801077496891836e-06, + "loss": 0.5339, + "step": 5307 + }, + { + "epoch": 0.43995026937422294, + "grad_norm": 0.25693097710609436, + "learning_rate": 2.800663075010361e-06, + "loss": 0.523, + "step": 5308 + }, + { + "epoch": 0.44003315375051805, + "grad_norm": 0.2599722445011139, + "learning_rate": 2.8002486531288855e-06, + "loss": 0.5476, + "step": 5309 + }, + { + "epoch": 0.4401160381268131, + "grad_norm": 0.25842025876045227, + "learning_rate": 2.7998342312474103e-06, + "loss": 0.536, + "step": 5310 + }, + { + "epoch": 0.44019892250310816, + "grad_norm": 0.2614330053329468, + "learning_rate": 2.7994198093659348e-06, + "loss": 0.5256, + "step": 5311 + }, + { + "epoch": 0.4402818068794032, + "grad_norm": 0.2867611050605774, + "learning_rate": 2.799005387484459e-06, + "loss": 0.547, + "step": 5312 + }, + { + "epoch": 0.4403646912556983, + "grad_norm": 0.29532676935195923, + "learning_rate": 2.798590965602984e-06, + "loss": 0.4996, + "step": 5313 + }, + { + "epoch": 0.4404475756319934, + "grad_norm": 0.24556805193424225, + "learning_rate": 2.7981765437215085e-06, + "loss": 0.5086, + "step": 5314 + }, + { + "epoch": 0.44053046000828844, + "grad_norm": 0.2616688907146454, + "learning_rate": 2.7977621218400338e-06, + "loss": 0.4923, + "step": 5315 + }, + { + "epoch": 0.4406133443845835, + "grad_norm": 0.23590333759784698, + "learning_rate": 2.797347699958558e-06, + "loss": 0.4438, + "step": 5316 + }, + { + "epoch": 0.44069622876087855, + "grad_norm": 0.2733282148838043, + "learning_rate": 2.7969332780770826e-06, + "loss": 0.5394, + "step": 5317 + }, + { + "epoch": 0.44077911313717366, + "grad_norm": 0.24374346435070038, + "learning_rate": 2.7965188561956075e-06, + "loss": 0.4786, + "step": 5318 + }, + { + "epoch": 0.4408619975134687, + "grad_norm": 0.23927830159664154, + "learning_rate": 2.796104434314132e-06, + "loss": 0.4893, + "step": 5319 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 0.256038099527359, + "learning_rate": 2.7956900124326568e-06, + "loss": 0.5244, + "step": 5320 + }, + { + "epoch": 0.44102776626605883, + "grad_norm": 0.25132226943969727, + "learning_rate": 2.795275590551181e-06, + "loss": 0.4981, + "step": 5321 + }, + { + "epoch": 0.44111065064235394, + "grad_norm": 0.2560637295246124, + "learning_rate": 2.7948611686697056e-06, + "loss": 0.5612, + "step": 5322 + }, + { + "epoch": 0.441193535018649, + "grad_norm": 0.25464773178100586, + "learning_rate": 2.794446746788231e-06, + "loss": 0.5165, + "step": 5323 + }, + { + "epoch": 0.44127641939494405, + "grad_norm": 0.2857580780982971, + "learning_rate": 2.7940323249067553e-06, + "loss": 0.5645, + "step": 5324 + }, + { + "epoch": 0.4413593037712391, + "grad_norm": 0.2552124261856079, + "learning_rate": 2.7936179030252798e-06, + "loss": 0.504, + "step": 5325 + }, + { + "epoch": 0.44144218814753416, + "grad_norm": 0.2598317861557007, + "learning_rate": 2.7932034811438046e-06, + "loss": 0.496, + "step": 5326 + }, + { + "epoch": 0.4415250725238293, + "grad_norm": 0.23825901746749878, + "learning_rate": 2.792789059262329e-06, + "loss": 0.4993, + "step": 5327 + }, + { + "epoch": 0.44160795690012433, + "grad_norm": 0.25819963216781616, + "learning_rate": 2.7923746373808543e-06, + "loss": 0.5417, + "step": 5328 + }, + { + "epoch": 0.4416908412764194, + "grad_norm": 0.25931891798973083, + "learning_rate": 2.7919602154993788e-06, + "loss": 0.524, + "step": 5329 + }, + { + "epoch": 0.44177372565271444, + "grad_norm": 0.25842466950416565, + "learning_rate": 2.791545793617903e-06, + "loss": 0.5451, + "step": 5330 + }, + { + "epoch": 0.44185661002900956, + "grad_norm": 0.2428404986858368, + "learning_rate": 2.791131371736428e-06, + "loss": 0.5086, + "step": 5331 + }, + { + "epoch": 0.4419394944053046, + "grad_norm": 0.2635623514652252, + "learning_rate": 2.7907169498549525e-06, + "loss": 0.5079, + "step": 5332 + }, + { + "epoch": 0.44202237878159967, + "grad_norm": 0.26045310497283936, + "learning_rate": 2.7903025279734773e-06, + "loss": 0.5153, + "step": 5333 + }, + { + "epoch": 0.4421052631578947, + "grad_norm": 0.2424784004688263, + "learning_rate": 2.7898881060920018e-06, + "loss": 0.4975, + "step": 5334 + }, + { + "epoch": 0.4421881475341898, + "grad_norm": 0.2653915584087372, + "learning_rate": 2.789473684210526e-06, + "loss": 0.5703, + "step": 5335 + }, + { + "epoch": 0.4422710319104849, + "grad_norm": 0.2573208808898926, + "learning_rate": 2.7890592623290515e-06, + "loss": 0.5116, + "step": 5336 + }, + { + "epoch": 0.44235391628677995, + "grad_norm": 0.2432820200920105, + "learning_rate": 2.788644840447576e-06, + "loss": 0.5092, + "step": 5337 + }, + { + "epoch": 0.442436800663075, + "grad_norm": 0.2880386710166931, + "learning_rate": 2.7882304185661008e-06, + "loss": 0.5532, + "step": 5338 + }, + { + "epoch": 0.44251968503937006, + "grad_norm": 0.237730011343956, + "learning_rate": 2.787815996684625e-06, + "loss": 0.5204, + "step": 5339 + }, + { + "epoch": 0.44260256941566517, + "grad_norm": 0.23831459879875183, + "learning_rate": 2.7874015748031496e-06, + "loss": 0.4924, + "step": 5340 + }, + { + "epoch": 0.4426854537919602, + "grad_norm": 0.23899486660957336, + "learning_rate": 2.7869871529216745e-06, + "loss": 0.5178, + "step": 5341 + }, + { + "epoch": 0.4427683381682553, + "grad_norm": 0.24707293510437012, + "learning_rate": 2.786572731040199e-06, + "loss": 0.5079, + "step": 5342 + }, + { + "epoch": 0.44285122254455034, + "grad_norm": 0.2688197195529938, + "learning_rate": 2.7861583091587233e-06, + "loss": 0.5429, + "step": 5343 + }, + { + "epoch": 0.4429341069208454, + "grad_norm": 0.2548159658908844, + "learning_rate": 2.7857438872772486e-06, + "loss": 0.5141, + "step": 5344 + }, + { + "epoch": 0.4430169912971405, + "grad_norm": 0.28531527519226074, + "learning_rate": 2.785329465395773e-06, + "loss": 0.5807, + "step": 5345 + }, + { + "epoch": 0.44309987567343556, + "grad_norm": 0.26467427611351013, + "learning_rate": 2.784915043514298e-06, + "loss": 0.5417, + "step": 5346 + }, + { + "epoch": 0.4431827600497306, + "grad_norm": 0.23687094449996948, + "learning_rate": 2.7845006216328223e-06, + "loss": 0.4868, + "step": 5347 + }, + { + "epoch": 0.4432656444260257, + "grad_norm": 0.26799798011779785, + "learning_rate": 2.7840861997513468e-06, + "loss": 0.5439, + "step": 5348 + }, + { + "epoch": 0.4433485288023208, + "grad_norm": 0.25248515605926514, + "learning_rate": 2.783671777869872e-06, + "loss": 0.5055, + "step": 5349 + }, + { + "epoch": 0.44343141317861584, + "grad_norm": 0.2344471663236618, + "learning_rate": 2.7832573559883965e-06, + "loss": 0.4849, + "step": 5350 + }, + { + "epoch": 0.4435142975549109, + "grad_norm": 0.2540130913257599, + "learning_rate": 2.7828429341069213e-06, + "loss": 0.5124, + "step": 5351 + }, + { + "epoch": 0.44359718193120595, + "grad_norm": 0.25236696004867554, + "learning_rate": 2.7824285122254458e-06, + "loss": 0.4946, + "step": 5352 + }, + { + "epoch": 0.443680066307501, + "grad_norm": 0.266066312789917, + "learning_rate": 2.78201409034397e-06, + "loss": 0.5579, + "step": 5353 + }, + { + "epoch": 0.4437629506837961, + "grad_norm": 0.23375727236270905, + "learning_rate": 2.781599668462495e-06, + "loss": 0.4657, + "step": 5354 + }, + { + "epoch": 0.4438458350600912, + "grad_norm": 0.2865810692310333, + "learning_rate": 2.7811852465810195e-06, + "loss": 0.48, + "step": 5355 + }, + { + "epoch": 0.44392871943638623, + "grad_norm": 0.2666076123714447, + "learning_rate": 2.7807708246995448e-06, + "loss": 0.4904, + "step": 5356 + }, + { + "epoch": 0.4440116038126813, + "grad_norm": 0.25318020582199097, + "learning_rate": 2.780356402818069e-06, + "loss": 0.5515, + "step": 5357 + }, + { + "epoch": 0.4440944881889764, + "grad_norm": 0.2513139247894287, + "learning_rate": 2.7799419809365936e-06, + "loss": 0.5187, + "step": 5358 + }, + { + "epoch": 0.44417737256527146, + "grad_norm": 0.25219833850860596, + "learning_rate": 2.7795275590551185e-06, + "loss": 0.5284, + "step": 5359 + }, + { + "epoch": 0.4442602569415665, + "grad_norm": 0.2564052641391754, + "learning_rate": 2.779113137173643e-06, + "loss": 0.5098, + "step": 5360 + }, + { + "epoch": 0.44434314131786157, + "grad_norm": 0.2689390778541565, + "learning_rate": 2.7786987152921674e-06, + "loss": 0.533, + "step": 5361 + }, + { + "epoch": 0.4444260256941567, + "grad_norm": 0.25876477360725403, + "learning_rate": 2.7782842934106922e-06, + "loss": 0.5034, + "step": 5362 + }, + { + "epoch": 0.44450891007045173, + "grad_norm": 0.26636141538619995, + "learning_rate": 2.7778698715292166e-06, + "loss": 0.5531, + "step": 5363 + }, + { + "epoch": 0.4445917944467468, + "grad_norm": 0.2514444887638092, + "learning_rate": 2.777455449647742e-06, + "loss": 0.5151, + "step": 5364 + }, + { + "epoch": 0.44467467882304185, + "grad_norm": 0.2619096338748932, + "learning_rate": 2.7770410277662664e-06, + "loss": 0.5281, + "step": 5365 + }, + { + "epoch": 0.4447575631993369, + "grad_norm": 0.267353892326355, + "learning_rate": 2.7766266058847908e-06, + "loss": 0.5408, + "step": 5366 + }, + { + "epoch": 0.444840447575632, + "grad_norm": 0.24639128148555756, + "learning_rate": 2.7762121840033156e-06, + "loss": 0.4968, + "step": 5367 + }, + { + "epoch": 0.44492333195192707, + "grad_norm": 0.23001672327518463, + "learning_rate": 2.77579776212184e-06, + "loss": 0.5131, + "step": 5368 + }, + { + "epoch": 0.4450062163282221, + "grad_norm": 0.259010374546051, + "learning_rate": 2.7753833402403654e-06, + "loss": 0.5036, + "step": 5369 + }, + { + "epoch": 0.4450891007045172, + "grad_norm": 0.25801706314086914, + "learning_rate": 2.7749689183588898e-06, + "loss": 0.5, + "step": 5370 + }, + { + "epoch": 0.4451719850808123, + "grad_norm": 0.2519853711128235, + "learning_rate": 2.7745544964774142e-06, + "loss": 0.5436, + "step": 5371 + }, + { + "epoch": 0.44525486945710735, + "grad_norm": 0.2512963116168976, + "learning_rate": 2.774140074595939e-06, + "loss": 0.5222, + "step": 5372 + }, + { + "epoch": 0.4453377538334024, + "grad_norm": 0.23538672924041748, + "learning_rate": 2.7737256527144635e-06, + "loss": 0.4876, + "step": 5373 + }, + { + "epoch": 0.44542063820969746, + "grad_norm": 0.27570515871047974, + "learning_rate": 2.7733112308329884e-06, + "loss": 0.5394, + "step": 5374 + }, + { + "epoch": 0.4455035225859925, + "grad_norm": 0.2745758891105652, + "learning_rate": 2.772896808951513e-06, + "loss": 0.5901, + "step": 5375 + }, + { + "epoch": 0.44558640696228763, + "grad_norm": 0.24171198904514313, + "learning_rate": 2.7724823870700372e-06, + "loss": 0.5021, + "step": 5376 + }, + { + "epoch": 0.4456692913385827, + "grad_norm": 0.25507861375808716, + "learning_rate": 2.7720679651885625e-06, + "loss": 0.579, + "step": 5377 + }, + { + "epoch": 0.44575217571487774, + "grad_norm": 0.2690682113170624, + "learning_rate": 2.771653543307087e-06, + "loss": 0.5815, + "step": 5378 + }, + { + "epoch": 0.4458350600911728, + "grad_norm": 0.2683745324611664, + "learning_rate": 2.7712391214256114e-06, + "loss": 0.56, + "step": 5379 + }, + { + "epoch": 0.4459179444674679, + "grad_norm": 0.2561626434326172, + "learning_rate": 2.7708246995441362e-06, + "loss": 0.5603, + "step": 5380 + }, + { + "epoch": 0.44600082884376296, + "grad_norm": 0.23280486464500427, + "learning_rate": 2.7704102776626607e-06, + "loss": 0.4918, + "step": 5381 + }, + { + "epoch": 0.446083713220058, + "grad_norm": 0.2627258896827698, + "learning_rate": 2.7699958557811855e-06, + "loss": 0.491, + "step": 5382 + }, + { + "epoch": 0.4461665975963531, + "grad_norm": 0.250647634267807, + "learning_rate": 2.76958143389971e-06, + "loss": 0.5452, + "step": 5383 + }, + { + "epoch": 0.44624948197264813, + "grad_norm": 0.2653030753135681, + "learning_rate": 2.7691670120182344e-06, + "loss": 0.5425, + "step": 5384 + }, + { + "epoch": 0.44633236634894324, + "grad_norm": 0.2957487404346466, + "learning_rate": 2.7687525901367596e-06, + "loss": 0.5316, + "step": 5385 + }, + { + "epoch": 0.4464152507252383, + "grad_norm": 0.24365569651126862, + "learning_rate": 2.768338168255284e-06, + "loss": 0.52, + "step": 5386 + }, + { + "epoch": 0.44649813510153336, + "grad_norm": 0.24733765423297882, + "learning_rate": 2.767923746373809e-06, + "loss": 0.4853, + "step": 5387 + }, + { + "epoch": 0.4465810194778284, + "grad_norm": 0.250834584236145, + "learning_rate": 2.7675093244923334e-06, + "loss": 0.535, + "step": 5388 + }, + { + "epoch": 0.4466639038541235, + "grad_norm": 0.29962968826293945, + "learning_rate": 2.767094902610858e-06, + "loss": 0.572, + "step": 5389 + }, + { + "epoch": 0.4467467882304186, + "grad_norm": 0.28093791007995605, + "learning_rate": 2.766680480729383e-06, + "loss": 0.5268, + "step": 5390 + }, + { + "epoch": 0.44682967260671363, + "grad_norm": 0.26209932565689087, + "learning_rate": 2.7662660588479075e-06, + "loss": 0.5393, + "step": 5391 + }, + { + "epoch": 0.4469125569830087, + "grad_norm": 0.25942838191986084, + "learning_rate": 2.7658516369664324e-06, + "loss": 0.5546, + "step": 5392 + }, + { + "epoch": 0.44699544135930375, + "grad_norm": 0.2671935260295868, + "learning_rate": 2.765437215084957e-06, + "loss": 0.5654, + "step": 5393 + }, + { + "epoch": 0.44707832573559886, + "grad_norm": 0.4194624722003937, + "learning_rate": 2.7650227932034812e-06, + "loss": 0.485, + "step": 5394 + }, + { + "epoch": 0.4471612101118939, + "grad_norm": 0.24579301476478577, + "learning_rate": 2.764608371322006e-06, + "loss": 0.4998, + "step": 5395 + }, + { + "epoch": 0.44724409448818897, + "grad_norm": 0.2508236765861511, + "learning_rate": 2.7641939494405305e-06, + "loss": 0.5252, + "step": 5396 + }, + { + "epoch": 0.447326978864484, + "grad_norm": 0.27596157789230347, + "learning_rate": 2.763779527559055e-06, + "loss": 0.5247, + "step": 5397 + }, + { + "epoch": 0.44740986324077914, + "grad_norm": 0.2644176483154297, + "learning_rate": 2.7633651056775802e-06, + "loss": 0.548, + "step": 5398 + }, + { + "epoch": 0.4474927476170742, + "grad_norm": 0.2557331621646881, + "learning_rate": 2.7629506837961047e-06, + "loss": 0.5241, + "step": 5399 + }, + { + "epoch": 0.44757563199336925, + "grad_norm": 0.2562386989593506, + "learning_rate": 2.7625362619146295e-06, + "loss": 0.4742, + "step": 5400 + }, + { + "epoch": 0.4476585163696643, + "grad_norm": 0.29048413038253784, + "learning_rate": 2.762121840033154e-06, + "loss": 0.5481, + "step": 5401 + }, + { + "epoch": 0.44774140074595936, + "grad_norm": 0.24003866314888, + "learning_rate": 2.7617074181516784e-06, + "loss": 0.4857, + "step": 5402 + }, + { + "epoch": 0.4478242851222545, + "grad_norm": 0.2286093533039093, + "learning_rate": 2.7612929962702032e-06, + "loss": 0.4776, + "step": 5403 + }, + { + "epoch": 0.44790716949854953, + "grad_norm": 0.24977876245975494, + "learning_rate": 2.7608785743887277e-06, + "loss": 0.5425, + "step": 5404 + }, + { + "epoch": 0.4479900538748446, + "grad_norm": 0.2547300159931183, + "learning_rate": 2.760464152507253e-06, + "loss": 0.4918, + "step": 5405 + }, + { + "epoch": 0.44807293825113964, + "grad_norm": 0.2840234637260437, + "learning_rate": 2.7600497306257774e-06, + "loss": 0.5803, + "step": 5406 + }, + { + "epoch": 0.44815582262743475, + "grad_norm": 0.237887904047966, + "learning_rate": 2.759635308744302e-06, + "loss": 0.4957, + "step": 5407 + }, + { + "epoch": 0.4482387070037298, + "grad_norm": 0.2527029812335968, + "learning_rate": 2.7592208868628267e-06, + "loss": 0.5131, + "step": 5408 + }, + { + "epoch": 0.44832159138002486, + "grad_norm": 0.2472274899482727, + "learning_rate": 2.758806464981351e-06, + "loss": 0.4935, + "step": 5409 + }, + { + "epoch": 0.4484044757563199, + "grad_norm": 0.2594021260738373, + "learning_rate": 2.7583920430998755e-06, + "loss": 0.5634, + "step": 5410 + }, + { + "epoch": 0.448487360132615, + "grad_norm": 0.2658671438694, + "learning_rate": 2.757977621218401e-06, + "loss": 0.5137, + "step": 5411 + }, + { + "epoch": 0.4485702445089101, + "grad_norm": 0.27461421489715576, + "learning_rate": 2.7575631993369252e-06, + "loss": 0.5531, + "step": 5412 + }, + { + "epoch": 0.44865312888520514, + "grad_norm": 0.24799565970897675, + "learning_rate": 2.75714877745545e-06, + "loss": 0.481, + "step": 5413 + }, + { + "epoch": 0.4487360132615002, + "grad_norm": 0.2545517086982727, + "learning_rate": 2.7567343555739745e-06, + "loss": 0.5112, + "step": 5414 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 0.26920658349990845, + "learning_rate": 2.756319933692499e-06, + "loss": 0.5671, + "step": 5415 + }, + { + "epoch": 0.44890178201409037, + "grad_norm": 0.28440555930137634, + "learning_rate": 2.755905511811024e-06, + "loss": 0.5318, + "step": 5416 + }, + { + "epoch": 0.4489846663903854, + "grad_norm": 0.2540252208709717, + "learning_rate": 2.7554910899295482e-06, + "loss": 0.4998, + "step": 5417 + }, + { + "epoch": 0.4490675507666805, + "grad_norm": 0.23816604912281036, + "learning_rate": 2.7550766680480735e-06, + "loss": 0.488, + "step": 5418 + }, + { + "epoch": 0.44915043514297553, + "grad_norm": 0.27286630868911743, + "learning_rate": 2.754662246166598e-06, + "loss": 0.5289, + "step": 5419 + }, + { + "epoch": 0.4492333195192706, + "grad_norm": 0.2422672063112259, + "learning_rate": 2.7542478242851224e-06, + "loss": 0.4957, + "step": 5420 + }, + { + "epoch": 0.4493162038955657, + "grad_norm": 0.27434247732162476, + "learning_rate": 2.7538334024036472e-06, + "loss": 0.5687, + "step": 5421 + }, + { + "epoch": 0.44939908827186076, + "grad_norm": 0.25345826148986816, + "learning_rate": 2.7534189805221717e-06, + "loss": 0.5023, + "step": 5422 + }, + { + "epoch": 0.4494819726481558, + "grad_norm": 0.250935822725296, + "learning_rate": 2.7530045586406965e-06, + "loss": 0.5383, + "step": 5423 + }, + { + "epoch": 0.44956485702445087, + "grad_norm": 0.25306686758995056, + "learning_rate": 2.752590136759221e-06, + "loss": 0.5116, + "step": 5424 + }, + { + "epoch": 0.449647741400746, + "grad_norm": 0.2517452538013458, + "learning_rate": 2.7521757148777454e-06, + "loss": 0.523, + "step": 5425 + }, + { + "epoch": 0.44973062577704104, + "grad_norm": 0.2951536476612091, + "learning_rate": 2.7517612929962707e-06, + "loss": 0.5706, + "step": 5426 + }, + { + "epoch": 0.4498135101533361, + "grad_norm": 0.2612990140914917, + "learning_rate": 2.751346871114795e-06, + "loss": 0.5073, + "step": 5427 + }, + { + "epoch": 0.44989639452963115, + "grad_norm": 0.2519262731075287, + "learning_rate": 2.7509324492333195e-06, + "loss": 0.5075, + "step": 5428 + }, + { + "epoch": 0.4499792789059262, + "grad_norm": 0.23174449801445007, + "learning_rate": 2.7505180273518444e-06, + "loss": 0.5134, + "step": 5429 + }, + { + "epoch": 0.4500621632822213, + "grad_norm": 0.25911906361579895, + "learning_rate": 2.750103605470369e-06, + "loss": 0.5468, + "step": 5430 + }, + { + "epoch": 0.4501450476585164, + "grad_norm": 0.25058743357658386, + "learning_rate": 2.749689183588894e-06, + "loss": 0.5674, + "step": 5431 + }, + { + "epoch": 0.45022793203481143, + "grad_norm": 0.24782726168632507, + "learning_rate": 2.7492747617074185e-06, + "loss": 0.513, + "step": 5432 + }, + { + "epoch": 0.4503108164111065, + "grad_norm": 0.23715656995773315, + "learning_rate": 2.748860339825943e-06, + "loss": 0.4863, + "step": 5433 + }, + { + "epoch": 0.4503937007874016, + "grad_norm": 0.2563387155532837, + "learning_rate": 2.748445917944468e-06, + "loss": 0.5515, + "step": 5434 + }, + { + "epoch": 0.45047658516369665, + "grad_norm": 0.2540087401866913, + "learning_rate": 2.7480314960629923e-06, + "loss": 0.5367, + "step": 5435 + }, + { + "epoch": 0.4505594695399917, + "grad_norm": 0.24475285410881042, + "learning_rate": 2.747617074181517e-06, + "loss": 0.4645, + "step": 5436 + }, + { + "epoch": 0.45064235391628676, + "grad_norm": 0.24878185987472534, + "learning_rate": 2.7472026523000415e-06, + "loss": 0.5676, + "step": 5437 + }, + { + "epoch": 0.4507252382925819, + "grad_norm": 0.25084736943244934, + "learning_rate": 2.746788230418566e-06, + "loss": 0.5257, + "step": 5438 + }, + { + "epoch": 0.45080812266887693, + "grad_norm": 0.25103065371513367, + "learning_rate": 2.7463738085370912e-06, + "loss": 0.4853, + "step": 5439 + }, + { + "epoch": 0.450891007045172, + "grad_norm": 0.25830215215682983, + "learning_rate": 2.7459593866556157e-06, + "loss": 0.5292, + "step": 5440 + }, + { + "epoch": 0.45097389142146704, + "grad_norm": 0.2513662874698639, + "learning_rate": 2.7455449647741405e-06, + "loss": 0.5531, + "step": 5441 + }, + { + "epoch": 0.4510567757977621, + "grad_norm": 0.2691269814968109, + "learning_rate": 2.745130542892665e-06, + "loss": 0.5338, + "step": 5442 + }, + { + "epoch": 0.4511396601740572, + "grad_norm": 0.257051020860672, + "learning_rate": 2.7447161210111894e-06, + "loss": 0.4919, + "step": 5443 + }, + { + "epoch": 0.45122254455035227, + "grad_norm": 0.2424156665802002, + "learning_rate": 2.7443016991297143e-06, + "loss": 0.4887, + "step": 5444 + }, + { + "epoch": 0.4513054289266473, + "grad_norm": 0.24229861795902252, + "learning_rate": 2.7438872772482387e-06, + "loss": 0.4937, + "step": 5445 + }, + { + "epoch": 0.4513883133029424, + "grad_norm": 0.23950466513633728, + "learning_rate": 2.7434728553667635e-06, + "loss": 0.4821, + "step": 5446 + }, + { + "epoch": 0.4514711976792375, + "grad_norm": 0.2722002863883972, + "learning_rate": 2.7430584334852884e-06, + "loss": 0.5199, + "step": 5447 + }, + { + "epoch": 0.45155408205553255, + "grad_norm": 0.3354588449001312, + "learning_rate": 2.742644011603813e-06, + "loss": 0.5059, + "step": 5448 + }, + { + "epoch": 0.4516369664318276, + "grad_norm": 0.25413602590560913, + "learning_rate": 2.7422295897223377e-06, + "loss": 0.5282, + "step": 5449 + }, + { + "epoch": 0.45171985080812266, + "grad_norm": 0.2543521523475647, + "learning_rate": 2.741815167840862e-06, + "loss": 0.53, + "step": 5450 + }, + { + "epoch": 0.4518027351844177, + "grad_norm": 0.24542881548404694, + "learning_rate": 2.7414007459593865e-06, + "loss": 0.5146, + "step": 5451 + }, + { + "epoch": 0.4518856195607128, + "grad_norm": 0.27676960825920105, + "learning_rate": 2.740986324077912e-06, + "loss": 0.5196, + "step": 5452 + }, + { + "epoch": 0.4519685039370079, + "grad_norm": 0.25565820932388306, + "learning_rate": 2.7405719021964363e-06, + "loss": 0.5581, + "step": 5453 + }, + { + "epoch": 0.45205138831330294, + "grad_norm": 0.25954726338386536, + "learning_rate": 2.740157480314961e-06, + "loss": 0.5063, + "step": 5454 + }, + { + "epoch": 0.452134272689598, + "grad_norm": 0.24261249601840973, + "learning_rate": 2.7397430584334855e-06, + "loss": 0.5435, + "step": 5455 + }, + { + "epoch": 0.4522171570658931, + "grad_norm": 0.2538028061389923, + "learning_rate": 2.73932863655201e-06, + "loss": 0.4877, + "step": 5456 + }, + { + "epoch": 0.45230004144218816, + "grad_norm": 0.23084957897663116, + "learning_rate": 2.738914214670535e-06, + "loss": 0.4789, + "step": 5457 + }, + { + "epoch": 0.4523829258184832, + "grad_norm": 0.24962309002876282, + "learning_rate": 2.7384997927890593e-06, + "loss": 0.5273, + "step": 5458 + }, + { + "epoch": 0.4524658101947783, + "grad_norm": 0.2646229863166809, + "learning_rate": 2.7380853709075845e-06, + "loss": 0.5394, + "step": 5459 + }, + { + "epoch": 0.45254869457107333, + "grad_norm": 0.24471883475780487, + "learning_rate": 2.737670949026109e-06, + "loss": 0.5089, + "step": 5460 + }, + { + "epoch": 0.45263157894736844, + "grad_norm": 0.24027326703071594, + "learning_rate": 2.7372565271446334e-06, + "loss": 0.5142, + "step": 5461 + }, + { + "epoch": 0.4527144633236635, + "grad_norm": 0.2387593537569046, + "learning_rate": 2.7368421052631583e-06, + "loss": 0.4936, + "step": 5462 + }, + { + "epoch": 0.45279734769995855, + "grad_norm": 0.2568579912185669, + "learning_rate": 2.7364276833816827e-06, + "loss": 0.501, + "step": 5463 + }, + { + "epoch": 0.4528802320762536, + "grad_norm": 0.24582450091838837, + "learning_rate": 2.736013261500207e-06, + "loss": 0.5299, + "step": 5464 + }, + { + "epoch": 0.4529631164525487, + "grad_norm": 0.2455950677394867, + "learning_rate": 2.735598839618732e-06, + "loss": 0.5366, + "step": 5465 + }, + { + "epoch": 0.4530460008288438, + "grad_norm": 0.2582765519618988, + "learning_rate": 2.735184417737257e-06, + "loss": 0.5006, + "step": 5466 + }, + { + "epoch": 0.45312888520513883, + "grad_norm": 0.2529853880405426, + "learning_rate": 2.7347699958557817e-06, + "loss": 0.5605, + "step": 5467 + }, + { + "epoch": 0.4532117695814339, + "grad_norm": 0.26070448756217957, + "learning_rate": 2.734355573974306e-06, + "loss": 0.5104, + "step": 5468 + }, + { + "epoch": 0.45329465395772894, + "grad_norm": 0.23943743109703064, + "learning_rate": 2.7339411520928306e-06, + "loss": 0.5132, + "step": 5469 + }, + { + "epoch": 0.45337753833402406, + "grad_norm": 0.2681555151939392, + "learning_rate": 2.7335267302113554e-06, + "loss": 0.5463, + "step": 5470 + }, + { + "epoch": 0.4534604227103191, + "grad_norm": 0.2324221283197403, + "learning_rate": 2.73311230832988e-06, + "loss": 0.4907, + "step": 5471 + }, + { + "epoch": 0.45354330708661417, + "grad_norm": 0.251694917678833, + "learning_rate": 2.732697886448405e-06, + "loss": 0.5094, + "step": 5472 + }, + { + "epoch": 0.4536261914629092, + "grad_norm": 0.2454596906900406, + "learning_rate": 2.7322834645669296e-06, + "loss": 0.5006, + "step": 5473 + }, + { + "epoch": 0.45370907583920433, + "grad_norm": 0.24945536255836487, + "learning_rate": 2.731869042685454e-06, + "loss": 0.5079, + "step": 5474 + }, + { + "epoch": 0.4537919602154994, + "grad_norm": 0.26783597469329834, + "learning_rate": 2.731454620803979e-06, + "loss": 0.5295, + "step": 5475 + }, + { + "epoch": 0.45387484459179445, + "grad_norm": 0.5495002865791321, + "learning_rate": 2.7310401989225033e-06, + "loss": 0.525, + "step": 5476 + }, + { + "epoch": 0.4539577289680895, + "grad_norm": 0.2332741767168045, + "learning_rate": 2.7306257770410277e-06, + "loss": 0.4676, + "step": 5477 + }, + { + "epoch": 0.45404061334438456, + "grad_norm": 0.25150519609451294, + "learning_rate": 2.7302113551595526e-06, + "loss": 0.54, + "step": 5478 + }, + { + "epoch": 0.45412349772067967, + "grad_norm": 0.2652362585067749, + "learning_rate": 2.729796933278077e-06, + "loss": 0.5645, + "step": 5479 + }, + { + "epoch": 0.4542063820969747, + "grad_norm": 0.24392199516296387, + "learning_rate": 2.7293825113966023e-06, + "loss": 0.4923, + "step": 5480 + }, + { + "epoch": 0.4542892664732698, + "grad_norm": 0.25300562381744385, + "learning_rate": 2.7289680895151267e-06, + "loss": 0.5148, + "step": 5481 + }, + { + "epoch": 0.45437215084956484, + "grad_norm": 0.2630462944507599, + "learning_rate": 2.728553667633651e-06, + "loss": 0.4819, + "step": 5482 + }, + { + "epoch": 0.45445503522585995, + "grad_norm": 0.2556450068950653, + "learning_rate": 2.728139245752176e-06, + "loss": 0.5273, + "step": 5483 + }, + { + "epoch": 0.454537919602155, + "grad_norm": 0.2842545509338379, + "learning_rate": 2.7277248238707004e-06, + "loss": 0.5072, + "step": 5484 + }, + { + "epoch": 0.45462080397845006, + "grad_norm": 0.2595987319946289, + "learning_rate": 2.7273104019892253e-06, + "loss": 0.4843, + "step": 5485 + }, + { + "epoch": 0.4547036883547451, + "grad_norm": 0.28753209114074707, + "learning_rate": 2.7268959801077497e-06, + "loss": 0.5539, + "step": 5486 + }, + { + "epoch": 0.4547865727310402, + "grad_norm": 0.29988011717796326, + "learning_rate": 2.7264815582262746e-06, + "loss": 0.4994, + "step": 5487 + }, + { + "epoch": 0.4548694571073353, + "grad_norm": 0.31416213512420654, + "learning_rate": 2.7260671363447994e-06, + "loss": 0.5288, + "step": 5488 + }, + { + "epoch": 0.45495234148363034, + "grad_norm": 0.2623971104621887, + "learning_rate": 2.725652714463324e-06, + "loss": 0.5305, + "step": 5489 + }, + { + "epoch": 0.4550352258599254, + "grad_norm": 0.2616049349308014, + "learning_rate": 2.7252382925818487e-06, + "loss": 0.4975, + "step": 5490 + }, + { + "epoch": 0.45511811023622045, + "grad_norm": 0.2944031357765198, + "learning_rate": 2.724823870700373e-06, + "loss": 0.5759, + "step": 5491 + }, + { + "epoch": 0.45520099461251556, + "grad_norm": 0.2845168709754944, + "learning_rate": 2.7244094488188976e-06, + "loss": 0.5276, + "step": 5492 + }, + { + "epoch": 0.4552838789888106, + "grad_norm": 0.2686086595058441, + "learning_rate": 2.723995026937423e-06, + "loss": 0.5392, + "step": 5493 + }, + { + "epoch": 0.4553667633651057, + "grad_norm": 0.27931055426597595, + "learning_rate": 2.7235806050559473e-06, + "loss": 0.5122, + "step": 5494 + }, + { + "epoch": 0.45544964774140073, + "grad_norm": 0.2653496265411377, + "learning_rate": 2.7231661831744717e-06, + "loss": 0.5314, + "step": 5495 + }, + { + "epoch": 0.4555325321176958, + "grad_norm": 0.24896423518657684, + "learning_rate": 2.7227517612929966e-06, + "loss": 0.5291, + "step": 5496 + }, + { + "epoch": 0.4556154164939909, + "grad_norm": 0.2644193768501282, + "learning_rate": 2.722337339411521e-06, + "loss": 0.5231, + "step": 5497 + }, + { + "epoch": 0.45569830087028595, + "grad_norm": 0.247967928647995, + "learning_rate": 2.721922917530046e-06, + "loss": 0.5309, + "step": 5498 + }, + { + "epoch": 0.455781185246581, + "grad_norm": 0.2584999203681946, + "learning_rate": 2.7215084956485703e-06, + "loss": 0.4958, + "step": 5499 + }, + { + "epoch": 0.45586406962287607, + "grad_norm": 0.25547829270362854, + "learning_rate": 2.7210940737670947e-06, + "loss": 0.5016, + "step": 5500 + }, + { + "epoch": 0.4559469539991712, + "grad_norm": 0.2671474814414978, + "learning_rate": 2.72067965188562e-06, + "loss": 0.5347, + "step": 5501 + }, + { + "epoch": 0.45602983837546623, + "grad_norm": 0.26050686836242676, + "learning_rate": 2.7202652300041444e-06, + "loss": 0.4577, + "step": 5502 + }, + { + "epoch": 0.4561127227517613, + "grad_norm": 0.2682153880596161, + "learning_rate": 2.7198508081226693e-06, + "loss": 0.5291, + "step": 5503 + }, + { + "epoch": 0.45619560712805635, + "grad_norm": 0.25987109541893005, + "learning_rate": 2.7194363862411937e-06, + "loss": 0.5554, + "step": 5504 + }, + { + "epoch": 0.4562784915043514, + "grad_norm": 0.25693613290786743, + "learning_rate": 2.719021964359718e-06, + "loss": 0.5272, + "step": 5505 + }, + { + "epoch": 0.4563613758806465, + "grad_norm": 0.22685031592845917, + "learning_rate": 2.718607542478243e-06, + "loss": 0.4521, + "step": 5506 + }, + { + "epoch": 0.45644426025694157, + "grad_norm": 0.2525438964366913, + "learning_rate": 2.718193120596768e-06, + "loss": 0.5266, + "step": 5507 + }, + { + "epoch": 0.4565271446332366, + "grad_norm": 0.2497277408838272, + "learning_rate": 2.7177786987152927e-06, + "loss": 0.4712, + "step": 5508 + }, + { + "epoch": 0.4566100290095317, + "grad_norm": 0.25798502564430237, + "learning_rate": 2.717364276833817e-06, + "loss": 0.5403, + "step": 5509 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 0.24877411127090454, + "learning_rate": 2.7169498549523416e-06, + "loss": 0.5385, + "step": 5510 + }, + { + "epoch": 0.45677579776212185, + "grad_norm": 0.25632768869400024, + "learning_rate": 2.7165354330708664e-06, + "loss": 0.4989, + "step": 5511 + }, + { + "epoch": 0.4568586821384169, + "grad_norm": 0.249245747923851, + "learning_rate": 2.716121011189391e-06, + "loss": 0.5093, + "step": 5512 + }, + { + "epoch": 0.45694156651471196, + "grad_norm": 0.27607637643814087, + "learning_rate": 2.7157065893079153e-06, + "loss": 0.5301, + "step": 5513 + }, + { + "epoch": 0.4570244508910071, + "grad_norm": 0.27108022570610046, + "learning_rate": 2.7152921674264406e-06, + "loss": 0.5862, + "step": 5514 + }, + { + "epoch": 0.45710733526730213, + "grad_norm": 0.2543375790119171, + "learning_rate": 2.714877745544965e-06, + "loss": 0.536, + "step": 5515 + }, + { + "epoch": 0.4571902196435972, + "grad_norm": 0.25257954001426697, + "learning_rate": 2.71446332366349e-06, + "loss": 0.515, + "step": 5516 + }, + { + "epoch": 0.45727310401989224, + "grad_norm": 0.27318552136421204, + "learning_rate": 2.7140489017820143e-06, + "loss": 0.5052, + "step": 5517 + }, + { + "epoch": 0.4573559883961873, + "grad_norm": 0.24823756515979767, + "learning_rate": 2.7136344799005387e-06, + "loss": 0.4988, + "step": 5518 + }, + { + "epoch": 0.4574388727724824, + "grad_norm": 0.2658519148826599, + "learning_rate": 2.7132200580190636e-06, + "loss": 0.5258, + "step": 5519 + }, + { + "epoch": 0.45752175714877746, + "grad_norm": 0.2472013235092163, + "learning_rate": 2.712805636137588e-06, + "loss": 0.4803, + "step": 5520 + }, + { + "epoch": 0.4576046415250725, + "grad_norm": 0.23741371929645538, + "learning_rate": 2.7123912142561133e-06, + "loss": 0.5022, + "step": 5521 + }, + { + "epoch": 0.4576875259013676, + "grad_norm": 0.2516511380672455, + "learning_rate": 2.7119767923746377e-06, + "loss": 0.506, + "step": 5522 + }, + { + "epoch": 0.4577704102776627, + "grad_norm": 0.2340400665998459, + "learning_rate": 2.711562370493162e-06, + "loss": 0.5231, + "step": 5523 + }, + { + "epoch": 0.45785329465395774, + "grad_norm": 0.25584378838539124, + "learning_rate": 2.711147948611687e-06, + "loss": 0.5209, + "step": 5524 + }, + { + "epoch": 0.4579361790302528, + "grad_norm": 0.2736526131629944, + "learning_rate": 2.7107335267302114e-06, + "loss": 0.5319, + "step": 5525 + }, + { + "epoch": 0.45801906340654785, + "grad_norm": 0.2486901581287384, + "learning_rate": 2.7103191048487363e-06, + "loss": 0.5289, + "step": 5526 + }, + { + "epoch": 0.4581019477828429, + "grad_norm": 0.25775036215782166, + "learning_rate": 2.709904682967261e-06, + "loss": 0.5418, + "step": 5527 + }, + { + "epoch": 0.458184832159138, + "grad_norm": 0.257315993309021, + "learning_rate": 2.7094902610857856e-06, + "loss": 0.5193, + "step": 5528 + }, + { + "epoch": 0.4582677165354331, + "grad_norm": 0.251610666513443, + "learning_rate": 2.7090758392043104e-06, + "loss": 0.5204, + "step": 5529 + }, + { + "epoch": 0.45835060091172813, + "grad_norm": 0.23908615112304688, + "learning_rate": 2.708661417322835e-06, + "loss": 0.5082, + "step": 5530 + }, + { + "epoch": 0.4584334852880232, + "grad_norm": 0.2593863904476166, + "learning_rate": 2.7082469954413593e-06, + "loss": 0.5421, + "step": 5531 + }, + { + "epoch": 0.4585163696643183, + "grad_norm": 0.2614327669143677, + "learning_rate": 2.707832573559884e-06, + "loss": 0.5172, + "step": 5532 + }, + { + "epoch": 0.45859925404061336, + "grad_norm": 0.2547266185283661, + "learning_rate": 2.7074181516784086e-06, + "loss": 0.5244, + "step": 5533 + }, + { + "epoch": 0.4586821384169084, + "grad_norm": 0.232369065284729, + "learning_rate": 2.707003729796934e-06, + "loss": 0.4858, + "step": 5534 + }, + { + "epoch": 0.45876502279320347, + "grad_norm": 0.248989075422287, + "learning_rate": 2.7065893079154583e-06, + "loss": 0.4921, + "step": 5535 + }, + { + "epoch": 0.4588479071694985, + "grad_norm": 0.23410922288894653, + "learning_rate": 2.7061748860339827e-06, + "loss": 0.4891, + "step": 5536 + }, + { + "epoch": 0.45893079154579364, + "grad_norm": 0.25730812549591064, + "learning_rate": 2.7057604641525076e-06, + "loss": 0.4879, + "step": 5537 + }, + { + "epoch": 0.4590136759220887, + "grad_norm": 0.2531355917453766, + "learning_rate": 2.705346042271032e-06, + "loss": 0.5053, + "step": 5538 + }, + { + "epoch": 0.45909656029838375, + "grad_norm": 0.24835211038589478, + "learning_rate": 2.704931620389557e-06, + "loss": 0.5062, + "step": 5539 + }, + { + "epoch": 0.4591794446746788, + "grad_norm": 0.2526538372039795, + "learning_rate": 2.7045171985080813e-06, + "loss": 0.5081, + "step": 5540 + }, + { + "epoch": 0.4592623290509739, + "grad_norm": 0.272857666015625, + "learning_rate": 2.7041027766266057e-06, + "loss": 0.5685, + "step": 5541 + }, + { + "epoch": 0.459345213427269, + "grad_norm": 0.2507550120353699, + "learning_rate": 2.703688354745131e-06, + "loss": 0.5216, + "step": 5542 + }, + { + "epoch": 0.45942809780356403, + "grad_norm": 0.2542041540145874, + "learning_rate": 2.7032739328636554e-06, + "loss": 0.5168, + "step": 5543 + }, + { + "epoch": 0.4595109821798591, + "grad_norm": 0.258242666721344, + "learning_rate": 2.7028595109821803e-06, + "loss": 0.5356, + "step": 5544 + }, + { + "epoch": 0.45959386655615414, + "grad_norm": 0.2332838475704193, + "learning_rate": 2.7024450891007047e-06, + "loss": 0.4854, + "step": 5545 + }, + { + "epoch": 0.45967675093244925, + "grad_norm": 0.2582071125507355, + "learning_rate": 2.702030667219229e-06, + "loss": 0.5519, + "step": 5546 + }, + { + "epoch": 0.4597596353087443, + "grad_norm": 0.26779428124427795, + "learning_rate": 2.7016162453377544e-06, + "loss": 0.5055, + "step": 5547 + }, + { + "epoch": 0.45984251968503936, + "grad_norm": 0.2523634433746338, + "learning_rate": 2.701201823456279e-06, + "loss": 0.4634, + "step": 5548 + }, + { + "epoch": 0.4599254040613344, + "grad_norm": 0.27597370743751526, + "learning_rate": 2.7007874015748033e-06, + "loss": 0.5434, + "step": 5549 + }, + { + "epoch": 0.46000828843762953, + "grad_norm": 0.24184872210025787, + "learning_rate": 2.700372979693328e-06, + "loss": 0.5048, + "step": 5550 + }, + { + "epoch": 0.4600911728139246, + "grad_norm": 0.2778463363647461, + "learning_rate": 2.6999585578118526e-06, + "loss": 0.5526, + "step": 5551 + }, + { + "epoch": 0.46017405719021964, + "grad_norm": 0.22983793914318085, + "learning_rate": 2.6995441359303775e-06, + "loss": 0.4613, + "step": 5552 + }, + { + "epoch": 0.4602569415665147, + "grad_norm": 0.27077817916870117, + "learning_rate": 2.699129714048902e-06, + "loss": 0.5408, + "step": 5553 + }, + { + "epoch": 0.46033982594280975, + "grad_norm": 0.23591966927051544, + "learning_rate": 2.6987152921674263e-06, + "loss": 0.505, + "step": 5554 + }, + { + "epoch": 0.46042271031910487, + "grad_norm": 0.26877039670944214, + "learning_rate": 2.6983008702859516e-06, + "loss": 0.5564, + "step": 5555 + }, + { + "epoch": 0.4605055946953999, + "grad_norm": 0.28493770956993103, + "learning_rate": 2.697886448404476e-06, + "loss": 0.5026, + "step": 5556 + }, + { + "epoch": 0.460588479071695, + "grad_norm": 0.2638489902019501, + "learning_rate": 2.697472026523001e-06, + "loss": 0.5349, + "step": 5557 + }, + { + "epoch": 0.46067136344799003, + "grad_norm": 0.28114834427833557, + "learning_rate": 2.6970576046415253e-06, + "loss": 0.5359, + "step": 5558 + }, + { + "epoch": 0.46075424782428515, + "grad_norm": 0.25402742624282837, + "learning_rate": 2.6966431827600497e-06, + "loss": 0.5568, + "step": 5559 + }, + { + "epoch": 0.4608371322005802, + "grad_norm": 0.24565526843070984, + "learning_rate": 2.6962287608785746e-06, + "loss": 0.5065, + "step": 5560 + }, + { + "epoch": 0.46092001657687526, + "grad_norm": 0.2647832930088043, + "learning_rate": 2.695814338997099e-06, + "loss": 0.5361, + "step": 5561 + }, + { + "epoch": 0.4610029009531703, + "grad_norm": 0.23811282217502594, + "learning_rate": 2.6953999171156235e-06, + "loss": 0.4967, + "step": 5562 + }, + { + "epoch": 0.46108578532946537, + "grad_norm": 0.24984201788902283, + "learning_rate": 2.6949854952341487e-06, + "loss": 0.5006, + "step": 5563 + }, + { + "epoch": 0.4611686697057605, + "grad_norm": 0.277915894985199, + "learning_rate": 2.694571073352673e-06, + "loss": 0.543, + "step": 5564 + }, + { + "epoch": 0.46125155408205554, + "grad_norm": 0.23899592459201813, + "learning_rate": 2.694156651471198e-06, + "loss": 0.51, + "step": 5565 + }, + { + "epoch": 0.4613344384583506, + "grad_norm": 0.24680830538272858, + "learning_rate": 2.6937422295897225e-06, + "loss": 0.5179, + "step": 5566 + }, + { + "epoch": 0.46141732283464565, + "grad_norm": 0.2575418949127197, + "learning_rate": 2.693327807708247e-06, + "loss": 0.5406, + "step": 5567 + }, + { + "epoch": 0.46150020721094076, + "grad_norm": 0.24173030257225037, + "learning_rate": 2.692913385826772e-06, + "loss": 0.5387, + "step": 5568 + }, + { + "epoch": 0.4615830915872358, + "grad_norm": 0.2786121070384979, + "learning_rate": 2.6924989639452966e-06, + "loss": 0.5606, + "step": 5569 + }, + { + "epoch": 0.4616659759635309, + "grad_norm": 0.2402992844581604, + "learning_rate": 2.6920845420638215e-06, + "loss": 0.5176, + "step": 5570 + }, + { + "epoch": 0.46174886033982593, + "grad_norm": 0.2543796896934509, + "learning_rate": 2.691670120182346e-06, + "loss": 0.5167, + "step": 5571 + }, + { + "epoch": 0.461831744716121, + "grad_norm": 0.2612086236476898, + "learning_rate": 2.6912556983008703e-06, + "loss": 0.5593, + "step": 5572 + }, + { + "epoch": 0.4619146290924161, + "grad_norm": 0.27284446358680725, + "learning_rate": 2.690841276419395e-06, + "loss": 0.5682, + "step": 5573 + }, + { + "epoch": 0.46199751346871115, + "grad_norm": 0.2703678011894226, + "learning_rate": 2.6904268545379196e-06, + "loss": 0.5164, + "step": 5574 + }, + { + "epoch": 0.4620803978450062, + "grad_norm": 0.2578887939453125, + "learning_rate": 2.690012432656445e-06, + "loss": 0.5041, + "step": 5575 + }, + { + "epoch": 0.46216328222130126, + "grad_norm": 0.23997502028942108, + "learning_rate": 2.6895980107749693e-06, + "loss": 0.5074, + "step": 5576 + }, + { + "epoch": 0.4622461665975964, + "grad_norm": 0.24371156096458435, + "learning_rate": 2.6891835888934938e-06, + "loss": 0.5056, + "step": 5577 + }, + { + "epoch": 0.46232905097389143, + "grad_norm": 0.24378842115402222, + "learning_rate": 2.6887691670120186e-06, + "loss": 0.5271, + "step": 5578 + }, + { + "epoch": 0.4624119353501865, + "grad_norm": 0.2479882687330246, + "learning_rate": 2.688354745130543e-06, + "loss": 0.4882, + "step": 5579 + }, + { + "epoch": 0.46249481972648154, + "grad_norm": 0.2404925376176834, + "learning_rate": 2.6879403232490675e-06, + "loss": 0.5072, + "step": 5580 + }, + { + "epoch": 0.4625777041027766, + "grad_norm": 0.26553335785865784, + "learning_rate": 2.6875259013675923e-06, + "loss": 0.5456, + "step": 5581 + }, + { + "epoch": 0.4626605884790717, + "grad_norm": 0.25075671076774597, + "learning_rate": 2.6871114794861168e-06, + "loss": 0.5028, + "step": 5582 + }, + { + "epoch": 0.46274347285536677, + "grad_norm": 0.2531648278236389, + "learning_rate": 2.686697057604642e-06, + "loss": 0.4977, + "step": 5583 + }, + { + "epoch": 0.4628263572316618, + "grad_norm": 0.2494460940361023, + "learning_rate": 2.6862826357231665e-06, + "loss": 0.5249, + "step": 5584 + }, + { + "epoch": 0.4629092416079569, + "grad_norm": 0.2630887031555176, + "learning_rate": 2.685868213841691e-06, + "loss": 0.5461, + "step": 5585 + }, + { + "epoch": 0.462992125984252, + "grad_norm": 0.26605626940727234, + "learning_rate": 2.6854537919602158e-06, + "loss": 0.5463, + "step": 5586 + }, + { + "epoch": 0.46307501036054705, + "grad_norm": 0.24170351028442383, + "learning_rate": 2.68503937007874e-06, + "loss": 0.4954, + "step": 5587 + }, + { + "epoch": 0.4631578947368421, + "grad_norm": 0.25585514307022095, + "learning_rate": 2.6846249481972655e-06, + "loss": 0.5009, + "step": 5588 + }, + { + "epoch": 0.46324077911313716, + "grad_norm": 0.22580252587795258, + "learning_rate": 2.68421052631579e-06, + "loss": 0.4839, + "step": 5589 + }, + { + "epoch": 0.46332366348943227, + "grad_norm": 0.25616133213043213, + "learning_rate": 2.6837961044343143e-06, + "loss": 0.5312, + "step": 5590 + }, + { + "epoch": 0.4634065478657273, + "grad_norm": 0.25658154487609863, + "learning_rate": 2.683381682552839e-06, + "loss": 0.5539, + "step": 5591 + }, + { + "epoch": 0.4634894322420224, + "grad_norm": 0.29882654547691345, + "learning_rate": 2.6829672606713636e-06, + "loss": 0.5656, + "step": 5592 + }, + { + "epoch": 0.46357231661831744, + "grad_norm": 0.24919159710407257, + "learning_rate": 2.6825528387898885e-06, + "loss": 0.5205, + "step": 5593 + }, + { + "epoch": 0.4636552009946125, + "grad_norm": 0.2731539309024811, + "learning_rate": 2.682138416908413e-06, + "loss": 0.5202, + "step": 5594 + }, + { + "epoch": 0.4637380853709076, + "grad_norm": 0.25126728415489197, + "learning_rate": 2.6817239950269373e-06, + "loss": 0.5528, + "step": 5595 + }, + { + "epoch": 0.46382096974720266, + "grad_norm": 0.2560661733150482, + "learning_rate": 2.6813095731454626e-06, + "loss": 0.5197, + "step": 5596 + }, + { + "epoch": 0.4639038541234977, + "grad_norm": 0.25521162152290344, + "learning_rate": 2.680895151263987e-06, + "loss": 0.5191, + "step": 5597 + }, + { + "epoch": 0.4639867384997928, + "grad_norm": 0.25352761149406433, + "learning_rate": 2.6804807293825115e-06, + "loss": 0.5284, + "step": 5598 + }, + { + "epoch": 0.4640696228760879, + "grad_norm": 0.2768426239490509, + "learning_rate": 2.6800663075010363e-06, + "loss": 0.4981, + "step": 5599 + }, + { + "epoch": 0.46415250725238294, + "grad_norm": 0.2460351586341858, + "learning_rate": 2.6796518856195608e-06, + "loss": 0.5089, + "step": 5600 + }, + { + "epoch": 0.464235391628678, + "grad_norm": 0.2573002278804779, + "learning_rate": 2.6792374637380856e-06, + "loss": 0.5107, + "step": 5601 + }, + { + "epoch": 0.46431827600497305, + "grad_norm": 0.26176729798316956, + "learning_rate": 2.67882304185661e-06, + "loss": 0.5046, + "step": 5602 + }, + { + "epoch": 0.4644011603812681, + "grad_norm": 0.2620142996311188, + "learning_rate": 2.6784086199751345e-06, + "loss": 0.5251, + "step": 5603 + }, + { + "epoch": 0.4644840447575632, + "grad_norm": 0.24973107874393463, + "learning_rate": 2.6779941980936598e-06, + "loss": 0.5137, + "step": 5604 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 0.27205246686935425, + "learning_rate": 2.677579776212184e-06, + "loss": 0.5041, + "step": 5605 + }, + { + "epoch": 0.46464981351015333, + "grad_norm": 0.2717180848121643, + "learning_rate": 2.677165354330709e-06, + "loss": 0.5521, + "step": 5606 + }, + { + "epoch": 0.4647326978864484, + "grad_norm": 0.25753089785575867, + "learning_rate": 2.6767509324492335e-06, + "loss": 0.5345, + "step": 5607 + }, + { + "epoch": 0.4648155822627435, + "grad_norm": 0.23957028985023499, + "learning_rate": 2.676336510567758e-06, + "loss": 0.5179, + "step": 5608 + }, + { + "epoch": 0.46489846663903855, + "grad_norm": 0.24205978214740753, + "learning_rate": 2.675922088686283e-06, + "loss": 0.4935, + "step": 5609 + }, + { + "epoch": 0.4649813510153336, + "grad_norm": 0.24133776128292084, + "learning_rate": 2.6755076668048076e-06, + "loss": 0.4922, + "step": 5610 + }, + { + "epoch": 0.46506423539162867, + "grad_norm": 0.25722837448120117, + "learning_rate": 2.6750932449233325e-06, + "loss": 0.507, + "step": 5611 + }, + { + "epoch": 0.4651471197679237, + "grad_norm": 0.25238922238349915, + "learning_rate": 2.674678823041857e-06, + "loss": 0.5224, + "step": 5612 + }, + { + "epoch": 0.46523000414421883, + "grad_norm": 0.2593724727630615, + "learning_rate": 2.6742644011603813e-06, + "loss": 0.5511, + "step": 5613 + }, + { + "epoch": 0.4653128885205139, + "grad_norm": 0.2447141855955124, + "learning_rate": 2.673849979278906e-06, + "loss": 0.5554, + "step": 5614 + }, + { + "epoch": 0.46539577289680895, + "grad_norm": 0.2497607320547104, + "learning_rate": 2.6734355573974306e-06, + "loss": 0.5226, + "step": 5615 + }, + { + "epoch": 0.465478657273104, + "grad_norm": 0.2562551200389862, + "learning_rate": 2.673021135515955e-06, + "loss": 0.5593, + "step": 5616 + }, + { + "epoch": 0.4655615416493991, + "grad_norm": 0.2540173828601837, + "learning_rate": 2.6726067136344803e-06, + "loss": 0.5038, + "step": 5617 + }, + { + "epoch": 0.46564442602569417, + "grad_norm": 0.25621452927589417, + "learning_rate": 2.6721922917530048e-06, + "loss": 0.5082, + "step": 5618 + }, + { + "epoch": 0.4657273104019892, + "grad_norm": 0.25581926107406616, + "learning_rate": 2.6717778698715296e-06, + "loss": 0.5074, + "step": 5619 + }, + { + "epoch": 0.4658101947782843, + "grad_norm": 0.2549000680446625, + "learning_rate": 2.671363447990054e-06, + "loss": 0.5116, + "step": 5620 + }, + { + "epoch": 0.46589307915457934, + "grad_norm": 0.24006257951259613, + "learning_rate": 2.6709490261085785e-06, + "loss": 0.4438, + "step": 5621 + }, + { + "epoch": 0.46597596353087445, + "grad_norm": 0.24180130660533905, + "learning_rate": 2.6705346042271033e-06, + "loss": 0.5095, + "step": 5622 + }, + { + "epoch": 0.4660588479071695, + "grad_norm": 0.28261277079582214, + "learning_rate": 2.6701201823456278e-06, + "loss": 0.5458, + "step": 5623 + }, + { + "epoch": 0.46614173228346456, + "grad_norm": 0.28123387694358826, + "learning_rate": 2.669705760464153e-06, + "loss": 0.548, + "step": 5624 + }, + { + "epoch": 0.4662246166597596, + "grad_norm": 0.28289586305618286, + "learning_rate": 2.6692913385826775e-06, + "loss": 0.5669, + "step": 5625 + }, + { + "epoch": 0.46630750103605473, + "grad_norm": 0.23230229318141937, + "learning_rate": 2.668876916701202e-06, + "loss": 0.4432, + "step": 5626 + }, + { + "epoch": 0.4663903854123498, + "grad_norm": 0.2482929527759552, + "learning_rate": 2.6684624948197268e-06, + "loss": 0.5065, + "step": 5627 + }, + { + "epoch": 0.46647326978864484, + "grad_norm": 0.2474246323108673, + "learning_rate": 2.668048072938251e-06, + "loss": 0.5124, + "step": 5628 + }, + { + "epoch": 0.4665561541649399, + "grad_norm": 0.2535220682621002, + "learning_rate": 2.6676336510567756e-06, + "loss": 0.5006, + "step": 5629 + }, + { + "epoch": 0.46663903854123495, + "grad_norm": 0.2585424780845642, + "learning_rate": 2.667219229175301e-06, + "loss": 0.5295, + "step": 5630 + }, + { + "epoch": 0.46672192291753006, + "grad_norm": 0.2509098947048187, + "learning_rate": 2.6668048072938254e-06, + "loss": 0.4707, + "step": 5631 + }, + { + "epoch": 0.4668048072938251, + "grad_norm": 0.24476446211338043, + "learning_rate": 2.66639038541235e-06, + "loss": 0.4982, + "step": 5632 + }, + { + "epoch": 0.4668876916701202, + "grad_norm": 0.2710858881473541, + "learning_rate": 2.6659759635308746e-06, + "loss": 0.5513, + "step": 5633 + }, + { + "epoch": 0.46697057604641523, + "grad_norm": 0.2465999573469162, + "learning_rate": 2.665561541649399e-06, + "loss": 0.5046, + "step": 5634 + }, + { + "epoch": 0.46705346042271034, + "grad_norm": 0.250912070274353, + "learning_rate": 2.665147119767924e-06, + "loss": 0.5168, + "step": 5635 + }, + { + "epoch": 0.4671363447990054, + "grad_norm": 0.2646140158176422, + "learning_rate": 2.6647326978864484e-06, + "loss": 0.5153, + "step": 5636 + }, + { + "epoch": 0.46721922917530045, + "grad_norm": 0.2359207272529602, + "learning_rate": 2.6643182760049736e-06, + "loss": 0.4771, + "step": 5637 + }, + { + "epoch": 0.4673021135515955, + "grad_norm": 0.23050457239151, + "learning_rate": 2.663903854123498e-06, + "loss": 0.4598, + "step": 5638 + }, + { + "epoch": 0.46738499792789057, + "grad_norm": 0.25947511196136475, + "learning_rate": 2.6634894322420225e-06, + "loss": 0.54, + "step": 5639 + }, + { + "epoch": 0.4674678823041857, + "grad_norm": 1.321203589439392, + "learning_rate": 2.6630750103605474e-06, + "loss": 0.4985, + "step": 5640 + }, + { + "epoch": 0.46755076668048073, + "grad_norm": 0.2619878649711609, + "learning_rate": 2.6626605884790718e-06, + "loss": 0.5389, + "step": 5641 + }, + { + "epoch": 0.4676336510567758, + "grad_norm": 0.2602897584438324, + "learning_rate": 2.6622461665975966e-06, + "loss": 0.5171, + "step": 5642 + }, + { + "epoch": 0.46771653543307085, + "grad_norm": 0.2633875608444214, + "learning_rate": 2.661831744716121e-06, + "loss": 0.5313, + "step": 5643 + }, + { + "epoch": 0.46779941980936596, + "grad_norm": 0.23841769993305206, + "learning_rate": 2.6614173228346455e-06, + "loss": 0.4856, + "step": 5644 + }, + { + "epoch": 0.467882304185661, + "grad_norm": 0.2558608055114746, + "learning_rate": 2.6610029009531708e-06, + "loss": 0.5221, + "step": 5645 + }, + { + "epoch": 0.46796518856195607, + "grad_norm": 0.2473994344472885, + "learning_rate": 2.6605884790716952e-06, + "loss": 0.4829, + "step": 5646 + }, + { + "epoch": 0.4680480729382511, + "grad_norm": 0.23911038041114807, + "learning_rate": 2.6601740571902196e-06, + "loss": 0.4907, + "step": 5647 + }, + { + "epoch": 0.4681309573145462, + "grad_norm": 0.25574541091918945, + "learning_rate": 2.6597596353087445e-06, + "loss": 0.5472, + "step": 5648 + }, + { + "epoch": 0.4682138416908413, + "grad_norm": 0.24398164451122284, + "learning_rate": 2.659345213427269e-06, + "loss": 0.4677, + "step": 5649 + }, + { + "epoch": 0.46829672606713635, + "grad_norm": 0.24919520318508148, + "learning_rate": 2.6589307915457942e-06, + "loss": 0.5135, + "step": 5650 + }, + { + "epoch": 0.4683796104434314, + "grad_norm": 0.24349498748779297, + "learning_rate": 2.6585163696643186e-06, + "loss": 0.4937, + "step": 5651 + }, + { + "epoch": 0.46846249481972646, + "grad_norm": 0.26165175437927246, + "learning_rate": 2.658101947782843e-06, + "loss": 0.5282, + "step": 5652 + }, + { + "epoch": 0.46854537919602157, + "grad_norm": 0.23115074634552002, + "learning_rate": 2.657687525901368e-06, + "loss": 0.4951, + "step": 5653 + }, + { + "epoch": 0.46862826357231663, + "grad_norm": 0.23263321816921234, + "learning_rate": 2.6572731040198924e-06, + "loss": 0.4603, + "step": 5654 + }, + { + "epoch": 0.4687111479486117, + "grad_norm": 0.244461789727211, + "learning_rate": 2.6568586821384172e-06, + "loss": 0.4851, + "step": 5655 + }, + { + "epoch": 0.46879403232490674, + "grad_norm": 0.2573271691799164, + "learning_rate": 2.6564442602569417e-06, + "loss": 0.4816, + "step": 5656 + }, + { + "epoch": 0.4688769167012018, + "grad_norm": 0.2535220980644226, + "learning_rate": 2.656029838375466e-06, + "loss": 0.4874, + "step": 5657 + }, + { + "epoch": 0.4689598010774969, + "grad_norm": 0.2700043022632599, + "learning_rate": 2.6556154164939914e-06, + "loss": 0.5529, + "step": 5658 + }, + { + "epoch": 0.46904268545379196, + "grad_norm": 0.23855096101760864, + "learning_rate": 2.655200994612516e-06, + "loss": 0.5199, + "step": 5659 + }, + { + "epoch": 0.469125569830087, + "grad_norm": 0.25236794352531433, + "learning_rate": 2.6547865727310406e-06, + "loss": 0.4971, + "step": 5660 + }, + { + "epoch": 0.4692084542063821, + "grad_norm": 0.2486615926027298, + "learning_rate": 2.654372150849565e-06, + "loss": 0.4903, + "step": 5661 + }, + { + "epoch": 0.4692913385826772, + "grad_norm": 0.25202858448028564, + "learning_rate": 2.6539577289680895e-06, + "loss": 0.5039, + "step": 5662 + }, + { + "epoch": 0.46937422295897224, + "grad_norm": 0.25210872292518616, + "learning_rate": 2.6535433070866144e-06, + "loss": 0.5155, + "step": 5663 + }, + { + "epoch": 0.4694571073352673, + "grad_norm": 0.25250813364982605, + "learning_rate": 2.653128885205139e-06, + "loss": 0.4966, + "step": 5664 + }, + { + "epoch": 0.46953999171156235, + "grad_norm": 0.2850164771080017, + "learning_rate": 2.6527144633236637e-06, + "loss": 0.5594, + "step": 5665 + }, + { + "epoch": 0.46962287608785747, + "grad_norm": 0.25405070185661316, + "learning_rate": 2.6523000414421885e-06, + "loss": 0.4796, + "step": 5666 + }, + { + "epoch": 0.4697057604641525, + "grad_norm": 0.25790584087371826, + "learning_rate": 2.651885619560713e-06, + "loss": 0.5311, + "step": 5667 + }, + { + "epoch": 0.4697886448404476, + "grad_norm": 0.26386210322380066, + "learning_rate": 2.651471197679238e-06, + "loss": 0.529, + "step": 5668 + }, + { + "epoch": 0.46987152921674263, + "grad_norm": 0.25121504068374634, + "learning_rate": 2.6510567757977622e-06, + "loss": 0.5069, + "step": 5669 + }, + { + "epoch": 0.4699544135930377, + "grad_norm": 0.2383774220943451, + "learning_rate": 2.6506423539162867e-06, + "loss": 0.4862, + "step": 5670 + }, + { + "epoch": 0.4700372979693328, + "grad_norm": 0.25487303733825684, + "learning_rate": 2.650227932034812e-06, + "loss": 0.5052, + "step": 5671 + }, + { + "epoch": 0.47012018234562786, + "grad_norm": 0.2643219530582428, + "learning_rate": 2.6498135101533364e-06, + "loss": 0.4975, + "step": 5672 + }, + { + "epoch": 0.4702030667219229, + "grad_norm": 0.25871512293815613, + "learning_rate": 2.6493990882718612e-06, + "loss": 0.5323, + "step": 5673 + }, + { + "epoch": 0.47028595109821797, + "grad_norm": 0.26304325461387634, + "learning_rate": 2.6489846663903857e-06, + "loss": 0.5377, + "step": 5674 + }, + { + "epoch": 0.4703688354745131, + "grad_norm": 0.2724239230155945, + "learning_rate": 2.64857024450891e-06, + "loss": 0.4584, + "step": 5675 + }, + { + "epoch": 0.47045171985080814, + "grad_norm": 0.25770503282546997, + "learning_rate": 2.648155822627435e-06, + "loss": 0.5499, + "step": 5676 + }, + { + "epoch": 0.4705346042271032, + "grad_norm": 0.27283188700675964, + "learning_rate": 2.6477414007459594e-06, + "loss": 0.4918, + "step": 5677 + }, + { + "epoch": 0.47061748860339825, + "grad_norm": 0.2742738127708435, + "learning_rate": 2.6473269788644847e-06, + "loss": 0.5107, + "step": 5678 + }, + { + "epoch": 0.4707003729796933, + "grad_norm": 0.23984992504119873, + "learning_rate": 2.646912556983009e-06, + "loss": 0.5379, + "step": 5679 + }, + { + "epoch": 0.4707832573559884, + "grad_norm": 0.26721838116645813, + "learning_rate": 2.6464981351015335e-06, + "loss": 0.5395, + "step": 5680 + }, + { + "epoch": 0.47086614173228347, + "grad_norm": 0.26553264260292053, + "learning_rate": 2.6460837132200584e-06, + "loss": 0.5088, + "step": 5681 + }, + { + "epoch": 0.47094902610857853, + "grad_norm": 0.2532651126384735, + "learning_rate": 2.645669291338583e-06, + "loss": 0.5181, + "step": 5682 + }, + { + "epoch": 0.4710319104848736, + "grad_norm": 0.2750825881958008, + "learning_rate": 2.6452548694571072e-06, + "loss": 0.5032, + "step": 5683 + }, + { + "epoch": 0.4711147948611687, + "grad_norm": 0.25606778264045715, + "learning_rate": 2.644840447575632e-06, + "loss": 0.5138, + "step": 5684 + }, + { + "epoch": 0.47119767923746375, + "grad_norm": 0.228797048330307, + "learning_rate": 2.644426025694157e-06, + "loss": 0.4822, + "step": 5685 + }, + { + "epoch": 0.4712805636137588, + "grad_norm": 0.24802790582180023, + "learning_rate": 2.644011603812682e-06, + "loss": 0.4963, + "step": 5686 + }, + { + "epoch": 0.47136344799005386, + "grad_norm": 0.24722741544246674, + "learning_rate": 2.6435971819312062e-06, + "loss": 0.5026, + "step": 5687 + }, + { + "epoch": 0.4714463323663489, + "grad_norm": 0.25182896852493286, + "learning_rate": 2.6431827600497307e-06, + "loss": 0.5527, + "step": 5688 + }, + { + "epoch": 0.47152921674264403, + "grad_norm": 0.26377904415130615, + "learning_rate": 2.6427683381682555e-06, + "loss": 0.5581, + "step": 5689 + }, + { + "epoch": 0.4716121011189391, + "grad_norm": 0.27694907784461975, + "learning_rate": 2.64235391628678e-06, + "loss": 0.5621, + "step": 5690 + }, + { + "epoch": 0.47169498549523414, + "grad_norm": 0.30454692244529724, + "learning_rate": 2.6419394944053052e-06, + "loss": 0.5849, + "step": 5691 + }, + { + "epoch": 0.4717778698715292, + "grad_norm": 0.27568235993385315, + "learning_rate": 2.6415250725238297e-06, + "loss": 0.5922, + "step": 5692 + }, + { + "epoch": 0.4718607542478243, + "grad_norm": 0.2467365860939026, + "learning_rate": 2.641110650642354e-06, + "loss": 0.472, + "step": 5693 + }, + { + "epoch": 0.47194363862411937, + "grad_norm": 0.2550916373729706, + "learning_rate": 2.640696228760879e-06, + "loss": 0.5115, + "step": 5694 + }, + { + "epoch": 0.4720265230004144, + "grad_norm": 0.27639544010162354, + "learning_rate": 2.6402818068794034e-06, + "loss": 0.4733, + "step": 5695 + }, + { + "epoch": 0.4721094073767095, + "grad_norm": 0.2598327696323395, + "learning_rate": 2.6398673849979282e-06, + "loss": 0.5349, + "step": 5696 + }, + { + "epoch": 0.47219229175300453, + "grad_norm": 0.2404664307832718, + "learning_rate": 2.6394529631164527e-06, + "loss": 0.4824, + "step": 5697 + }, + { + "epoch": 0.47227517612929965, + "grad_norm": 0.2676791846752167, + "learning_rate": 2.639038541234977e-06, + "loss": 0.5376, + "step": 5698 + }, + { + "epoch": 0.4723580605055947, + "grad_norm": 0.2431785762310028, + "learning_rate": 2.6386241193535024e-06, + "loss": 0.4932, + "step": 5699 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 0.25426366925239563, + "learning_rate": 2.638209697472027e-06, + "loss": 0.4946, + "step": 5700 + }, + { + "epoch": 0.4725238292581848, + "grad_norm": 0.2481471449136734, + "learning_rate": 2.6377952755905512e-06, + "loss": 0.5109, + "step": 5701 + }, + { + "epoch": 0.4726067136344799, + "grad_norm": 0.27360236644744873, + "learning_rate": 2.637380853709076e-06, + "loss": 0.5616, + "step": 5702 + }, + { + "epoch": 0.472689598010775, + "grad_norm": 0.24652649462223053, + "learning_rate": 2.6369664318276005e-06, + "loss": 0.5107, + "step": 5703 + }, + { + "epoch": 0.47277248238707004, + "grad_norm": 0.27641943097114563, + "learning_rate": 2.6365520099461254e-06, + "loss": 0.5466, + "step": 5704 + }, + { + "epoch": 0.4728553667633651, + "grad_norm": 0.2474568784236908, + "learning_rate": 2.63613758806465e-06, + "loss": 0.5138, + "step": 5705 + }, + { + "epoch": 0.47293825113966015, + "grad_norm": 0.262660950422287, + "learning_rate": 2.6357231661831747e-06, + "loss": 0.5066, + "step": 5706 + }, + { + "epoch": 0.47302113551595526, + "grad_norm": 0.26123684644699097, + "learning_rate": 2.6353087443016995e-06, + "loss": 0.5176, + "step": 5707 + }, + { + "epoch": 0.4731040198922503, + "grad_norm": 0.2613990008831024, + "learning_rate": 2.634894322420224e-06, + "loss": 0.546, + "step": 5708 + }, + { + "epoch": 0.47318690426854537, + "grad_norm": 0.2548488676548004, + "learning_rate": 2.634479900538749e-06, + "loss": 0.5105, + "step": 5709 + }, + { + "epoch": 0.47326978864484043, + "grad_norm": 0.25436681509017944, + "learning_rate": 2.6340654786572733e-06, + "loss": 0.5239, + "step": 5710 + }, + { + "epoch": 0.47335267302113554, + "grad_norm": 0.26693040132522583, + "learning_rate": 2.6336510567757977e-06, + "loss": 0.4723, + "step": 5711 + }, + { + "epoch": 0.4734355573974306, + "grad_norm": 0.24167867004871368, + "learning_rate": 2.633236634894323e-06, + "loss": 0.4807, + "step": 5712 + }, + { + "epoch": 0.47351844177372565, + "grad_norm": 0.2566489577293396, + "learning_rate": 2.6328222130128474e-06, + "loss": 0.5305, + "step": 5713 + }, + { + "epoch": 0.4736013261500207, + "grad_norm": 0.2699125111103058, + "learning_rate": 2.632407791131372e-06, + "loss": 0.5049, + "step": 5714 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 0.241461843252182, + "learning_rate": 2.6319933692498967e-06, + "loss": 0.5132, + "step": 5715 + }, + { + "epoch": 0.4737670949026109, + "grad_norm": 0.26280149817466736, + "learning_rate": 2.631578947368421e-06, + "loss": 0.506, + "step": 5716 + }, + { + "epoch": 0.47384997927890593, + "grad_norm": 0.27395063638687134, + "learning_rate": 2.631164525486946e-06, + "loss": 0.5337, + "step": 5717 + }, + { + "epoch": 0.473932863655201, + "grad_norm": 0.2877342700958252, + "learning_rate": 2.6307501036054704e-06, + "loss": 0.5701, + "step": 5718 + }, + { + "epoch": 0.47401574803149604, + "grad_norm": 0.26247671246528625, + "learning_rate": 2.630335681723995e-06, + "loss": 0.5637, + "step": 5719 + }, + { + "epoch": 0.47409863240779115, + "grad_norm": 0.2537963092327118, + "learning_rate": 2.62992125984252e-06, + "loss": 0.5631, + "step": 5720 + }, + { + "epoch": 0.4741815167840862, + "grad_norm": 0.2553640902042389, + "learning_rate": 2.6295068379610445e-06, + "loss": 0.5042, + "step": 5721 + }, + { + "epoch": 0.47426440116038127, + "grad_norm": 0.2554572522640228, + "learning_rate": 2.6290924160795694e-06, + "loss": 0.5222, + "step": 5722 + }, + { + "epoch": 0.4743472855366763, + "grad_norm": 0.26621195673942566, + "learning_rate": 2.628677994198094e-06, + "loss": 0.5174, + "step": 5723 + }, + { + "epoch": 0.4744301699129714, + "grad_norm": 0.2656167447566986, + "learning_rate": 2.6282635723166183e-06, + "loss": 0.5509, + "step": 5724 + }, + { + "epoch": 0.4745130542892665, + "grad_norm": 0.2591210603713989, + "learning_rate": 2.627849150435143e-06, + "loss": 0.5024, + "step": 5725 + }, + { + "epoch": 0.47459593866556155, + "grad_norm": 0.23908960819244385, + "learning_rate": 2.627434728553668e-06, + "loss": 0.5016, + "step": 5726 + }, + { + "epoch": 0.4746788230418566, + "grad_norm": 0.25411587953567505, + "learning_rate": 2.627020306672193e-06, + "loss": 0.4968, + "step": 5727 + }, + { + "epoch": 0.47476170741815166, + "grad_norm": 0.2468300610780716, + "learning_rate": 2.6266058847907173e-06, + "loss": 0.5017, + "step": 5728 + }, + { + "epoch": 0.47484459179444677, + "grad_norm": 0.24687737226486206, + "learning_rate": 2.6261914629092417e-06, + "loss": 0.4691, + "step": 5729 + }, + { + "epoch": 0.4749274761707418, + "grad_norm": 0.23766262829303741, + "learning_rate": 2.6257770410277665e-06, + "loss": 0.5187, + "step": 5730 + }, + { + "epoch": 0.4750103605470369, + "grad_norm": 0.26478224992752075, + "learning_rate": 2.625362619146291e-06, + "loss": 0.5301, + "step": 5731 + }, + { + "epoch": 0.47509324492333194, + "grad_norm": 0.25079140067100525, + "learning_rate": 2.6249481972648154e-06, + "loss": 0.5287, + "step": 5732 + }, + { + "epoch": 0.47517612929962705, + "grad_norm": 0.24601583182811737, + "learning_rate": 2.6245337753833407e-06, + "loss": 0.4731, + "step": 5733 + }, + { + "epoch": 0.4752590136759221, + "grad_norm": 0.27005863189697266, + "learning_rate": 2.624119353501865e-06, + "loss": 0.5305, + "step": 5734 + }, + { + "epoch": 0.47534189805221716, + "grad_norm": 0.23725152015686035, + "learning_rate": 2.62370493162039e-06, + "loss": 0.4911, + "step": 5735 + }, + { + "epoch": 0.4754247824285122, + "grad_norm": 0.26156020164489746, + "learning_rate": 2.6232905097389144e-06, + "loss": 0.4987, + "step": 5736 + }, + { + "epoch": 0.47550766680480727, + "grad_norm": 0.24334146082401276, + "learning_rate": 2.622876087857439e-06, + "loss": 0.5205, + "step": 5737 + }, + { + "epoch": 0.4755905511811024, + "grad_norm": 0.23657043278217316, + "learning_rate": 2.6224616659759637e-06, + "loss": 0.5073, + "step": 5738 + }, + { + "epoch": 0.47567343555739744, + "grad_norm": 0.2557370364665985, + "learning_rate": 2.622047244094488e-06, + "loss": 0.5348, + "step": 5739 + }, + { + "epoch": 0.4757563199336925, + "grad_norm": 0.24571412801742554, + "learning_rate": 2.6216328222130134e-06, + "loss": 0.5205, + "step": 5740 + }, + { + "epoch": 0.47583920430998755, + "grad_norm": 0.2931174039840698, + "learning_rate": 2.621218400331538e-06, + "loss": 0.5409, + "step": 5741 + }, + { + "epoch": 0.47592208868628266, + "grad_norm": 0.24147149920463562, + "learning_rate": 2.6208039784500623e-06, + "loss": 0.5186, + "step": 5742 + }, + { + "epoch": 0.4760049730625777, + "grad_norm": 0.25372087955474854, + "learning_rate": 2.620389556568587e-06, + "loss": 0.5294, + "step": 5743 + }, + { + "epoch": 0.4760878574388728, + "grad_norm": 0.25367823243141174, + "learning_rate": 2.6199751346871116e-06, + "loss": 0.5552, + "step": 5744 + }, + { + "epoch": 0.47617074181516783, + "grad_norm": 0.25404128432273865, + "learning_rate": 2.6195607128056364e-06, + "loss": 0.505, + "step": 5745 + }, + { + "epoch": 0.4762536261914629, + "grad_norm": 0.27092230319976807, + "learning_rate": 2.6191462909241613e-06, + "loss": 0.5404, + "step": 5746 + }, + { + "epoch": 0.476336510567758, + "grad_norm": 0.28024283051490784, + "learning_rate": 2.6187318690426857e-06, + "loss": 0.5546, + "step": 5747 + }, + { + "epoch": 0.47641939494405305, + "grad_norm": 0.24908216297626495, + "learning_rate": 2.6183174471612106e-06, + "loss": 0.4942, + "step": 5748 + }, + { + "epoch": 0.4765022793203481, + "grad_norm": 0.25008198618888855, + "learning_rate": 2.617903025279735e-06, + "loss": 0.5275, + "step": 5749 + }, + { + "epoch": 0.47658516369664317, + "grad_norm": 0.24647140502929688, + "learning_rate": 2.6174886033982594e-06, + "loss": 0.519, + "step": 5750 + }, + { + "epoch": 0.4766680480729383, + "grad_norm": 0.2632022202014923, + "learning_rate": 2.6170741815167843e-06, + "loss": 0.5405, + "step": 5751 + }, + { + "epoch": 0.47675093244923333, + "grad_norm": 0.2503550946712494, + "learning_rate": 2.6166597596353087e-06, + "loss": 0.5297, + "step": 5752 + }, + { + "epoch": 0.4768338168255284, + "grad_norm": 0.25704634189605713, + "learning_rate": 2.616245337753834e-06, + "loss": 0.5405, + "step": 5753 + }, + { + "epoch": 0.47691670120182345, + "grad_norm": 0.24542924761772156, + "learning_rate": 2.6158309158723584e-06, + "loss": 0.5231, + "step": 5754 + }, + { + "epoch": 0.4769995855781185, + "grad_norm": 0.23503181338310242, + "learning_rate": 2.615416493990883e-06, + "loss": 0.4673, + "step": 5755 + }, + { + "epoch": 0.4770824699544136, + "grad_norm": 0.26469293236732483, + "learning_rate": 2.6150020721094077e-06, + "loss": 0.5095, + "step": 5756 + }, + { + "epoch": 0.47716535433070867, + "grad_norm": 0.2584437429904938, + "learning_rate": 2.614587650227932e-06, + "loss": 0.4866, + "step": 5757 + }, + { + "epoch": 0.4772482387070037, + "grad_norm": 0.29157301783561707, + "learning_rate": 2.614173228346457e-06, + "loss": 0.5226, + "step": 5758 + }, + { + "epoch": 0.4773311230832988, + "grad_norm": 0.24596603214740753, + "learning_rate": 2.6137588064649814e-06, + "loss": 0.4946, + "step": 5759 + }, + { + "epoch": 0.4774140074595939, + "grad_norm": 0.28377410769462585, + "learning_rate": 2.613344384583506e-06, + "loss": 0.5327, + "step": 5760 + }, + { + "epoch": 0.47749689183588895, + "grad_norm": 0.25890737771987915, + "learning_rate": 2.612929962702031e-06, + "loss": 0.5093, + "step": 5761 + }, + { + "epoch": 0.477579776212184, + "grad_norm": 0.26149943470954895, + "learning_rate": 2.6125155408205556e-06, + "loss": 0.5217, + "step": 5762 + }, + { + "epoch": 0.47766266058847906, + "grad_norm": 0.24482212960720062, + "learning_rate": 2.6121011189390804e-06, + "loss": 0.4885, + "step": 5763 + }, + { + "epoch": 0.4777455449647741, + "grad_norm": 0.23957198858261108, + "learning_rate": 2.611686697057605e-06, + "loss": 0.5078, + "step": 5764 + }, + { + "epoch": 0.4778284293410692, + "grad_norm": 0.2704732120037079, + "learning_rate": 2.6112722751761293e-06, + "loss": 0.5454, + "step": 5765 + }, + { + "epoch": 0.4779113137173643, + "grad_norm": 0.26083439588546753, + "learning_rate": 2.6108578532946546e-06, + "loss": 0.5259, + "step": 5766 + }, + { + "epoch": 0.47799419809365934, + "grad_norm": 0.2468865066766739, + "learning_rate": 2.610443431413179e-06, + "loss": 0.4935, + "step": 5767 + }, + { + "epoch": 0.4780770824699544, + "grad_norm": 0.2650865316390991, + "learning_rate": 2.6100290095317034e-06, + "loss": 0.5288, + "step": 5768 + }, + { + "epoch": 0.4781599668462495, + "grad_norm": 0.23245882987976074, + "learning_rate": 2.6096145876502283e-06, + "loss": 0.4994, + "step": 5769 + }, + { + "epoch": 0.47824285122254456, + "grad_norm": 0.25583791732788086, + "learning_rate": 2.6092001657687527e-06, + "loss": 0.5377, + "step": 5770 + }, + { + "epoch": 0.4783257355988396, + "grad_norm": 0.22981128096580505, + "learning_rate": 2.6087857438872776e-06, + "loss": 0.479, + "step": 5771 + }, + { + "epoch": 0.4784086199751347, + "grad_norm": 0.27677616477012634, + "learning_rate": 2.608371322005802e-06, + "loss": 0.5231, + "step": 5772 + }, + { + "epoch": 0.47849150435142973, + "grad_norm": 0.2589089274406433, + "learning_rate": 2.6079569001243264e-06, + "loss": 0.5292, + "step": 5773 + }, + { + "epoch": 0.47857438872772484, + "grad_norm": 0.2407745122909546, + "learning_rate": 2.6075424782428517e-06, + "loss": 0.4897, + "step": 5774 + }, + { + "epoch": 0.4786572731040199, + "grad_norm": 0.2578321695327759, + "learning_rate": 2.607128056361376e-06, + "loss": 0.5128, + "step": 5775 + }, + { + "epoch": 0.47874015748031495, + "grad_norm": 0.2624182105064392, + "learning_rate": 2.606713634479901e-06, + "loss": 0.4928, + "step": 5776 + }, + { + "epoch": 0.47882304185661, + "grad_norm": 0.2744714319705963, + "learning_rate": 2.6062992125984254e-06, + "loss": 0.513, + "step": 5777 + }, + { + "epoch": 0.4789059262329051, + "grad_norm": 0.25471898913383484, + "learning_rate": 2.60588479071695e-06, + "loss": 0.5583, + "step": 5778 + }, + { + "epoch": 0.4789888106092002, + "grad_norm": 0.2604016959667206, + "learning_rate": 2.6054703688354747e-06, + "loss": 0.5015, + "step": 5779 + }, + { + "epoch": 0.47907169498549523, + "grad_norm": 0.2709046006202698, + "learning_rate": 2.605055946953999e-06, + "loss": 0.5578, + "step": 5780 + }, + { + "epoch": 0.4791545793617903, + "grad_norm": 0.28705915808677673, + "learning_rate": 2.6046415250725236e-06, + "loss": 0.586, + "step": 5781 + }, + { + "epoch": 0.47923746373808535, + "grad_norm": 0.25185656547546387, + "learning_rate": 2.604227103191049e-06, + "loss": 0.4971, + "step": 5782 + }, + { + "epoch": 0.47932034811438046, + "grad_norm": 0.25847291946411133, + "learning_rate": 2.6038126813095733e-06, + "loss": 0.5002, + "step": 5783 + }, + { + "epoch": 0.4794032324906755, + "grad_norm": 0.24797576665878296, + "learning_rate": 2.603398259428098e-06, + "loss": 0.5173, + "step": 5784 + }, + { + "epoch": 0.47948611686697057, + "grad_norm": 0.24691501259803772, + "learning_rate": 2.6029838375466226e-06, + "loss": 0.4677, + "step": 5785 + }, + { + "epoch": 0.4795690012432656, + "grad_norm": 0.24458158016204834, + "learning_rate": 2.602569415665147e-06, + "loss": 0.4869, + "step": 5786 + }, + { + "epoch": 0.47965188561956074, + "grad_norm": 0.2478497475385666, + "learning_rate": 2.6021549937836723e-06, + "loss": 0.4993, + "step": 5787 + }, + { + "epoch": 0.4797347699958558, + "grad_norm": 0.26062577962875366, + "learning_rate": 2.6017405719021967e-06, + "loss": 0.4788, + "step": 5788 + }, + { + "epoch": 0.47981765437215085, + "grad_norm": 0.26445531845092773, + "learning_rate": 2.6013261500207216e-06, + "loss": 0.4872, + "step": 5789 + }, + { + "epoch": 0.4799005387484459, + "grad_norm": 0.23103898763656616, + "learning_rate": 2.600911728139246e-06, + "loss": 0.4802, + "step": 5790 + }, + { + "epoch": 0.47998342312474096, + "grad_norm": 0.2546352744102478, + "learning_rate": 2.6004973062577704e-06, + "loss": 0.5081, + "step": 5791 + }, + { + "epoch": 0.48006630750103607, + "grad_norm": 0.2558031678199768, + "learning_rate": 2.6000828843762953e-06, + "loss": 0.5058, + "step": 5792 + }, + { + "epoch": 0.4801491918773311, + "grad_norm": 0.2607472836971283, + "learning_rate": 2.5996684624948197e-06, + "loss": 0.5697, + "step": 5793 + }, + { + "epoch": 0.4802320762536262, + "grad_norm": 0.2722508907318115, + "learning_rate": 2.599254040613345e-06, + "loss": 0.4991, + "step": 5794 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 0.2583223879337311, + "learning_rate": 2.5988396187318694e-06, + "loss": 0.5067, + "step": 5795 + }, + { + "epoch": 0.48039784500621635, + "grad_norm": 0.2624218165874481, + "learning_rate": 2.598425196850394e-06, + "loss": 0.5323, + "step": 5796 + }, + { + "epoch": 0.4804807293825114, + "grad_norm": 0.2645767033100128, + "learning_rate": 2.5980107749689187e-06, + "loss": 0.5403, + "step": 5797 + }, + { + "epoch": 0.48056361375880646, + "grad_norm": 0.2621195316314697, + "learning_rate": 2.597596353087443e-06, + "loss": 0.5146, + "step": 5798 + }, + { + "epoch": 0.4806464981351015, + "grad_norm": 0.23939131200313568, + "learning_rate": 2.5971819312059676e-06, + "loss": 0.4899, + "step": 5799 + }, + { + "epoch": 0.4807293825113966, + "grad_norm": 0.22253605723381042, + "learning_rate": 2.5967675093244924e-06, + "loss": 0.4857, + "step": 5800 + }, + { + "epoch": 0.4808122668876917, + "grad_norm": 0.26596906781196594, + "learning_rate": 2.596353087443017e-06, + "loss": 0.5205, + "step": 5801 + }, + { + "epoch": 0.48089515126398674, + "grad_norm": 0.2762465476989746, + "learning_rate": 2.595938665561542e-06, + "loss": 0.5244, + "step": 5802 + }, + { + "epoch": 0.4809780356402818, + "grad_norm": 0.2752887010574341, + "learning_rate": 2.5955242436800666e-06, + "loss": 0.5414, + "step": 5803 + }, + { + "epoch": 0.48106092001657685, + "grad_norm": 0.2369517982006073, + "learning_rate": 2.595109821798591e-06, + "loss": 0.4974, + "step": 5804 + }, + { + "epoch": 0.48114380439287197, + "grad_norm": 0.24807307124137878, + "learning_rate": 2.594695399917116e-06, + "loss": 0.5185, + "step": 5805 + }, + { + "epoch": 0.481226688769167, + "grad_norm": 0.2893737852573395, + "learning_rate": 2.5942809780356403e-06, + "loss": 0.5144, + "step": 5806 + }, + { + "epoch": 0.4813095731454621, + "grad_norm": 0.2548023760318756, + "learning_rate": 2.5938665561541656e-06, + "loss": 0.5268, + "step": 5807 + }, + { + "epoch": 0.48139245752175713, + "grad_norm": 0.23907768726348877, + "learning_rate": 2.59345213427269e-06, + "loss": 0.4878, + "step": 5808 + }, + { + "epoch": 0.48147534189805224, + "grad_norm": 0.2634356915950775, + "learning_rate": 2.5930377123912144e-06, + "loss": 0.5361, + "step": 5809 + }, + { + "epoch": 0.4815582262743473, + "grad_norm": 0.2545965611934662, + "learning_rate": 2.5926232905097393e-06, + "loss": 0.5272, + "step": 5810 + }, + { + "epoch": 0.48164111065064236, + "grad_norm": 0.2619999349117279, + "learning_rate": 2.5922088686282637e-06, + "loss": 0.5452, + "step": 5811 + }, + { + "epoch": 0.4817239950269374, + "grad_norm": 0.2414710968732834, + "learning_rate": 2.5917944467467886e-06, + "loss": 0.4886, + "step": 5812 + }, + { + "epoch": 0.48180687940323247, + "grad_norm": 0.24702104926109314, + "learning_rate": 2.591380024865313e-06, + "loss": 0.5029, + "step": 5813 + }, + { + "epoch": 0.4818897637795276, + "grad_norm": 0.25117990374565125, + "learning_rate": 2.5909656029838375e-06, + "loss": 0.4792, + "step": 5814 + }, + { + "epoch": 0.48197264815582264, + "grad_norm": 0.2582029700279236, + "learning_rate": 2.5905511811023627e-06, + "loss": 0.5062, + "step": 5815 + }, + { + "epoch": 0.4820555325321177, + "grad_norm": 0.26894083619117737, + "learning_rate": 2.590136759220887e-06, + "loss": 0.5291, + "step": 5816 + }, + { + "epoch": 0.48213841690841275, + "grad_norm": 0.2567923665046692, + "learning_rate": 2.5897223373394116e-06, + "loss": 0.532, + "step": 5817 + }, + { + "epoch": 0.48222130128470786, + "grad_norm": 0.26115885376930237, + "learning_rate": 2.5893079154579364e-06, + "loss": 0.5099, + "step": 5818 + }, + { + "epoch": 0.4823041856610029, + "grad_norm": 0.26032954454421997, + "learning_rate": 2.588893493576461e-06, + "loss": 0.5316, + "step": 5819 + }, + { + "epoch": 0.48238707003729797, + "grad_norm": 0.2690945565700531, + "learning_rate": 2.5884790716949857e-06, + "loss": 0.5308, + "step": 5820 + }, + { + "epoch": 0.482469954413593, + "grad_norm": 0.2474420815706253, + "learning_rate": 2.58806464981351e-06, + "loss": 0.5092, + "step": 5821 + }, + { + "epoch": 0.4825528387898881, + "grad_norm": 0.2556314766407013, + "learning_rate": 2.5876502279320346e-06, + "loss": 0.5435, + "step": 5822 + }, + { + "epoch": 0.4826357231661832, + "grad_norm": 0.27080750465393066, + "learning_rate": 2.58723580605056e-06, + "loss": 0.5114, + "step": 5823 + }, + { + "epoch": 0.48271860754247825, + "grad_norm": 0.26844868063926697, + "learning_rate": 2.5868213841690843e-06, + "loss": 0.5591, + "step": 5824 + }, + { + "epoch": 0.4828014919187733, + "grad_norm": 0.25653505325317383, + "learning_rate": 2.586406962287609e-06, + "loss": 0.5475, + "step": 5825 + }, + { + "epoch": 0.48288437629506836, + "grad_norm": 0.2530212998390198, + "learning_rate": 2.5859925404061336e-06, + "loss": 0.4915, + "step": 5826 + }, + { + "epoch": 0.4829672606713635, + "grad_norm": 0.2536596357822418, + "learning_rate": 2.585578118524658e-06, + "loss": 0.5183, + "step": 5827 + }, + { + "epoch": 0.48305014504765853, + "grad_norm": 0.2665823698043823, + "learning_rate": 2.5851636966431833e-06, + "loss": 0.5542, + "step": 5828 + }, + { + "epoch": 0.4831330294239536, + "grad_norm": 0.25507017970085144, + "learning_rate": 2.5847492747617077e-06, + "loss": 0.5401, + "step": 5829 + }, + { + "epoch": 0.48321591380024864, + "grad_norm": 0.25582748651504517, + "learning_rate": 2.5843348528802326e-06, + "loss": 0.5057, + "step": 5830 + }, + { + "epoch": 0.4832987981765437, + "grad_norm": 0.25517383217811584, + "learning_rate": 2.583920430998757e-06, + "loss": 0.5757, + "step": 5831 + }, + { + "epoch": 0.4833816825528388, + "grad_norm": 0.2601606845855713, + "learning_rate": 2.5835060091172815e-06, + "loss": 0.5125, + "step": 5832 + }, + { + "epoch": 0.48346456692913387, + "grad_norm": 0.24849390983581543, + "learning_rate": 2.5830915872358063e-06, + "loss": 0.5112, + "step": 5833 + }, + { + "epoch": 0.4835474513054289, + "grad_norm": 0.23640567064285278, + "learning_rate": 2.5826771653543307e-06, + "loss": 0.4872, + "step": 5834 + }, + { + "epoch": 0.483630335681724, + "grad_norm": 0.25871697068214417, + "learning_rate": 2.582262743472855e-06, + "loss": 0.5343, + "step": 5835 + }, + { + "epoch": 0.4837132200580191, + "grad_norm": 0.2454739511013031, + "learning_rate": 2.5818483215913805e-06, + "loss": 0.5007, + "step": 5836 + }, + { + "epoch": 0.48379610443431414, + "grad_norm": 0.2642514705657959, + "learning_rate": 2.581433899709905e-06, + "loss": 0.5145, + "step": 5837 + }, + { + "epoch": 0.4838789888106092, + "grad_norm": 0.24081431329250336, + "learning_rate": 2.5810194778284297e-06, + "loss": 0.5369, + "step": 5838 + }, + { + "epoch": 0.48396187318690426, + "grad_norm": 0.24277663230895996, + "learning_rate": 2.580605055946954e-06, + "loss": 0.4628, + "step": 5839 + }, + { + "epoch": 0.4840447575631993, + "grad_norm": 0.2594192326068878, + "learning_rate": 2.5801906340654786e-06, + "loss": 0.524, + "step": 5840 + }, + { + "epoch": 0.4841276419394944, + "grad_norm": 0.257331520318985, + "learning_rate": 2.5797762121840035e-06, + "loss": 0.509, + "step": 5841 + }, + { + "epoch": 0.4842105263157895, + "grad_norm": 0.2341834455728531, + "learning_rate": 2.579361790302528e-06, + "loss": 0.5223, + "step": 5842 + }, + { + "epoch": 0.48429341069208454, + "grad_norm": 0.26102593541145325, + "learning_rate": 2.578947368421053e-06, + "loss": 0.5029, + "step": 5843 + }, + { + "epoch": 0.4843762950683796, + "grad_norm": 0.24391363561153412, + "learning_rate": 2.5785329465395776e-06, + "loss": 0.5045, + "step": 5844 + }, + { + "epoch": 0.4844591794446747, + "grad_norm": 0.2507432699203491, + "learning_rate": 2.578118524658102e-06, + "loss": 0.5173, + "step": 5845 + }, + { + "epoch": 0.48454206382096976, + "grad_norm": 0.2952359616756439, + "learning_rate": 2.577704102776627e-06, + "loss": 0.542, + "step": 5846 + }, + { + "epoch": 0.4846249481972648, + "grad_norm": 0.2524542808532715, + "learning_rate": 2.5772896808951513e-06, + "loss": 0.4968, + "step": 5847 + }, + { + "epoch": 0.48470783257355987, + "grad_norm": 0.2695816159248352, + "learning_rate": 2.5768752590136758e-06, + "loss": 0.559, + "step": 5848 + }, + { + "epoch": 0.4847907169498549, + "grad_norm": 0.23882946372032166, + "learning_rate": 2.576460837132201e-06, + "loss": 0.4901, + "step": 5849 + }, + { + "epoch": 0.48487360132615004, + "grad_norm": 0.2681351602077484, + "learning_rate": 2.5760464152507255e-06, + "loss": 0.531, + "step": 5850 + }, + { + "epoch": 0.4849564857024451, + "grad_norm": 0.2668311595916748, + "learning_rate": 2.5756319933692503e-06, + "loss": 0.5316, + "step": 5851 + }, + { + "epoch": 0.48503937007874015, + "grad_norm": 0.2652503550052643, + "learning_rate": 2.5752175714877748e-06, + "loss": 0.528, + "step": 5852 + }, + { + "epoch": 0.4851222544550352, + "grad_norm": 0.2538180351257324, + "learning_rate": 2.574803149606299e-06, + "loss": 0.5294, + "step": 5853 + }, + { + "epoch": 0.4852051388313303, + "grad_norm": 0.2499082386493683, + "learning_rate": 2.574388727724824e-06, + "loss": 0.4592, + "step": 5854 + }, + { + "epoch": 0.4852880232076254, + "grad_norm": 0.2537841498851776, + "learning_rate": 2.5739743058433485e-06, + "loss": 0.5132, + "step": 5855 + }, + { + "epoch": 0.48537090758392043, + "grad_norm": 0.25334635376930237, + "learning_rate": 2.5735598839618738e-06, + "loss": 0.5399, + "step": 5856 + }, + { + "epoch": 0.4854537919602155, + "grad_norm": 0.25488021969795227, + "learning_rate": 2.573145462080398e-06, + "loss": 0.5344, + "step": 5857 + }, + { + "epoch": 0.48553667633651054, + "grad_norm": 0.25886785984039307, + "learning_rate": 2.5727310401989226e-06, + "loss": 0.5035, + "step": 5858 + }, + { + "epoch": 0.48561956071280565, + "grad_norm": 0.2627842128276825, + "learning_rate": 2.5723166183174475e-06, + "loss": 0.5273, + "step": 5859 + }, + { + "epoch": 0.4857024450891007, + "grad_norm": 0.25087428092956543, + "learning_rate": 2.571902196435972e-06, + "loss": 0.5039, + "step": 5860 + }, + { + "epoch": 0.48578532946539577, + "grad_norm": 0.25915393233299255, + "learning_rate": 2.5714877745544968e-06, + "loss": 0.5417, + "step": 5861 + }, + { + "epoch": 0.4858682138416908, + "grad_norm": 0.24924366176128387, + "learning_rate": 2.571073352673021e-06, + "loss": 0.5507, + "step": 5862 + }, + { + "epoch": 0.48595109821798593, + "grad_norm": 0.2605861723423004, + "learning_rate": 2.5706589307915456e-06, + "loss": 0.5354, + "step": 5863 + }, + { + "epoch": 0.486033982594281, + "grad_norm": 0.25789007544517517, + "learning_rate": 2.570244508910071e-06, + "loss": 0.5419, + "step": 5864 + }, + { + "epoch": 0.48611686697057604, + "grad_norm": 0.25209346413612366, + "learning_rate": 2.5698300870285953e-06, + "loss": 0.4892, + "step": 5865 + }, + { + "epoch": 0.4861997513468711, + "grad_norm": 0.26945358514785767, + "learning_rate": 2.5694156651471198e-06, + "loss": 0.5439, + "step": 5866 + }, + { + "epoch": 0.48628263572316616, + "grad_norm": 0.2506932020187378, + "learning_rate": 2.5690012432656446e-06, + "loss": 0.4808, + "step": 5867 + }, + { + "epoch": 0.48636552009946127, + "grad_norm": 0.2511362135410309, + "learning_rate": 2.568586821384169e-06, + "loss": 0.5351, + "step": 5868 + }, + { + "epoch": 0.4864484044757563, + "grad_norm": 0.26517897844314575, + "learning_rate": 2.5681723995026943e-06, + "loss": 0.5513, + "step": 5869 + }, + { + "epoch": 0.4865312888520514, + "grad_norm": 0.23773321509361267, + "learning_rate": 2.5677579776212188e-06, + "loss": 0.5118, + "step": 5870 + }, + { + "epoch": 0.48661417322834644, + "grad_norm": 0.265591561794281, + "learning_rate": 2.567343555739743e-06, + "loss": 0.564, + "step": 5871 + }, + { + "epoch": 0.48669705760464155, + "grad_norm": 0.24049176275730133, + "learning_rate": 2.566929133858268e-06, + "loss": 0.5248, + "step": 5872 + }, + { + "epoch": 0.4867799419809366, + "grad_norm": 0.2457093596458435, + "learning_rate": 2.5665147119767925e-06, + "loss": 0.4654, + "step": 5873 + }, + { + "epoch": 0.48686282635723166, + "grad_norm": 0.2764091491699219, + "learning_rate": 2.5661002900953173e-06, + "loss": 0.5306, + "step": 5874 + }, + { + "epoch": 0.4869457107335267, + "grad_norm": 0.26920151710510254, + "learning_rate": 2.5656858682138418e-06, + "loss": 0.5178, + "step": 5875 + }, + { + "epoch": 0.48702859510982177, + "grad_norm": 0.2392544001340866, + "learning_rate": 2.565271446332366e-06, + "loss": 0.5021, + "step": 5876 + }, + { + "epoch": 0.4871114794861169, + "grad_norm": 0.28225451707839966, + "learning_rate": 2.5648570244508915e-06, + "loss": 0.5529, + "step": 5877 + }, + { + "epoch": 0.48719436386241194, + "grad_norm": 0.25344157218933105, + "learning_rate": 2.564442602569416e-06, + "loss": 0.5238, + "step": 5878 + }, + { + "epoch": 0.487277248238707, + "grad_norm": 0.2512068748474121, + "learning_rate": 2.5640281806879408e-06, + "loss": 0.5527, + "step": 5879 + }, + { + "epoch": 0.48736013261500205, + "grad_norm": 0.2537855803966522, + "learning_rate": 2.563613758806465e-06, + "loss": 0.5532, + "step": 5880 + }, + { + "epoch": 0.48744301699129716, + "grad_norm": 0.24672168493270874, + "learning_rate": 2.5631993369249896e-06, + "loss": 0.4923, + "step": 5881 + }, + { + "epoch": 0.4875259013675922, + "grad_norm": 0.2817804515361786, + "learning_rate": 2.5627849150435145e-06, + "loss": 0.5818, + "step": 5882 + }, + { + "epoch": 0.4876087857438873, + "grad_norm": 0.26795998215675354, + "learning_rate": 2.562370493162039e-06, + "loss": 0.5372, + "step": 5883 + }, + { + "epoch": 0.48769167012018233, + "grad_norm": 0.24918022751808167, + "learning_rate": 2.5619560712805638e-06, + "loss": 0.5065, + "step": 5884 + }, + { + "epoch": 0.48777455449647744, + "grad_norm": 0.26688987016677856, + "learning_rate": 2.5615416493990886e-06, + "loss": 0.53, + "step": 5885 + }, + { + "epoch": 0.4878574388727725, + "grad_norm": 0.2470758557319641, + "learning_rate": 2.561127227517613e-06, + "loss": 0.4857, + "step": 5886 + }, + { + "epoch": 0.48794032324906755, + "grad_norm": 0.23014038801193237, + "learning_rate": 2.560712805636138e-06, + "loss": 0.5029, + "step": 5887 + }, + { + "epoch": 0.4880232076253626, + "grad_norm": 0.2475268691778183, + "learning_rate": 2.5602983837546623e-06, + "loss": 0.552, + "step": 5888 + }, + { + "epoch": 0.48810609200165767, + "grad_norm": 0.2547178566455841, + "learning_rate": 2.5598839618731868e-06, + "loss": 0.5328, + "step": 5889 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 0.27943122386932373, + "learning_rate": 2.559469539991712e-06, + "loss": 0.5479, + "step": 5890 + }, + { + "epoch": 0.48827186075424783, + "grad_norm": 0.27502384781837463, + "learning_rate": 2.5590551181102365e-06, + "loss": 0.5192, + "step": 5891 + }, + { + "epoch": 0.4883547451305429, + "grad_norm": 0.2728174030780792, + "learning_rate": 2.5586406962287613e-06, + "loss": 0.5623, + "step": 5892 + }, + { + "epoch": 0.48843762950683794, + "grad_norm": 0.2510896325111389, + "learning_rate": 2.5582262743472858e-06, + "loss": 0.4936, + "step": 5893 + }, + { + "epoch": 0.48852051388313306, + "grad_norm": 0.2418404221534729, + "learning_rate": 2.55781185246581e-06, + "loss": 0.5083, + "step": 5894 + }, + { + "epoch": 0.4886033982594281, + "grad_norm": 0.2388509064912796, + "learning_rate": 2.557397430584335e-06, + "loss": 0.4909, + "step": 5895 + }, + { + "epoch": 0.48868628263572317, + "grad_norm": 0.2525002062320709, + "learning_rate": 2.5569830087028595e-06, + "loss": 0.5375, + "step": 5896 + }, + { + "epoch": 0.4887691670120182, + "grad_norm": 0.2648436427116394, + "learning_rate": 2.5565685868213848e-06, + "loss": 0.5412, + "step": 5897 + }, + { + "epoch": 0.4888520513883133, + "grad_norm": 0.24487504363059998, + "learning_rate": 2.556154164939909e-06, + "loss": 0.477, + "step": 5898 + }, + { + "epoch": 0.4889349357646084, + "grad_norm": 0.27556517720222473, + "learning_rate": 2.5557397430584336e-06, + "loss": 0.5281, + "step": 5899 + }, + { + "epoch": 0.48901782014090345, + "grad_norm": 0.2510826885700226, + "learning_rate": 2.5553253211769585e-06, + "loss": 0.4998, + "step": 5900 + }, + { + "epoch": 0.4891007045171985, + "grad_norm": 0.3286239206790924, + "learning_rate": 2.554910899295483e-06, + "loss": 0.5576, + "step": 5901 + }, + { + "epoch": 0.48918358889349356, + "grad_norm": 0.23658527433872223, + "learning_rate": 2.5544964774140074e-06, + "loss": 0.4921, + "step": 5902 + }, + { + "epoch": 0.48926647326978867, + "grad_norm": 0.2517411708831787, + "learning_rate": 2.554082055532532e-06, + "loss": 0.483, + "step": 5903 + }, + { + "epoch": 0.4893493576460837, + "grad_norm": 0.25992804765701294, + "learning_rate": 2.553667633651057e-06, + "loss": 0.5041, + "step": 5904 + }, + { + "epoch": 0.4894322420223788, + "grad_norm": 0.24978138506412506, + "learning_rate": 2.553253211769582e-06, + "loss": 0.5268, + "step": 5905 + }, + { + "epoch": 0.48951512639867384, + "grad_norm": 0.254518985748291, + "learning_rate": 2.5528387898881064e-06, + "loss": 0.5278, + "step": 5906 + }, + { + "epoch": 0.4895980107749689, + "grad_norm": 0.2768123149871826, + "learning_rate": 2.5524243680066308e-06, + "loss": 0.5403, + "step": 5907 + }, + { + "epoch": 0.489680895151264, + "grad_norm": 0.2631494402885437, + "learning_rate": 2.5520099461251556e-06, + "loss": 0.5114, + "step": 5908 + }, + { + "epoch": 0.48976377952755906, + "grad_norm": 0.24022458493709564, + "learning_rate": 2.55159552424368e-06, + "loss": 0.5216, + "step": 5909 + }, + { + "epoch": 0.4898466639038541, + "grad_norm": 0.2688482999801636, + "learning_rate": 2.5511811023622053e-06, + "loss": 0.5044, + "step": 5910 + }, + { + "epoch": 0.4899295482801492, + "grad_norm": 0.2682500183582306, + "learning_rate": 2.5507666804807298e-06, + "loss": 0.5603, + "step": 5911 + }, + { + "epoch": 0.4900124326564443, + "grad_norm": 0.2483581155538559, + "learning_rate": 2.5503522585992542e-06, + "loss": 0.5323, + "step": 5912 + }, + { + "epoch": 0.49009531703273934, + "grad_norm": 0.23039260506629944, + "learning_rate": 2.549937836717779e-06, + "loss": 0.4693, + "step": 5913 + }, + { + "epoch": 0.4901782014090344, + "grad_norm": 0.24972036480903625, + "learning_rate": 2.5495234148363035e-06, + "loss": 0.4942, + "step": 5914 + }, + { + "epoch": 0.49026108578532945, + "grad_norm": 0.2615238428115845, + "learning_rate": 2.5491089929548284e-06, + "loss": 0.5322, + "step": 5915 + }, + { + "epoch": 0.4903439701616245, + "grad_norm": 0.26579147577285767, + "learning_rate": 2.5486945710733528e-06, + "loss": 0.5566, + "step": 5916 + }, + { + "epoch": 0.4904268545379196, + "grad_norm": 0.24552784860134125, + "learning_rate": 2.5482801491918772e-06, + "loss": 0.4762, + "step": 5917 + }, + { + "epoch": 0.4905097389142147, + "grad_norm": 0.25553956627845764, + "learning_rate": 2.5478657273104025e-06, + "loss": 0.53, + "step": 5918 + }, + { + "epoch": 0.49059262329050973, + "grad_norm": 0.26213860511779785, + "learning_rate": 2.547451305428927e-06, + "loss": 0.5446, + "step": 5919 + }, + { + "epoch": 0.4906755076668048, + "grad_norm": 0.28659361600875854, + "learning_rate": 2.5470368835474514e-06, + "loss": 0.5288, + "step": 5920 + }, + { + "epoch": 0.4907583920430999, + "grad_norm": 0.27178218960762024, + "learning_rate": 2.5466224616659762e-06, + "loss": 0.5155, + "step": 5921 + }, + { + "epoch": 0.49084127641939496, + "grad_norm": 0.2513246238231659, + "learning_rate": 2.5462080397845006e-06, + "loss": 0.5002, + "step": 5922 + }, + { + "epoch": 0.49092416079569, + "grad_norm": 0.24954025447368622, + "learning_rate": 2.5457936179030255e-06, + "loss": 0.5392, + "step": 5923 + }, + { + "epoch": 0.49100704517198507, + "grad_norm": 0.2703867256641388, + "learning_rate": 2.54537919602155e-06, + "loss": 0.5214, + "step": 5924 + }, + { + "epoch": 0.4910899295482801, + "grad_norm": 0.25962987542152405, + "learning_rate": 2.544964774140075e-06, + "loss": 0.5593, + "step": 5925 + }, + { + "epoch": 0.49117281392457524, + "grad_norm": 0.2414228469133377, + "learning_rate": 2.5445503522585996e-06, + "loss": 0.492, + "step": 5926 + }, + { + "epoch": 0.4912556983008703, + "grad_norm": 0.25645026564598083, + "learning_rate": 2.544135930377124e-06, + "loss": 0.5335, + "step": 5927 + }, + { + "epoch": 0.49133858267716535, + "grad_norm": 0.2550852596759796, + "learning_rate": 2.543721508495649e-06, + "loss": 0.5474, + "step": 5928 + }, + { + "epoch": 0.4914214670534604, + "grad_norm": 0.2600247263908386, + "learning_rate": 2.5433070866141734e-06, + "loss": 0.4943, + "step": 5929 + }, + { + "epoch": 0.4915043514297555, + "grad_norm": 0.25070667266845703, + "learning_rate": 2.542892664732698e-06, + "loss": 0.5099, + "step": 5930 + }, + { + "epoch": 0.49158723580605057, + "grad_norm": 0.24991928040981293, + "learning_rate": 2.542478242851223e-06, + "loss": 0.5113, + "step": 5931 + }, + { + "epoch": 0.4916701201823456, + "grad_norm": 0.26536256074905396, + "learning_rate": 2.5420638209697475e-06, + "loss": 0.5622, + "step": 5932 + }, + { + "epoch": 0.4917530045586407, + "grad_norm": 0.2855004072189331, + "learning_rate": 2.541649399088272e-06, + "loss": 0.519, + "step": 5933 + }, + { + "epoch": 0.49183588893493574, + "grad_norm": 0.2612053155899048, + "learning_rate": 2.541234977206797e-06, + "loss": 0.5403, + "step": 5934 + }, + { + "epoch": 0.49191877331123085, + "grad_norm": 0.2663303315639496, + "learning_rate": 2.5408205553253212e-06, + "loss": 0.5328, + "step": 5935 + }, + { + "epoch": 0.4920016576875259, + "grad_norm": 0.25691601634025574, + "learning_rate": 2.540406133443846e-06, + "loss": 0.5064, + "step": 5936 + }, + { + "epoch": 0.49208454206382096, + "grad_norm": 0.2666735053062439, + "learning_rate": 2.5399917115623705e-06, + "loss": 0.5812, + "step": 5937 + }, + { + "epoch": 0.492167426440116, + "grad_norm": 0.5552436709403992, + "learning_rate": 2.539577289680895e-06, + "loss": 0.4937, + "step": 5938 + }, + { + "epoch": 0.49225031081641113, + "grad_norm": 0.26110705733299255, + "learning_rate": 2.5391628677994202e-06, + "loss": 0.5801, + "step": 5939 + }, + { + "epoch": 0.4923331951927062, + "grad_norm": 0.2581813931465149, + "learning_rate": 2.5387484459179447e-06, + "loss": 0.5446, + "step": 5940 + }, + { + "epoch": 0.49241607956900124, + "grad_norm": 0.26382243633270264, + "learning_rate": 2.5383340240364695e-06, + "loss": 0.5117, + "step": 5941 + }, + { + "epoch": 0.4924989639452963, + "grad_norm": 0.2504177391529083, + "learning_rate": 2.537919602154994e-06, + "loss": 0.5208, + "step": 5942 + }, + { + "epoch": 0.49258184832159135, + "grad_norm": 0.2554609775543213, + "learning_rate": 2.5375051802735184e-06, + "loss": 0.5196, + "step": 5943 + }, + { + "epoch": 0.49266473269788646, + "grad_norm": 0.2607099711894989, + "learning_rate": 2.5370907583920432e-06, + "loss": 0.53, + "step": 5944 + }, + { + "epoch": 0.4927476170741815, + "grad_norm": 0.2628728747367859, + "learning_rate": 2.536676336510568e-06, + "loss": 0.535, + "step": 5945 + }, + { + "epoch": 0.4928305014504766, + "grad_norm": 0.2578107714653015, + "learning_rate": 2.536261914629093e-06, + "loss": 0.5079, + "step": 5946 + }, + { + "epoch": 0.49291338582677163, + "grad_norm": 0.283162921667099, + "learning_rate": 2.5358474927476174e-06, + "loss": 0.5231, + "step": 5947 + }, + { + "epoch": 0.49299627020306674, + "grad_norm": 0.2648470401763916, + "learning_rate": 2.535433070866142e-06, + "loss": 0.5273, + "step": 5948 + }, + { + "epoch": 0.4930791545793618, + "grad_norm": 0.2572108805179596, + "learning_rate": 2.5350186489846667e-06, + "loss": 0.555, + "step": 5949 + }, + { + "epoch": 0.49316203895565686, + "grad_norm": 0.2395741194486618, + "learning_rate": 2.534604227103191e-06, + "loss": 0.4636, + "step": 5950 + }, + { + "epoch": 0.4932449233319519, + "grad_norm": 0.25794196128845215, + "learning_rate": 2.5341898052217155e-06, + "loss": 0.5098, + "step": 5951 + }, + { + "epoch": 0.49332780770824697, + "grad_norm": 0.2594955861568451, + "learning_rate": 2.533775383340241e-06, + "loss": 0.5504, + "step": 5952 + }, + { + "epoch": 0.4934106920845421, + "grad_norm": 0.24762989580631256, + "learning_rate": 2.5333609614587652e-06, + "loss": 0.4915, + "step": 5953 + }, + { + "epoch": 0.49349357646083714, + "grad_norm": 0.23073823750019073, + "learning_rate": 2.53294653957729e-06, + "loss": 0.4763, + "step": 5954 + }, + { + "epoch": 0.4935764608371322, + "grad_norm": 0.25652778148651123, + "learning_rate": 2.5325321176958145e-06, + "loss": 0.5138, + "step": 5955 + }, + { + "epoch": 0.49365934521342725, + "grad_norm": 0.24902647733688354, + "learning_rate": 2.532117695814339e-06, + "loss": 0.5101, + "step": 5956 + }, + { + "epoch": 0.49374222958972236, + "grad_norm": 0.24607335031032562, + "learning_rate": 2.531703273932864e-06, + "loss": 0.542, + "step": 5957 + }, + { + "epoch": 0.4938251139660174, + "grad_norm": 0.24535340070724487, + "learning_rate": 2.5312888520513882e-06, + "loss": 0.5812, + "step": 5958 + }, + { + "epoch": 0.49390799834231247, + "grad_norm": 0.2685573399066925, + "learning_rate": 2.5308744301699135e-06, + "loss": 0.5123, + "step": 5959 + }, + { + "epoch": 0.4939908827186075, + "grad_norm": 0.22972063720226288, + "learning_rate": 2.530460008288438e-06, + "loss": 0.4453, + "step": 5960 + }, + { + "epoch": 0.49407376709490264, + "grad_norm": 0.241910919547081, + "learning_rate": 2.5300455864069624e-06, + "loss": 0.4866, + "step": 5961 + }, + { + "epoch": 0.4941566514711977, + "grad_norm": 0.2438254952430725, + "learning_rate": 2.5296311645254872e-06, + "loss": 0.5055, + "step": 5962 + }, + { + "epoch": 0.49423953584749275, + "grad_norm": 0.24424074590206146, + "learning_rate": 2.5292167426440117e-06, + "loss": 0.4961, + "step": 5963 + }, + { + "epoch": 0.4943224202237878, + "grad_norm": 0.22569286823272705, + "learning_rate": 2.5288023207625365e-06, + "loss": 0.4998, + "step": 5964 + }, + { + "epoch": 0.49440530460008286, + "grad_norm": 0.27231070399284363, + "learning_rate": 2.5283878988810614e-06, + "loss": 0.5004, + "step": 5965 + }, + { + "epoch": 0.494488188976378, + "grad_norm": 0.24913284182548523, + "learning_rate": 2.527973476999586e-06, + "loss": 0.4805, + "step": 5966 + }, + { + "epoch": 0.49457107335267303, + "grad_norm": 0.23282620310783386, + "learning_rate": 2.5275590551181107e-06, + "loss": 0.5039, + "step": 5967 + }, + { + "epoch": 0.4946539577289681, + "grad_norm": 0.2286771982908249, + "learning_rate": 2.527144633236635e-06, + "loss": 0.5052, + "step": 5968 + }, + { + "epoch": 0.49473684210526314, + "grad_norm": 0.25059249997138977, + "learning_rate": 2.5267302113551595e-06, + "loss": 0.5193, + "step": 5969 + }, + { + "epoch": 0.49481972648155825, + "grad_norm": 0.26101961731910706, + "learning_rate": 2.5263157894736844e-06, + "loss": 0.5132, + "step": 5970 + }, + { + "epoch": 0.4949026108578533, + "grad_norm": 0.25982069969177246, + "learning_rate": 2.525901367592209e-06, + "loss": 0.5419, + "step": 5971 + }, + { + "epoch": 0.49498549523414836, + "grad_norm": 0.2585727274417877, + "learning_rate": 2.525486945710734e-06, + "loss": 0.5244, + "step": 5972 + }, + { + "epoch": 0.4950683796104434, + "grad_norm": 0.2563199996948242, + "learning_rate": 2.5250725238292585e-06, + "loss": 0.5282, + "step": 5973 + }, + { + "epoch": 0.4951512639867385, + "grad_norm": 0.2615974545478821, + "learning_rate": 2.524658101947783e-06, + "loss": 0.548, + "step": 5974 + }, + { + "epoch": 0.4952341483630336, + "grad_norm": 0.2507331371307373, + "learning_rate": 2.524243680066308e-06, + "loss": 0.4815, + "step": 5975 + }, + { + "epoch": 0.49531703273932864, + "grad_norm": 0.2684566378593445, + "learning_rate": 2.5238292581848322e-06, + "loss": 0.5193, + "step": 5976 + }, + { + "epoch": 0.4953999171156237, + "grad_norm": 0.2564339339733124, + "learning_rate": 2.523414836303357e-06, + "loss": 0.5283, + "step": 5977 + }, + { + "epoch": 0.49548280149191876, + "grad_norm": 0.2500789761543274, + "learning_rate": 2.5230004144218815e-06, + "loss": 0.4718, + "step": 5978 + }, + { + "epoch": 0.49556568586821387, + "grad_norm": 0.26546645164489746, + "learning_rate": 2.522585992540406e-06, + "loss": 0.5374, + "step": 5979 + }, + { + "epoch": 0.4956485702445089, + "grad_norm": 0.2304716259241104, + "learning_rate": 2.5221715706589312e-06, + "loss": 0.4947, + "step": 5980 + }, + { + "epoch": 0.495731454620804, + "grad_norm": 0.2614201307296753, + "learning_rate": 2.5217571487774557e-06, + "loss": 0.5631, + "step": 5981 + }, + { + "epoch": 0.49581433899709904, + "grad_norm": 0.2566671669483185, + "learning_rate": 2.5213427268959805e-06, + "loss": 0.5176, + "step": 5982 + }, + { + "epoch": 0.4958972233733941, + "grad_norm": 0.2495611608028412, + "learning_rate": 2.520928305014505e-06, + "loss": 0.5146, + "step": 5983 + }, + { + "epoch": 0.4959801077496892, + "grad_norm": 0.23607859015464783, + "learning_rate": 2.5205138831330294e-06, + "loss": 0.4618, + "step": 5984 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 0.24934765696525574, + "learning_rate": 2.5200994612515547e-06, + "loss": 0.493, + "step": 5985 + }, + { + "epoch": 0.4961458765022793, + "grad_norm": 0.2801060974597931, + "learning_rate": 2.519685039370079e-06, + "loss": 0.5508, + "step": 5986 + }, + { + "epoch": 0.49622876087857437, + "grad_norm": 0.25918081402778625, + "learning_rate": 2.5192706174886035e-06, + "loss": 0.5171, + "step": 5987 + }, + { + "epoch": 0.4963116452548695, + "grad_norm": 0.2453443855047226, + "learning_rate": 2.5188561956071284e-06, + "loss": 0.5128, + "step": 5988 + }, + { + "epoch": 0.49639452963116454, + "grad_norm": 0.24668391048908234, + "learning_rate": 2.518441773725653e-06, + "loss": 0.504, + "step": 5989 + }, + { + "epoch": 0.4964774140074596, + "grad_norm": 0.24252478778362274, + "learning_rate": 2.5180273518441777e-06, + "loss": 0.503, + "step": 5990 + }, + { + "epoch": 0.49656029838375465, + "grad_norm": 0.2582523226737976, + "learning_rate": 2.517612929962702e-06, + "loss": 0.5236, + "step": 5991 + }, + { + "epoch": 0.4966431827600497, + "grad_norm": 0.2585163116455078, + "learning_rate": 2.5171985080812265e-06, + "loss": 0.535, + "step": 5992 + }, + { + "epoch": 0.4967260671363448, + "grad_norm": 0.2583881914615631, + "learning_rate": 2.516784086199752e-06, + "loss": 0.5095, + "step": 5993 + }, + { + "epoch": 0.4968089515126399, + "grad_norm": 0.25251564383506775, + "learning_rate": 2.5163696643182763e-06, + "loss": 0.4861, + "step": 5994 + }, + { + "epoch": 0.49689183588893493, + "grad_norm": 0.23722219467163086, + "learning_rate": 2.515955242436801e-06, + "loss": 0.4747, + "step": 5995 + }, + { + "epoch": 0.49697472026523, + "grad_norm": 0.23447595536708832, + "learning_rate": 2.5155408205553255e-06, + "loss": 0.4522, + "step": 5996 + }, + { + "epoch": 0.4970576046415251, + "grad_norm": 0.47907763719558716, + "learning_rate": 2.51512639867385e-06, + "loss": 0.5374, + "step": 5997 + }, + { + "epoch": 0.49714048901782015, + "grad_norm": 0.25361165404319763, + "learning_rate": 2.514711976792375e-06, + "loss": 0.5161, + "step": 5998 + }, + { + "epoch": 0.4972233733941152, + "grad_norm": 0.24865508079528809, + "learning_rate": 2.5142975549108993e-06, + "loss": 0.4615, + "step": 5999 + }, + { + "epoch": 0.49730625777041026, + "grad_norm": 0.24020589888095856, + "learning_rate": 2.5138831330294237e-06, + "loss": 0.4976, + "step": 6000 + }, + { + "epoch": 0.4973891421467053, + "grad_norm": 0.26028764247894287, + "learning_rate": 2.513468711147949e-06, + "loss": 0.5237, + "step": 6001 + }, + { + "epoch": 0.49747202652300043, + "grad_norm": 0.2504473030567169, + "learning_rate": 2.5130542892664734e-06, + "loss": 0.5445, + "step": 6002 + }, + { + "epoch": 0.4975549108992955, + "grad_norm": 0.24820564687252045, + "learning_rate": 2.5126398673849983e-06, + "loss": 0.509, + "step": 6003 + }, + { + "epoch": 0.49763779527559054, + "grad_norm": 0.25453662872314453, + "learning_rate": 2.5122254455035227e-06, + "loss": 0.5388, + "step": 6004 + }, + { + "epoch": 0.4977206796518856, + "grad_norm": 0.2521790862083435, + "learning_rate": 2.511811023622047e-06, + "loss": 0.4967, + "step": 6005 + }, + { + "epoch": 0.4978035640281807, + "grad_norm": 0.25307977199554443, + "learning_rate": 2.5113966017405724e-06, + "loss": 0.4906, + "step": 6006 + }, + { + "epoch": 0.49788644840447577, + "grad_norm": 0.2673411965370178, + "learning_rate": 2.510982179859097e-06, + "loss": 0.5194, + "step": 6007 + }, + { + "epoch": 0.4979693327807708, + "grad_norm": 0.2479352056980133, + "learning_rate": 2.5105677579776217e-06, + "loss": 0.5587, + "step": 6008 + }, + { + "epoch": 0.4980522171570659, + "grad_norm": 0.2661097049713135, + "learning_rate": 2.510153336096146e-06, + "loss": 0.5122, + "step": 6009 + }, + { + "epoch": 0.49813510153336094, + "grad_norm": 0.25009873509407043, + "learning_rate": 2.5097389142146706e-06, + "loss": 0.5367, + "step": 6010 + }, + { + "epoch": 0.49821798590965605, + "grad_norm": 0.25982677936553955, + "learning_rate": 2.5093244923331954e-06, + "loss": 0.4925, + "step": 6011 + }, + { + "epoch": 0.4983008702859511, + "grad_norm": 0.26191386580467224, + "learning_rate": 2.50891007045172e-06, + "loss": 0.5098, + "step": 6012 + }, + { + "epoch": 0.49838375466224616, + "grad_norm": 0.24398814141750336, + "learning_rate": 2.508495648570245e-06, + "loss": 0.5085, + "step": 6013 + }, + { + "epoch": 0.4984666390385412, + "grad_norm": 0.24752146005630493, + "learning_rate": 2.5080812266887695e-06, + "loss": 0.5186, + "step": 6014 + }, + { + "epoch": 0.4985495234148363, + "grad_norm": 0.2464447021484375, + "learning_rate": 2.507666804807294e-06, + "loss": 0.5417, + "step": 6015 + }, + { + "epoch": 0.4986324077911314, + "grad_norm": 0.26439228653907776, + "learning_rate": 2.507252382925819e-06, + "loss": 0.5562, + "step": 6016 + }, + { + "epoch": 0.49871529216742644, + "grad_norm": 0.2429608404636383, + "learning_rate": 2.5068379610443433e-06, + "loss": 0.4608, + "step": 6017 + }, + { + "epoch": 0.4987981765437215, + "grad_norm": 0.274142324924469, + "learning_rate": 2.5064235391628677e-06, + "loss": 0.4733, + "step": 6018 + }, + { + "epoch": 0.49888106092001655, + "grad_norm": 0.24354715645313263, + "learning_rate": 2.5060091172813926e-06, + "loss": 0.5536, + "step": 6019 + }, + { + "epoch": 0.49896394529631166, + "grad_norm": 0.24443238973617554, + "learning_rate": 2.505594695399917e-06, + "loss": 0.5071, + "step": 6020 + }, + { + "epoch": 0.4990468296726067, + "grad_norm": 0.24575810134410858, + "learning_rate": 2.5051802735184423e-06, + "loss": 0.5068, + "step": 6021 + }, + { + "epoch": 0.4991297140489018, + "grad_norm": 0.2561238706111908, + "learning_rate": 2.5047658516369667e-06, + "loss": 0.5226, + "step": 6022 + }, + { + "epoch": 0.49921259842519683, + "grad_norm": 0.24153490364551544, + "learning_rate": 2.504351429755491e-06, + "loss": 0.4978, + "step": 6023 + }, + { + "epoch": 0.49929548280149194, + "grad_norm": 0.25205546617507935, + "learning_rate": 2.503937007874016e-06, + "loss": 0.5167, + "step": 6024 + }, + { + "epoch": 0.499378367177787, + "grad_norm": 0.24924877285957336, + "learning_rate": 2.5035225859925404e-06, + "loss": 0.5181, + "step": 6025 + }, + { + "epoch": 0.49946125155408205, + "grad_norm": 0.23615002632141113, + "learning_rate": 2.5031081641110657e-06, + "loss": 0.4856, + "step": 6026 + }, + { + "epoch": 0.4995441359303771, + "grad_norm": 0.2507381737232208, + "learning_rate": 2.50269374222959e-06, + "loss": 0.498, + "step": 6027 + }, + { + "epoch": 0.49962702030667216, + "grad_norm": 0.26874956488609314, + "learning_rate": 2.5022793203481146e-06, + "loss": 0.5309, + "step": 6028 + }, + { + "epoch": 0.4997099046829673, + "grad_norm": 0.23894082009792328, + "learning_rate": 2.5018648984666394e-06, + "loss": 0.5016, + "step": 6029 + }, + { + "epoch": 0.49979278905926233, + "grad_norm": 0.2760389745235443, + "learning_rate": 2.501450476585164e-06, + "loss": 0.5016, + "step": 6030 + }, + { + "epoch": 0.4998756734355574, + "grad_norm": 0.2781631052494049, + "learning_rate": 2.5010360547036887e-06, + "loss": 0.519, + "step": 6031 + }, + { + "epoch": 0.49995855781185244, + "grad_norm": 0.2497565895318985, + "learning_rate": 2.500621632822213e-06, + "loss": 0.5018, + "step": 6032 + }, + { + "epoch": 0.5000414421881475, + "grad_norm": 0.2598586082458496, + "learning_rate": 2.5002072109407376e-06, + "loss": 0.5549, + "step": 6033 + }, + { + "epoch": 0.5001243265644426, + "grad_norm": 0.2558212876319885, + "learning_rate": 2.4997927890592624e-06, + "loss": 0.5074, + "step": 6034 + }, + { + "epoch": 0.5002072109407377, + "grad_norm": 0.2786058187484741, + "learning_rate": 2.4993783671777873e-06, + "loss": 0.5475, + "step": 6035 + }, + { + "epoch": 0.5002900953170327, + "grad_norm": 0.2557070255279541, + "learning_rate": 2.4989639452963117e-06, + "loss": 0.5011, + "step": 6036 + }, + { + "epoch": 0.5003729796933278, + "grad_norm": 0.2785012125968933, + "learning_rate": 2.4985495234148366e-06, + "loss": 0.5021, + "step": 6037 + }, + { + "epoch": 0.5004558640696228, + "grad_norm": 0.24984565377235413, + "learning_rate": 2.498135101533361e-06, + "loss": 0.4901, + "step": 6038 + }, + { + "epoch": 0.500538748445918, + "grad_norm": 0.2492893636226654, + "learning_rate": 2.497720679651886e-06, + "loss": 0.478, + "step": 6039 + }, + { + "epoch": 0.5006216328222131, + "grad_norm": 0.2631351351737976, + "learning_rate": 2.4973062577704103e-06, + "loss": 0.5248, + "step": 6040 + }, + { + "epoch": 0.5007045171985081, + "grad_norm": 0.24937421083450317, + "learning_rate": 2.496891835888935e-06, + "loss": 0.4725, + "step": 6041 + }, + { + "epoch": 0.5007874015748032, + "grad_norm": 0.27234452962875366, + "learning_rate": 2.49647741400746e-06, + "loss": 0.594, + "step": 6042 + }, + { + "epoch": 0.5008702859510982, + "grad_norm": 0.26605913043022156, + "learning_rate": 2.4960629921259844e-06, + "loss": 0.5076, + "step": 6043 + }, + { + "epoch": 0.5009531703273933, + "grad_norm": 0.2851162850856781, + "learning_rate": 2.4956485702445093e-06, + "loss": 0.5803, + "step": 6044 + }, + { + "epoch": 0.5010360547036884, + "grad_norm": 0.24649663269519806, + "learning_rate": 2.4952341483630337e-06, + "loss": 0.487, + "step": 6045 + }, + { + "epoch": 0.5011189390799834, + "grad_norm": 0.26688897609710693, + "learning_rate": 2.4948197264815586e-06, + "loss": 0.5248, + "step": 6046 + }, + { + "epoch": 0.5012018234562785, + "grad_norm": 0.2824138104915619, + "learning_rate": 2.494405304600083e-06, + "loss": 0.5063, + "step": 6047 + }, + { + "epoch": 0.5012847078325735, + "grad_norm": 0.2677099406719208, + "learning_rate": 2.493990882718608e-06, + "loss": 0.5789, + "step": 6048 + }, + { + "epoch": 0.5013675922088686, + "grad_norm": 0.2545158863067627, + "learning_rate": 2.4935764608371323e-06, + "loss": 0.5214, + "step": 6049 + }, + { + "epoch": 0.5014504765851637, + "grad_norm": 0.257363885641098, + "learning_rate": 2.493162038955657e-06, + "loss": 0.5238, + "step": 6050 + }, + { + "epoch": 0.5015333609614587, + "grad_norm": 0.2378970980644226, + "learning_rate": 2.492747617074182e-06, + "loss": 0.4854, + "step": 6051 + }, + { + "epoch": 0.5016162453377538, + "grad_norm": 0.25181227922439575, + "learning_rate": 2.4923331951927064e-06, + "loss": 0.5114, + "step": 6052 + }, + { + "epoch": 0.501699129714049, + "grad_norm": 0.26207756996154785, + "learning_rate": 2.491918773311231e-06, + "loss": 0.5632, + "step": 6053 + }, + { + "epoch": 0.501782014090344, + "grad_norm": 0.25115668773651123, + "learning_rate": 2.4915043514297557e-06, + "loss": 0.5039, + "step": 6054 + }, + { + "epoch": 0.5018648984666391, + "grad_norm": 0.24857164919376373, + "learning_rate": 2.4910899295482806e-06, + "loss": 0.5177, + "step": 6055 + }, + { + "epoch": 0.5019477828429341, + "grad_norm": 0.25395679473876953, + "learning_rate": 2.490675507666805e-06, + "loss": 0.5249, + "step": 6056 + }, + { + "epoch": 0.5020306672192292, + "grad_norm": 0.2563682496547699, + "learning_rate": 2.4902610857853294e-06, + "loss": 0.5178, + "step": 6057 + }, + { + "epoch": 0.5021135515955243, + "grad_norm": 0.3070979714393616, + "learning_rate": 2.4898466639038543e-06, + "loss": 0.5519, + "step": 6058 + }, + { + "epoch": 0.5021964359718193, + "grad_norm": 0.28455525636672974, + "learning_rate": 2.489432242022379e-06, + "loss": 0.535, + "step": 6059 + }, + { + "epoch": 0.5022793203481144, + "grad_norm": 0.2569061815738678, + "learning_rate": 2.4890178201409036e-06, + "loss": 0.5101, + "step": 6060 + }, + { + "epoch": 0.5023622047244094, + "grad_norm": 0.26312756538391113, + "learning_rate": 2.488603398259428e-06, + "loss": 0.5345, + "step": 6061 + }, + { + "epoch": 0.5024450891007045, + "grad_norm": 0.2519277036190033, + "learning_rate": 2.488188976377953e-06, + "loss": 0.5167, + "step": 6062 + }, + { + "epoch": 0.5025279734769996, + "grad_norm": 0.24182425439357758, + "learning_rate": 2.4877745544964777e-06, + "loss": 0.525, + "step": 6063 + }, + { + "epoch": 0.5026108578532946, + "grad_norm": 0.23368947207927704, + "learning_rate": 2.4873601326150026e-06, + "loss": 0.4865, + "step": 6064 + }, + { + "epoch": 0.5026937422295897, + "grad_norm": 0.26484522223472595, + "learning_rate": 2.486945710733527e-06, + "loss": 0.5508, + "step": 6065 + }, + { + "epoch": 0.5027766266058847, + "grad_norm": 0.23831899464130402, + "learning_rate": 2.4865312888520514e-06, + "loss": 0.5035, + "step": 6066 + }, + { + "epoch": 0.5028595109821798, + "grad_norm": 0.22997315227985382, + "learning_rate": 2.4861168669705763e-06, + "loss": 0.5084, + "step": 6067 + }, + { + "epoch": 0.502942395358475, + "grad_norm": 0.24539481103420258, + "learning_rate": 2.485702445089101e-06, + "loss": 0.5093, + "step": 6068 + }, + { + "epoch": 0.50302527973477, + "grad_norm": 0.2465042918920517, + "learning_rate": 2.4852880232076256e-06, + "loss": 0.4837, + "step": 6069 + }, + { + "epoch": 0.5031081641110651, + "grad_norm": 0.26290950179100037, + "learning_rate": 2.48487360132615e-06, + "loss": 0.5069, + "step": 6070 + }, + { + "epoch": 0.5031910484873602, + "grad_norm": 0.2374269813299179, + "learning_rate": 2.484459179444675e-06, + "loss": 0.5112, + "step": 6071 + }, + { + "epoch": 0.5032739328636552, + "grad_norm": 0.25855231285095215, + "learning_rate": 2.4840447575631997e-06, + "loss": 0.4872, + "step": 6072 + }, + { + "epoch": 0.5033568172399503, + "grad_norm": 0.2618601322174072, + "learning_rate": 2.483630335681724e-06, + "loss": 0.5241, + "step": 6073 + }, + { + "epoch": 0.5034397016162453, + "grad_norm": 0.24248115718364716, + "learning_rate": 2.4832159138002486e-06, + "loss": 0.5067, + "step": 6074 + }, + { + "epoch": 0.5035225859925404, + "grad_norm": 0.23962156474590302, + "learning_rate": 2.4828014919187734e-06, + "loss": 0.5149, + "step": 6075 + }, + { + "epoch": 0.5036054703688355, + "grad_norm": 0.24508622288703918, + "learning_rate": 2.4823870700372983e-06, + "loss": 0.4898, + "step": 6076 + }, + { + "epoch": 0.5036883547451305, + "grad_norm": 0.2583557665348053, + "learning_rate": 2.4819726481558227e-06, + "loss": 0.5272, + "step": 6077 + }, + { + "epoch": 0.5037712391214256, + "grad_norm": 0.2646426260471344, + "learning_rate": 2.4815582262743476e-06, + "loss": 0.5035, + "step": 6078 + }, + { + "epoch": 0.5038541234977206, + "grad_norm": 0.23374664783477783, + "learning_rate": 2.481143804392872e-06, + "loss": 0.4994, + "step": 6079 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 0.2606086730957031, + "learning_rate": 2.480729382511397e-06, + "loss": 0.561, + "step": 6080 + }, + { + "epoch": 0.5040198922503109, + "grad_norm": 0.23678122460842133, + "learning_rate": 2.4803149606299213e-06, + "loss": 0.4879, + "step": 6081 + }, + { + "epoch": 0.5041027766266059, + "grad_norm": 0.30286917090415955, + "learning_rate": 2.479900538748446e-06, + "loss": 0.5665, + "step": 6082 + }, + { + "epoch": 0.504185661002901, + "grad_norm": 0.24408362805843353, + "learning_rate": 2.4794861168669706e-06, + "loss": 0.4983, + "step": 6083 + }, + { + "epoch": 0.5042685453791961, + "grad_norm": 0.2462170124053955, + "learning_rate": 2.4790716949854954e-06, + "loss": 0.5316, + "step": 6084 + }, + { + "epoch": 0.5043514297554911, + "grad_norm": 0.24677705764770508, + "learning_rate": 2.4786572731040203e-06, + "loss": 0.4933, + "step": 6085 + }, + { + "epoch": 0.5044343141317862, + "grad_norm": 0.27014026045799255, + "learning_rate": 2.4782428512225447e-06, + "loss": 0.5169, + "step": 6086 + }, + { + "epoch": 0.5045171985080812, + "grad_norm": 0.2555750906467438, + "learning_rate": 2.4778284293410696e-06, + "loss": 0.5435, + "step": 6087 + }, + { + "epoch": 0.5046000828843763, + "grad_norm": 0.24175994098186493, + "learning_rate": 2.477414007459594e-06, + "loss": 0.5373, + "step": 6088 + }, + { + "epoch": 0.5046829672606714, + "grad_norm": 0.3128836452960968, + "learning_rate": 2.476999585578119e-06, + "loss": 0.5123, + "step": 6089 + }, + { + "epoch": 0.5047658516369664, + "grad_norm": 0.2542758882045746, + "learning_rate": 2.4765851636966433e-06, + "loss": 0.495, + "step": 6090 + }, + { + "epoch": 0.5048487360132615, + "grad_norm": 0.2612980604171753, + "learning_rate": 2.476170741815168e-06, + "loss": 0.5408, + "step": 6091 + }, + { + "epoch": 0.5049316203895565, + "grad_norm": 0.2391999065876007, + "learning_rate": 2.4757563199336926e-06, + "loss": 0.4886, + "step": 6092 + }, + { + "epoch": 0.5050145047658516, + "grad_norm": 0.25136858224868774, + "learning_rate": 2.4753418980522174e-06, + "loss": 0.4808, + "step": 6093 + }, + { + "epoch": 0.5050973891421467, + "grad_norm": 0.24799901247024536, + "learning_rate": 2.474927476170742e-06, + "loss": 0.483, + "step": 6094 + }, + { + "epoch": 0.5051802735184417, + "grad_norm": 0.24457229673862457, + "learning_rate": 2.4745130542892667e-06, + "loss": 0.5008, + "step": 6095 + }, + { + "epoch": 0.5052631578947369, + "grad_norm": 0.25707194209098816, + "learning_rate": 2.4740986324077916e-06, + "loss": 0.5136, + "step": 6096 + }, + { + "epoch": 0.5053460422710319, + "grad_norm": 0.268247127532959, + "learning_rate": 2.473684210526316e-06, + "loss": 0.5516, + "step": 6097 + }, + { + "epoch": 0.505428926647327, + "grad_norm": 0.24434413015842438, + "learning_rate": 2.4732697886448405e-06, + "loss": 0.4784, + "step": 6098 + }, + { + "epoch": 0.5055118110236221, + "grad_norm": 0.26397132873535156, + "learning_rate": 2.4728553667633653e-06, + "loss": 0.5326, + "step": 6099 + }, + { + "epoch": 0.5055946953999171, + "grad_norm": 0.23967309296131134, + "learning_rate": 2.47244094488189e-06, + "loss": 0.5051, + "step": 6100 + }, + { + "epoch": 0.5056775797762122, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.4720265230004146e-06, + "loss": 0.5096, + "step": 6101 + }, + { + "epoch": 0.5057604641525073, + "grad_norm": 0.2513044774532318, + "learning_rate": 2.471612101118939e-06, + "loss": 0.5147, + "step": 6102 + }, + { + "epoch": 0.5058433485288023, + "grad_norm": 0.25753656029701233, + "learning_rate": 2.471197679237464e-06, + "loss": 0.5478, + "step": 6103 + }, + { + "epoch": 0.5059262329050974, + "grad_norm": 0.2589559555053711, + "learning_rate": 2.4707832573559887e-06, + "loss": 0.5928, + "step": 6104 + }, + { + "epoch": 0.5060091172813924, + "grad_norm": 0.24775353074073792, + "learning_rate": 2.4703688354745136e-06, + "loss": 0.516, + "step": 6105 + }, + { + "epoch": 0.5060920016576875, + "grad_norm": 0.26786506175994873, + "learning_rate": 2.469954413593038e-06, + "loss": 0.514, + "step": 6106 + }, + { + "epoch": 0.5061748860339826, + "grad_norm": 0.24304378032684326, + "learning_rate": 2.4695399917115625e-06, + "loss": 0.5048, + "step": 6107 + }, + { + "epoch": 0.5062577704102776, + "grad_norm": 0.23674997687339783, + "learning_rate": 2.4691255698300873e-06, + "loss": 0.5275, + "step": 6108 + }, + { + "epoch": 0.5063406547865728, + "grad_norm": 0.2748830020427704, + "learning_rate": 2.468711147948612e-06, + "loss": 0.5538, + "step": 6109 + }, + { + "epoch": 0.5064235391628678, + "grad_norm": 0.26856812834739685, + "learning_rate": 2.4682967260671366e-06, + "loss": 0.501, + "step": 6110 + }, + { + "epoch": 0.5065064235391629, + "grad_norm": 0.23661142587661743, + "learning_rate": 2.467882304185661e-06, + "loss": 0.4764, + "step": 6111 + }, + { + "epoch": 0.506589307915458, + "grad_norm": 0.24911783635616302, + "learning_rate": 2.467467882304186e-06, + "loss": 0.5145, + "step": 6112 + }, + { + "epoch": 0.506672192291753, + "grad_norm": 0.2274685949087143, + "learning_rate": 2.4670534604227107e-06, + "loss": 0.4897, + "step": 6113 + }, + { + "epoch": 0.5067550766680481, + "grad_norm": 0.2606127858161926, + "learning_rate": 2.466639038541235e-06, + "loss": 0.478, + "step": 6114 + }, + { + "epoch": 0.5068379610443431, + "grad_norm": 0.2518637776374817, + "learning_rate": 2.4662246166597596e-06, + "loss": 0.5427, + "step": 6115 + }, + { + "epoch": 0.5069208454206382, + "grad_norm": 0.22962473332881927, + "learning_rate": 2.4658101947782845e-06, + "loss": 0.5057, + "step": 6116 + }, + { + "epoch": 0.5070037297969333, + "grad_norm": 0.2556267976760864, + "learning_rate": 2.4653957728968093e-06, + "loss": 0.5072, + "step": 6117 + }, + { + "epoch": 0.5070866141732283, + "grad_norm": 0.2820703685283661, + "learning_rate": 2.4649813510153337e-06, + "loss": 0.5359, + "step": 6118 + }, + { + "epoch": 0.5071694985495234, + "grad_norm": 0.255122572183609, + "learning_rate": 2.464566929133858e-06, + "loss": 0.5399, + "step": 6119 + }, + { + "epoch": 0.5072523829258185, + "grad_norm": 0.24921561777591705, + "learning_rate": 2.464152507252383e-06, + "loss": 0.5405, + "step": 6120 + }, + { + "epoch": 0.5073352673021135, + "grad_norm": 0.23815420269966125, + "learning_rate": 2.463738085370908e-06, + "loss": 0.4673, + "step": 6121 + }, + { + "epoch": 0.5074181516784086, + "grad_norm": 0.2431754171848297, + "learning_rate": 2.4633236634894323e-06, + "loss": 0.5244, + "step": 6122 + }, + { + "epoch": 0.5075010360547036, + "grad_norm": 0.27018365263938904, + "learning_rate": 2.462909241607957e-06, + "loss": 0.5732, + "step": 6123 + }, + { + "epoch": 0.5075839204309988, + "grad_norm": 0.2621809244155884, + "learning_rate": 2.4624948197264816e-06, + "loss": 0.5466, + "step": 6124 + }, + { + "epoch": 0.5076668048072939, + "grad_norm": 0.26592785120010376, + "learning_rate": 2.4620803978450065e-06, + "loss": 0.5438, + "step": 6125 + }, + { + "epoch": 0.5077496891835889, + "grad_norm": 0.2413918673992157, + "learning_rate": 2.4616659759635313e-06, + "loss": 0.5029, + "step": 6126 + }, + { + "epoch": 0.507832573559884, + "grad_norm": 0.24541451036930084, + "learning_rate": 2.4612515540820558e-06, + "loss": 0.5079, + "step": 6127 + }, + { + "epoch": 0.507915457936179, + "grad_norm": 0.25483986735343933, + "learning_rate": 2.46083713220058e-06, + "loss": 0.5183, + "step": 6128 + }, + { + "epoch": 0.5079983423124741, + "grad_norm": 0.2764640748500824, + "learning_rate": 2.460422710319105e-06, + "loss": 0.5258, + "step": 6129 + }, + { + "epoch": 0.5080812266887692, + "grad_norm": 0.23899559676647186, + "learning_rate": 2.46000828843763e-06, + "loss": 0.4693, + "step": 6130 + }, + { + "epoch": 0.5081641110650642, + "grad_norm": 0.23739323019981384, + "learning_rate": 2.4595938665561543e-06, + "loss": 0.5038, + "step": 6131 + }, + { + "epoch": 0.5082469954413593, + "grad_norm": 0.2460920363664627, + "learning_rate": 2.4591794446746788e-06, + "loss": 0.5124, + "step": 6132 + }, + { + "epoch": 0.5083298798176543, + "grad_norm": 0.23726104199886322, + "learning_rate": 2.4587650227932036e-06, + "loss": 0.4815, + "step": 6133 + }, + { + "epoch": 0.5084127641939494, + "grad_norm": 0.2658695578575134, + "learning_rate": 2.4583506009117285e-06, + "loss": 0.5164, + "step": 6134 + }, + { + "epoch": 0.5084956485702445, + "grad_norm": 0.2495201677083969, + "learning_rate": 2.457936179030253e-06, + "loss": 0.5114, + "step": 6135 + }, + { + "epoch": 0.5085785329465395, + "grad_norm": 0.2647700011730194, + "learning_rate": 2.4575217571487778e-06, + "loss": 0.5061, + "step": 6136 + }, + { + "epoch": 0.5086614173228347, + "grad_norm": 0.23019640147686005, + "learning_rate": 2.457107335267302e-06, + "loss": 0.5084, + "step": 6137 + }, + { + "epoch": 0.5087443016991298, + "grad_norm": 0.25772416591644287, + "learning_rate": 2.456692913385827e-06, + "loss": 0.516, + "step": 6138 + }, + { + "epoch": 0.5088271860754248, + "grad_norm": 0.2460571527481079, + "learning_rate": 2.4562784915043515e-06, + "loss": 0.5225, + "step": 6139 + }, + { + "epoch": 0.5089100704517199, + "grad_norm": 0.25247812271118164, + "learning_rate": 2.4558640696228763e-06, + "loss": 0.5098, + "step": 6140 + }, + { + "epoch": 0.5089929548280149, + "grad_norm": 0.24691146612167358, + "learning_rate": 2.4554496477414008e-06, + "loss": 0.4866, + "step": 6141 + }, + { + "epoch": 0.50907583920431, + "grad_norm": 0.24729841947555542, + "learning_rate": 2.4550352258599256e-06, + "loss": 0.5175, + "step": 6142 + }, + { + "epoch": 0.5091587235806051, + "grad_norm": 0.24483293294906616, + "learning_rate": 2.4546208039784505e-06, + "loss": 0.5056, + "step": 6143 + }, + { + "epoch": 0.5092416079569001, + "grad_norm": 0.26354485750198364, + "learning_rate": 2.454206382096975e-06, + "loss": 0.5338, + "step": 6144 + }, + { + "epoch": 0.5093244923331952, + "grad_norm": 0.2541813254356384, + "learning_rate": 2.4537919602154998e-06, + "loss": 0.5485, + "step": 6145 + }, + { + "epoch": 0.5094073767094902, + "grad_norm": 0.24662117660045624, + "learning_rate": 2.453377538334024e-06, + "loss": 0.4882, + "step": 6146 + }, + { + "epoch": 0.5094902610857853, + "grad_norm": 0.28011354804039, + "learning_rate": 2.452963116452549e-06, + "loss": 0.5295, + "step": 6147 + }, + { + "epoch": 0.5095731454620804, + "grad_norm": 0.3212886154651642, + "learning_rate": 2.4525486945710735e-06, + "loss": 0.5117, + "step": 6148 + }, + { + "epoch": 0.5096560298383754, + "grad_norm": 0.24747565388679504, + "learning_rate": 2.4521342726895983e-06, + "loss": 0.4934, + "step": 6149 + }, + { + "epoch": 0.5097389142146705, + "grad_norm": 0.2527866065502167, + "learning_rate": 2.4517198508081228e-06, + "loss": 0.5236, + "step": 6150 + }, + { + "epoch": 0.5098217985909657, + "grad_norm": 0.2732298672199249, + "learning_rate": 2.4513054289266476e-06, + "loss": 0.5327, + "step": 6151 + }, + { + "epoch": 0.5099046829672607, + "grad_norm": 0.25683701038360596, + "learning_rate": 2.450891007045172e-06, + "loss": 0.4951, + "step": 6152 + }, + { + "epoch": 0.5099875673435558, + "grad_norm": 0.26006919145584106, + "learning_rate": 2.450476585163697e-06, + "loss": 0.52, + "step": 6153 + }, + { + "epoch": 0.5100704517198508, + "grad_norm": 0.22417645156383514, + "learning_rate": 2.4500621632822218e-06, + "loss": 0.4581, + "step": 6154 + }, + { + "epoch": 0.5101533360961459, + "grad_norm": 0.2524721622467041, + "learning_rate": 2.449647741400746e-06, + "loss": 0.4967, + "step": 6155 + }, + { + "epoch": 0.510236220472441, + "grad_norm": 0.2561906576156616, + "learning_rate": 2.4492333195192706e-06, + "loss": 0.5265, + "step": 6156 + }, + { + "epoch": 0.510319104848736, + "grad_norm": 0.23837152123451233, + "learning_rate": 2.4488188976377955e-06, + "loss": 0.4583, + "step": 6157 + }, + { + "epoch": 0.5104019892250311, + "grad_norm": 0.2673163115978241, + "learning_rate": 2.4484044757563203e-06, + "loss": 0.567, + "step": 6158 + }, + { + "epoch": 0.5104848736013261, + "grad_norm": 0.2512507140636444, + "learning_rate": 2.4479900538748448e-06, + "loss": 0.5545, + "step": 6159 + }, + { + "epoch": 0.5105677579776212, + "grad_norm": 0.24111244082450867, + "learning_rate": 2.447575631993369e-06, + "loss": 0.4999, + "step": 6160 + }, + { + "epoch": 0.5106506423539163, + "grad_norm": 0.2562327980995178, + "learning_rate": 2.447161210111894e-06, + "loss": 0.5026, + "step": 6161 + }, + { + "epoch": 0.5107335267302113, + "grad_norm": 0.26181817054748535, + "learning_rate": 2.446746788230419e-06, + "loss": 0.549, + "step": 6162 + }, + { + "epoch": 0.5108164111065064, + "grad_norm": 0.238749697804451, + "learning_rate": 2.4463323663489433e-06, + "loss": 0.524, + "step": 6163 + }, + { + "epoch": 0.5108992954828014, + "grad_norm": 0.26137420535087585, + "learning_rate": 2.445917944467468e-06, + "loss": 0.5441, + "step": 6164 + }, + { + "epoch": 0.5109821798590966, + "grad_norm": 0.2577533423900604, + "learning_rate": 2.4455035225859926e-06, + "loss": 0.5116, + "step": 6165 + }, + { + "epoch": 0.5110650642353917, + "grad_norm": 0.24093413352966309, + "learning_rate": 2.4450891007045175e-06, + "loss": 0.5024, + "step": 6166 + }, + { + "epoch": 0.5111479486116867, + "grad_norm": 0.2610907256603241, + "learning_rate": 2.4446746788230423e-06, + "loss": 0.5043, + "step": 6167 + }, + { + "epoch": 0.5112308329879818, + "grad_norm": 0.24128367006778717, + "learning_rate": 2.4442602569415668e-06, + "loss": 0.5095, + "step": 6168 + }, + { + "epoch": 0.5113137173642769, + "grad_norm": 0.2497491091489792, + "learning_rate": 2.443845835060091e-06, + "loss": 0.5207, + "step": 6169 + }, + { + "epoch": 0.5113966017405719, + "grad_norm": 0.2503413259983063, + "learning_rate": 2.443431413178616e-06, + "loss": 0.5174, + "step": 6170 + }, + { + "epoch": 0.511479486116867, + "grad_norm": 0.24550452828407288, + "learning_rate": 2.443016991297141e-06, + "loss": 0.495, + "step": 6171 + }, + { + "epoch": 0.511562370493162, + "grad_norm": 0.24685175716876984, + "learning_rate": 2.4426025694156653e-06, + "loss": 0.548, + "step": 6172 + }, + { + "epoch": 0.5116452548694571, + "grad_norm": 0.2452838122844696, + "learning_rate": 2.4421881475341898e-06, + "loss": 0.4643, + "step": 6173 + }, + { + "epoch": 0.5117281392457522, + "grad_norm": 0.24791289865970612, + "learning_rate": 2.4417737256527146e-06, + "loss": 0.5188, + "step": 6174 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 0.2515656650066376, + "learning_rate": 2.4413593037712395e-06, + "loss": 0.5244, + "step": 6175 + }, + { + "epoch": 0.5118939079983423, + "grad_norm": 0.25699859857559204, + "learning_rate": 2.440944881889764e-06, + "loss": 0.5118, + "step": 6176 + }, + { + "epoch": 0.5119767923746373, + "grad_norm": 0.23708128929138184, + "learning_rate": 2.4405304600082884e-06, + "loss": 0.4923, + "step": 6177 + }, + { + "epoch": 0.5120596767509324, + "grad_norm": 0.25501978397369385, + "learning_rate": 2.440116038126813e-06, + "loss": 0.5006, + "step": 6178 + }, + { + "epoch": 0.5121425611272276, + "grad_norm": 0.2503023147583008, + "learning_rate": 2.439701616245338e-06, + "loss": 0.4503, + "step": 6179 + }, + { + "epoch": 0.5122254455035226, + "grad_norm": 0.25470060110092163, + "learning_rate": 2.4392871943638625e-06, + "loss": 0.5264, + "step": 6180 + }, + { + "epoch": 0.5123083298798177, + "grad_norm": 0.2372523993253708, + "learning_rate": 2.4388727724823874e-06, + "loss": 0.4732, + "step": 6181 + }, + { + "epoch": 0.5123912142561127, + "grad_norm": 0.2827899158000946, + "learning_rate": 2.4384583506009118e-06, + "loss": 0.5482, + "step": 6182 + }, + { + "epoch": 0.5124740986324078, + "grad_norm": 0.2450476884841919, + "learning_rate": 2.4380439287194366e-06, + "loss": 0.4837, + "step": 6183 + }, + { + "epoch": 0.5125569830087029, + "grad_norm": 0.2828364670276642, + "learning_rate": 2.4376295068379615e-06, + "loss": 0.5181, + "step": 6184 + }, + { + "epoch": 0.5126398673849979, + "grad_norm": 0.2425706535577774, + "learning_rate": 2.437215084956486e-06, + "loss": 0.4774, + "step": 6185 + }, + { + "epoch": 0.512722751761293, + "grad_norm": 0.2754089832305908, + "learning_rate": 2.4368006630750104e-06, + "loss": 0.4987, + "step": 6186 + }, + { + "epoch": 0.5128056361375881, + "grad_norm": 0.25899186730384827, + "learning_rate": 2.4363862411935352e-06, + "loss": 0.5184, + "step": 6187 + }, + { + "epoch": 0.5128885205138831, + "grad_norm": 0.25861310958862305, + "learning_rate": 2.43597181931206e-06, + "loss": 0.5069, + "step": 6188 + }, + { + "epoch": 0.5129714048901782, + "grad_norm": 0.24647994339466095, + "learning_rate": 2.4355573974305845e-06, + "loss": 0.5126, + "step": 6189 + }, + { + "epoch": 0.5130542892664732, + "grad_norm": 0.2584373354911804, + "learning_rate": 2.435142975549109e-06, + "loss": 0.528, + "step": 6190 + }, + { + "epoch": 0.5131371736427683, + "grad_norm": 0.2611032724380493, + "learning_rate": 2.4347285536676338e-06, + "loss": 0.5624, + "step": 6191 + }, + { + "epoch": 0.5132200580190635, + "grad_norm": 0.2586449980735779, + "learning_rate": 2.4343141317861586e-06, + "loss": 0.4927, + "step": 6192 + }, + { + "epoch": 0.5133029423953585, + "grad_norm": 0.22415751218795776, + "learning_rate": 2.433899709904683e-06, + "loss": 0.4509, + "step": 6193 + }, + { + "epoch": 0.5133858267716536, + "grad_norm": 0.26420140266418457, + "learning_rate": 2.433485288023208e-06, + "loss": 0.5046, + "step": 6194 + }, + { + "epoch": 0.5134687111479486, + "grad_norm": 0.25795894861221313, + "learning_rate": 2.4330708661417324e-06, + "loss": 0.4915, + "step": 6195 + }, + { + "epoch": 0.5135515955242437, + "grad_norm": 0.2536274194717407, + "learning_rate": 2.4326564442602572e-06, + "loss": 0.535, + "step": 6196 + }, + { + "epoch": 0.5136344799005388, + "grad_norm": 0.2766583561897278, + "learning_rate": 2.4322420223787816e-06, + "loss": 0.51, + "step": 6197 + }, + { + "epoch": 0.5137173642768338, + "grad_norm": 0.26052963733673096, + "learning_rate": 2.4318276004973065e-06, + "loss": 0.5155, + "step": 6198 + }, + { + "epoch": 0.5138002486531289, + "grad_norm": 0.2568257451057434, + "learning_rate": 2.431413178615831e-06, + "loss": 0.5294, + "step": 6199 + }, + { + "epoch": 0.5138831330294239, + "grad_norm": 0.23941263556480408, + "learning_rate": 2.430998756734356e-06, + "loss": 0.507, + "step": 6200 + }, + { + "epoch": 0.513966017405719, + "grad_norm": 0.23756755888462067, + "learning_rate": 2.4305843348528802e-06, + "loss": 0.5213, + "step": 6201 + }, + { + "epoch": 0.5140489017820141, + "grad_norm": 0.2457093745470047, + "learning_rate": 2.430169912971405e-06, + "loss": 0.4965, + "step": 6202 + }, + { + "epoch": 0.5141317861583091, + "grad_norm": 0.24624893069267273, + "learning_rate": 2.42975549108993e-06, + "loss": 0.5331, + "step": 6203 + }, + { + "epoch": 0.5142146705346042, + "grad_norm": 0.24985045194625854, + "learning_rate": 2.4293410692084544e-06, + "loss": 0.5153, + "step": 6204 + }, + { + "epoch": 0.5142975549108993, + "grad_norm": 0.2834433615207672, + "learning_rate": 2.4289266473269792e-06, + "loss": 0.5632, + "step": 6205 + }, + { + "epoch": 0.5143804392871943, + "grad_norm": 0.25473538041114807, + "learning_rate": 2.4285122254455037e-06, + "loss": 0.5316, + "step": 6206 + }, + { + "epoch": 0.5144633236634895, + "grad_norm": 0.24349436163902283, + "learning_rate": 2.4280978035640285e-06, + "loss": 0.528, + "step": 6207 + }, + { + "epoch": 0.5145462080397845, + "grad_norm": 0.2628271281719208, + "learning_rate": 2.427683381682553e-06, + "loss": 0.5596, + "step": 6208 + }, + { + "epoch": 0.5146290924160796, + "grad_norm": 0.252024382352829, + "learning_rate": 2.427268959801078e-06, + "loss": 0.4751, + "step": 6209 + }, + { + "epoch": 0.5147119767923747, + "grad_norm": 0.2524823844432831, + "learning_rate": 2.4268545379196022e-06, + "loss": 0.5121, + "step": 6210 + }, + { + "epoch": 0.5147948611686697, + "grad_norm": 0.24200159311294556, + "learning_rate": 2.426440116038127e-06, + "loss": 0.5065, + "step": 6211 + }, + { + "epoch": 0.5148777455449648, + "grad_norm": 0.2566699981689453, + "learning_rate": 2.426025694156652e-06, + "loss": 0.5149, + "step": 6212 + }, + { + "epoch": 0.5149606299212598, + "grad_norm": 0.247910737991333, + "learning_rate": 2.4256112722751764e-06, + "loss": 0.4968, + "step": 6213 + }, + { + "epoch": 0.5150435142975549, + "grad_norm": 0.23352760076522827, + "learning_rate": 2.425196850393701e-06, + "loss": 0.4615, + "step": 6214 + }, + { + "epoch": 0.51512639867385, + "grad_norm": 0.26611319184303284, + "learning_rate": 2.4247824285122257e-06, + "loss": 0.5258, + "step": 6215 + }, + { + "epoch": 0.515209283050145, + "grad_norm": 0.2625701427459717, + "learning_rate": 2.4243680066307505e-06, + "loss": 0.4978, + "step": 6216 + }, + { + "epoch": 0.5152921674264401, + "grad_norm": 0.21929383277893066, + "learning_rate": 2.423953584749275e-06, + "loss": 0.4778, + "step": 6217 + }, + { + "epoch": 0.5153750518027351, + "grad_norm": 0.24517984688282013, + "learning_rate": 2.4235391628677994e-06, + "loss": 0.5167, + "step": 6218 + }, + { + "epoch": 0.5154579361790302, + "grad_norm": 0.25458914041519165, + "learning_rate": 2.4231247409863242e-06, + "loss": 0.4949, + "step": 6219 + }, + { + "epoch": 0.5155408205553254, + "grad_norm": 0.25140732526779175, + "learning_rate": 2.422710319104849e-06, + "loss": 0.4935, + "step": 6220 + }, + { + "epoch": 0.5156237049316204, + "grad_norm": 0.26815342903137207, + "learning_rate": 2.4222958972233735e-06, + "loss": 0.5066, + "step": 6221 + }, + { + "epoch": 0.5157065893079155, + "grad_norm": 0.2521127164363861, + "learning_rate": 2.421881475341898e-06, + "loss": 0.5404, + "step": 6222 + }, + { + "epoch": 0.5157894736842106, + "grad_norm": 0.271482914686203, + "learning_rate": 2.421467053460423e-06, + "loss": 0.5132, + "step": 6223 + }, + { + "epoch": 0.5158723580605056, + "grad_norm": 0.2629507780075073, + "learning_rate": 2.4210526315789477e-06, + "loss": 0.517, + "step": 6224 + }, + { + "epoch": 0.5159552424368007, + "grad_norm": 0.28024086356163025, + "learning_rate": 2.4206382096974725e-06, + "loss": 0.5521, + "step": 6225 + }, + { + "epoch": 0.5160381268130957, + "grad_norm": 0.24867132306098938, + "learning_rate": 2.420223787815997e-06, + "loss": 0.5209, + "step": 6226 + }, + { + "epoch": 0.5161210111893908, + "grad_norm": 0.2536858320236206, + "learning_rate": 2.4198093659345214e-06, + "loss": 0.5356, + "step": 6227 + }, + { + "epoch": 0.5162038955656859, + "grad_norm": 0.29869142174720764, + "learning_rate": 2.4193949440530462e-06, + "loss": 0.5385, + "step": 6228 + }, + { + "epoch": 0.5162867799419809, + "grad_norm": 0.25012534856796265, + "learning_rate": 2.418980522171571e-06, + "loss": 0.4876, + "step": 6229 + }, + { + "epoch": 0.516369664318276, + "grad_norm": 0.2768109440803528, + "learning_rate": 2.4185661002900955e-06, + "loss": 0.5205, + "step": 6230 + }, + { + "epoch": 0.516452548694571, + "grad_norm": 0.25550955533981323, + "learning_rate": 2.41815167840862e-06, + "loss": 0.5391, + "step": 6231 + }, + { + "epoch": 0.5165354330708661, + "grad_norm": 0.2679550349712372, + "learning_rate": 2.417737256527145e-06, + "loss": 0.4966, + "step": 6232 + }, + { + "epoch": 0.5166183174471612, + "grad_norm": 0.24553929269313812, + "learning_rate": 2.4173228346456697e-06, + "loss": 0.5317, + "step": 6233 + }, + { + "epoch": 0.5167012018234562, + "grad_norm": 0.2509346008300781, + "learning_rate": 2.416908412764194e-06, + "loss": 0.5129, + "step": 6234 + }, + { + "epoch": 0.5167840861997514, + "grad_norm": 0.2537175118923187, + "learning_rate": 2.4164939908827185e-06, + "loss": 0.4821, + "step": 6235 + }, + { + "epoch": 0.5168669705760465, + "grad_norm": 0.2539937496185303, + "learning_rate": 2.4160795690012434e-06, + "loss": 0.4798, + "step": 6236 + }, + { + "epoch": 0.5169498549523415, + "grad_norm": 0.25680992007255554, + "learning_rate": 2.4156651471197682e-06, + "loss": 0.5186, + "step": 6237 + }, + { + "epoch": 0.5170327393286366, + "grad_norm": 0.2218734472990036, + "learning_rate": 2.4152507252382927e-06, + "loss": 0.4731, + "step": 6238 + }, + { + "epoch": 0.5171156237049316, + "grad_norm": 0.248477965593338, + "learning_rate": 2.4148363033568175e-06, + "loss": 0.4982, + "step": 6239 + }, + { + "epoch": 0.5171985080812267, + "grad_norm": 0.27785447239875793, + "learning_rate": 2.414421881475342e-06, + "loss": 0.5044, + "step": 6240 + }, + { + "epoch": 0.5172813924575218, + "grad_norm": 0.24546565115451813, + "learning_rate": 2.414007459593867e-06, + "loss": 0.482, + "step": 6241 + }, + { + "epoch": 0.5173642768338168, + "grad_norm": 0.27508896589279175, + "learning_rate": 2.4135930377123912e-06, + "loss": 0.5207, + "step": 6242 + }, + { + "epoch": 0.5174471612101119, + "grad_norm": 0.23618194460868835, + "learning_rate": 2.413178615830916e-06, + "loss": 0.4862, + "step": 6243 + }, + { + "epoch": 0.5175300455864069, + "grad_norm": 0.25140032172203064, + "learning_rate": 2.4127641939494405e-06, + "loss": 0.5285, + "step": 6244 + }, + { + "epoch": 0.517612929962702, + "grad_norm": 0.27826040983200073, + "learning_rate": 2.4123497720679654e-06, + "loss": 0.5229, + "step": 6245 + }, + { + "epoch": 0.5176958143389971, + "grad_norm": 0.256656676530838, + "learning_rate": 2.4119353501864902e-06, + "loss": 0.5054, + "step": 6246 + }, + { + "epoch": 0.5177786987152921, + "grad_norm": 0.2548012137413025, + "learning_rate": 2.4115209283050147e-06, + "loss": 0.5056, + "step": 6247 + }, + { + "epoch": 0.5178615830915873, + "grad_norm": 0.2332487255334854, + "learning_rate": 2.4111065064235395e-06, + "loss": 0.4682, + "step": 6248 + }, + { + "epoch": 0.5179444674678823, + "grad_norm": 0.25487226247787476, + "learning_rate": 2.410692084542064e-06, + "loss": 0.5147, + "step": 6249 + }, + { + "epoch": 0.5180273518441774, + "grad_norm": 0.2491910606622696, + "learning_rate": 2.410277662660589e-06, + "loss": 0.545, + "step": 6250 + }, + { + "epoch": 0.5181102362204725, + "grad_norm": 0.26368531584739685, + "learning_rate": 2.4098632407791132e-06, + "loss": 0.5194, + "step": 6251 + }, + { + "epoch": 0.5181931205967675, + "grad_norm": 0.2408839613199234, + "learning_rate": 2.409448818897638e-06, + "loss": 0.4887, + "step": 6252 + }, + { + "epoch": 0.5182760049730626, + "grad_norm": 0.27526745200157166, + "learning_rate": 2.4090343970161625e-06, + "loss": 0.5287, + "step": 6253 + }, + { + "epoch": 0.5183588893493577, + "grad_norm": 0.2835533022880554, + "learning_rate": 2.4086199751346874e-06, + "loss": 0.4965, + "step": 6254 + }, + { + "epoch": 0.5184417737256527, + "grad_norm": 0.2635200619697571, + "learning_rate": 2.408205553253212e-06, + "loss": 0.5372, + "step": 6255 + }, + { + "epoch": 0.5185246581019478, + "grad_norm": 0.2599073052406311, + "learning_rate": 2.4077911313717367e-06, + "loss": 0.5312, + "step": 6256 + }, + { + "epoch": 0.5186075424782428, + "grad_norm": 0.23983134329319, + "learning_rate": 2.4073767094902615e-06, + "loss": 0.4843, + "step": 6257 + }, + { + "epoch": 0.5186904268545379, + "grad_norm": 0.26278141140937805, + "learning_rate": 2.406962287608786e-06, + "loss": 0.4714, + "step": 6258 + }, + { + "epoch": 0.518773311230833, + "grad_norm": 0.27547335624694824, + "learning_rate": 2.4065478657273104e-06, + "loss": 0.5084, + "step": 6259 + }, + { + "epoch": 0.518856195607128, + "grad_norm": 0.2508985102176666, + "learning_rate": 2.4061334438458353e-06, + "loss": 0.4891, + "step": 6260 + }, + { + "epoch": 0.5189390799834231, + "grad_norm": 0.2716338336467743, + "learning_rate": 2.40571902196436e-06, + "loss": 0.5271, + "step": 6261 + }, + { + "epoch": 0.5190219643597181, + "grad_norm": 0.25849470496177673, + "learning_rate": 2.4053046000828845e-06, + "loss": 0.5509, + "step": 6262 + }, + { + "epoch": 0.5191048487360133, + "grad_norm": 0.26523837447166443, + "learning_rate": 2.4048901782014094e-06, + "loss": 0.5019, + "step": 6263 + }, + { + "epoch": 0.5191877331123084, + "grad_norm": 0.2661520838737488, + "learning_rate": 2.404475756319934e-06, + "loss": 0.5292, + "step": 6264 + }, + { + "epoch": 0.5192706174886034, + "grad_norm": 0.23894859850406647, + "learning_rate": 2.4040613344384587e-06, + "loss": 0.4963, + "step": 6265 + }, + { + "epoch": 0.5193535018648985, + "grad_norm": 0.260564386844635, + "learning_rate": 2.403646912556983e-06, + "loss": 0.5327, + "step": 6266 + }, + { + "epoch": 0.5194363862411935, + "grad_norm": 0.2549040615558624, + "learning_rate": 2.403232490675508e-06, + "loss": 0.5158, + "step": 6267 + }, + { + "epoch": 0.5195192706174886, + "grad_norm": 0.26590806245803833, + "learning_rate": 2.4028180687940324e-06, + "loss": 0.5444, + "step": 6268 + }, + { + "epoch": 0.5196021549937837, + "grad_norm": 0.2523225247859955, + "learning_rate": 2.4024036469125573e-06, + "loss": 0.535, + "step": 6269 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 0.2540481686592102, + "learning_rate": 2.401989225031082e-06, + "loss": 0.5077, + "step": 6270 + }, + { + "epoch": 0.5197679237463738, + "grad_norm": 0.2513411343097687, + "learning_rate": 2.4015748031496065e-06, + "loss": 0.5259, + "step": 6271 + }, + { + "epoch": 0.5198508081226689, + "grad_norm": 0.27512454986572266, + "learning_rate": 2.401160381268131e-06, + "loss": 0.5302, + "step": 6272 + }, + { + "epoch": 0.5199336924989639, + "grad_norm": 0.23963749408721924, + "learning_rate": 2.400745959386656e-06, + "loss": 0.4895, + "step": 6273 + }, + { + "epoch": 0.520016576875259, + "grad_norm": 0.24748453497886658, + "learning_rate": 2.4003315375051807e-06, + "loss": 0.5127, + "step": 6274 + }, + { + "epoch": 0.520099461251554, + "grad_norm": 0.25132274627685547, + "learning_rate": 2.399917115623705e-06, + "loss": 0.5579, + "step": 6275 + }, + { + "epoch": 0.5201823456278492, + "grad_norm": 0.2578490078449249, + "learning_rate": 2.3995026937422295e-06, + "loss": 0.547, + "step": 6276 + }, + { + "epoch": 0.5202652300041443, + "grad_norm": 0.26320335268974304, + "learning_rate": 2.3990882718607544e-06, + "loss": 0.5404, + "step": 6277 + }, + { + "epoch": 0.5203481143804393, + "grad_norm": 0.24796119332313538, + "learning_rate": 2.3986738499792793e-06, + "loss": 0.4798, + "step": 6278 + }, + { + "epoch": 0.5204309987567344, + "grad_norm": 0.24733442068099976, + "learning_rate": 2.3982594280978037e-06, + "loss": 0.5111, + "step": 6279 + }, + { + "epoch": 0.5205138831330294, + "grad_norm": 0.25351086258888245, + "learning_rate": 2.397845006216328e-06, + "loss": 0.519, + "step": 6280 + }, + { + "epoch": 0.5205967675093245, + "grad_norm": 0.27116328477859497, + "learning_rate": 2.397430584334853e-06, + "loss": 0.5375, + "step": 6281 + }, + { + "epoch": 0.5206796518856196, + "grad_norm": 0.2702968716621399, + "learning_rate": 2.397016162453378e-06, + "loss": 0.5106, + "step": 6282 + }, + { + "epoch": 0.5207625362619146, + "grad_norm": 0.2576724588871002, + "learning_rate": 2.3966017405719027e-06, + "loss": 0.5231, + "step": 6283 + }, + { + "epoch": 0.5208454206382097, + "grad_norm": 0.2596430778503418, + "learning_rate": 2.396187318690427e-06, + "loss": 0.5266, + "step": 6284 + }, + { + "epoch": 0.5209283050145047, + "grad_norm": 0.2509610056877136, + "learning_rate": 2.3957728968089516e-06, + "loss": 0.5135, + "step": 6285 + }, + { + "epoch": 0.5210111893907998, + "grad_norm": 0.2716809809207916, + "learning_rate": 2.3953584749274764e-06, + "loss": 0.4872, + "step": 6286 + }, + { + "epoch": 0.5210940737670949, + "grad_norm": 0.26643800735473633, + "learning_rate": 2.3949440530460013e-06, + "loss": 0.5557, + "step": 6287 + }, + { + "epoch": 0.5211769581433899, + "grad_norm": 0.24204020202159882, + "learning_rate": 2.3945296311645257e-06, + "loss": 0.4921, + "step": 6288 + }, + { + "epoch": 0.521259842519685, + "grad_norm": 0.25111740827560425, + "learning_rate": 2.39411520928305e-06, + "loss": 0.5011, + "step": 6289 + }, + { + "epoch": 0.5213427268959802, + "grad_norm": 0.23753005266189575, + "learning_rate": 2.393700787401575e-06, + "loss": 0.4917, + "step": 6290 + }, + { + "epoch": 0.5214256112722752, + "grad_norm": 0.24855484068393707, + "learning_rate": 2.3932863655201e-06, + "loss": 0.5518, + "step": 6291 + }, + { + "epoch": 0.5215084956485703, + "grad_norm": 0.2822589576244354, + "learning_rate": 2.3928719436386243e-06, + "loss": 0.5285, + "step": 6292 + }, + { + "epoch": 0.5215913800248653, + "grad_norm": 0.2685754597187042, + "learning_rate": 2.3924575217571487e-06, + "loss": 0.5016, + "step": 6293 + }, + { + "epoch": 0.5216742644011604, + "grad_norm": 0.23552730679512024, + "learning_rate": 2.3920430998756736e-06, + "loss": 0.4605, + "step": 6294 + }, + { + "epoch": 0.5217571487774555, + "grad_norm": 0.2684164047241211, + "learning_rate": 2.3916286779941984e-06, + "loss": 0.54, + "step": 6295 + }, + { + "epoch": 0.5218400331537505, + "grad_norm": 0.25782936811447144, + "learning_rate": 2.391214256112723e-06, + "loss": 0.5291, + "step": 6296 + }, + { + "epoch": 0.5219229175300456, + "grad_norm": 0.25952082872390747, + "learning_rate": 2.3907998342312477e-06, + "loss": 0.5481, + "step": 6297 + }, + { + "epoch": 0.5220058019063406, + "grad_norm": 0.25399336218833923, + "learning_rate": 2.390385412349772e-06, + "loss": 0.5379, + "step": 6298 + }, + { + "epoch": 0.5220886862826357, + "grad_norm": 0.2691109776496887, + "learning_rate": 2.389970990468297e-06, + "loss": 0.4992, + "step": 6299 + }, + { + "epoch": 0.5221715706589308, + "grad_norm": 0.26331397891044617, + "learning_rate": 2.3895565685868214e-06, + "loss": 0.5244, + "step": 6300 + }, + { + "epoch": 0.5222544550352258, + "grad_norm": 0.23934200406074524, + "learning_rate": 2.3891421467053463e-06, + "loss": 0.4751, + "step": 6301 + }, + { + "epoch": 0.5223373394115209, + "grad_norm": 0.24232764542102814, + "learning_rate": 2.3887277248238707e-06, + "loss": 0.5408, + "step": 6302 + }, + { + "epoch": 0.522420223787816, + "grad_norm": 0.2710106372833252, + "learning_rate": 2.3883133029423956e-06, + "loss": 0.5204, + "step": 6303 + }, + { + "epoch": 0.522503108164111, + "grad_norm": 1.4844732284545898, + "learning_rate": 2.3878988810609204e-06, + "loss": 0.4881, + "step": 6304 + }, + { + "epoch": 0.5225859925404062, + "grad_norm": 0.25909677147865295, + "learning_rate": 2.387484459179445e-06, + "loss": 0.5055, + "step": 6305 + }, + { + "epoch": 0.5226688769167012, + "grad_norm": 0.24364051222801208, + "learning_rate": 2.3870700372979697e-06, + "loss": 0.4717, + "step": 6306 + }, + { + "epoch": 0.5227517612929963, + "grad_norm": 0.24241317808628082, + "learning_rate": 2.386655615416494e-06, + "loss": 0.4951, + "step": 6307 + }, + { + "epoch": 0.5228346456692914, + "grad_norm": 0.27640458941459656, + "learning_rate": 2.386241193535019e-06, + "loss": 0.5772, + "step": 6308 + }, + { + "epoch": 0.5229175300455864, + "grad_norm": 0.26261502504348755, + "learning_rate": 2.3858267716535434e-06, + "loss": 0.502, + "step": 6309 + }, + { + "epoch": 0.5230004144218815, + "grad_norm": 0.24261975288391113, + "learning_rate": 2.3854123497720683e-06, + "loss": 0.489, + "step": 6310 + }, + { + "epoch": 0.5230832987981765, + "grad_norm": 0.25761377811431885, + "learning_rate": 2.3849979278905927e-06, + "loss": 0.5211, + "step": 6311 + }, + { + "epoch": 0.5231661831744716, + "grad_norm": 0.24896498024463654, + "learning_rate": 2.3845835060091176e-06, + "loss": 0.5151, + "step": 6312 + }, + { + "epoch": 0.5232490675507667, + "grad_norm": 0.2759874761104584, + "learning_rate": 2.384169084127642e-06, + "loss": 0.5299, + "step": 6313 + }, + { + "epoch": 0.5233319519270617, + "grad_norm": 0.2376255840063095, + "learning_rate": 2.383754662246167e-06, + "loss": 0.4922, + "step": 6314 + }, + { + "epoch": 0.5234148363033568, + "grad_norm": 0.2653898596763611, + "learning_rate": 2.3833402403646917e-06, + "loss": 0.515, + "step": 6315 + }, + { + "epoch": 0.5234977206796518, + "grad_norm": 0.2534277141094208, + "learning_rate": 2.382925818483216e-06, + "loss": 0.5045, + "step": 6316 + }, + { + "epoch": 0.523580605055947, + "grad_norm": 0.25924932956695557, + "learning_rate": 2.3825113966017406e-06, + "loss": 0.5307, + "step": 6317 + }, + { + "epoch": 0.5236634894322421, + "grad_norm": 0.2389492392539978, + "learning_rate": 2.3820969747202654e-06, + "loss": 0.4559, + "step": 6318 + }, + { + "epoch": 0.5237463738085371, + "grad_norm": 0.27069151401519775, + "learning_rate": 2.3816825528387903e-06, + "loss": 0.4828, + "step": 6319 + }, + { + "epoch": 0.5238292581848322, + "grad_norm": 0.27537307143211365, + "learning_rate": 2.3812681309573147e-06, + "loss": 0.517, + "step": 6320 + }, + { + "epoch": 0.5239121425611273, + "grad_norm": 0.2634919285774231, + "learning_rate": 2.380853709075839e-06, + "loss": 0.5483, + "step": 6321 + }, + { + "epoch": 0.5239950269374223, + "grad_norm": 0.24671733379364014, + "learning_rate": 2.380439287194364e-06, + "loss": 0.5175, + "step": 6322 + }, + { + "epoch": 0.5240779113137174, + "grad_norm": 0.2325306236743927, + "learning_rate": 2.380024865312889e-06, + "loss": 0.503, + "step": 6323 + }, + { + "epoch": 0.5241607956900124, + "grad_norm": 0.26248323917388916, + "learning_rate": 2.3796104434314137e-06, + "loss": 0.5095, + "step": 6324 + }, + { + "epoch": 0.5242436800663075, + "grad_norm": 0.2602270543575287, + "learning_rate": 2.379196021549938e-06, + "loss": 0.4861, + "step": 6325 + }, + { + "epoch": 0.5243265644426026, + "grad_norm": 0.24077925086021423, + "learning_rate": 2.3787815996684626e-06, + "loss": 0.5188, + "step": 6326 + }, + { + "epoch": 0.5244094488188976, + "grad_norm": 0.2485741376876831, + "learning_rate": 2.3783671777869874e-06, + "loss": 0.5082, + "step": 6327 + }, + { + "epoch": 0.5244923331951927, + "grad_norm": 0.26194751262664795, + "learning_rate": 2.3779527559055123e-06, + "loss": 0.4959, + "step": 6328 + }, + { + "epoch": 0.5245752175714877, + "grad_norm": 0.24631695449352264, + "learning_rate": 2.3775383340240367e-06, + "loss": 0.522, + "step": 6329 + }, + { + "epoch": 0.5246581019477828, + "grad_norm": 0.25294584035873413, + "learning_rate": 2.377123912142561e-06, + "loss": 0.526, + "step": 6330 + }, + { + "epoch": 0.524740986324078, + "grad_norm": 0.2295425683259964, + "learning_rate": 2.376709490261086e-06, + "loss": 0.4694, + "step": 6331 + }, + { + "epoch": 0.524823870700373, + "grad_norm": 0.3641234040260315, + "learning_rate": 2.376295068379611e-06, + "loss": 0.4815, + "step": 6332 + }, + { + "epoch": 0.5249067550766681, + "grad_norm": 0.2522130012512207, + "learning_rate": 2.3758806464981353e-06, + "loss": 0.5253, + "step": 6333 + }, + { + "epoch": 0.5249896394529631, + "grad_norm": 0.23895667493343353, + "learning_rate": 2.3754662246166597e-06, + "loss": 0.4852, + "step": 6334 + }, + { + "epoch": 0.5250725238292582, + "grad_norm": 0.2563622295856476, + "learning_rate": 2.3750518027351846e-06, + "loss": 0.5231, + "step": 6335 + }, + { + "epoch": 0.5251554082055533, + "grad_norm": 0.25018808245658875, + "learning_rate": 2.3746373808537094e-06, + "loss": 0.5154, + "step": 6336 + }, + { + "epoch": 0.5252382925818483, + "grad_norm": 0.24506638944149017, + "learning_rate": 2.374222958972234e-06, + "loss": 0.4769, + "step": 6337 + }, + { + "epoch": 0.5253211769581434, + "grad_norm": 0.24945957958698273, + "learning_rate": 2.3738085370907583e-06, + "loss": 0.5323, + "step": 6338 + }, + { + "epoch": 0.5254040613344385, + "grad_norm": 0.2545723617076874, + "learning_rate": 2.373394115209283e-06, + "loss": 0.4992, + "step": 6339 + }, + { + "epoch": 0.5254869457107335, + "grad_norm": 0.24065318703651428, + "learning_rate": 2.372979693327808e-06, + "loss": 0.5172, + "step": 6340 + }, + { + "epoch": 0.5255698300870286, + "grad_norm": 0.25046178698539734, + "learning_rate": 2.3725652714463324e-06, + "loss": 0.5123, + "step": 6341 + }, + { + "epoch": 0.5256527144633236, + "grad_norm": 0.2719615697860718, + "learning_rate": 2.3721508495648573e-06, + "loss": 0.5735, + "step": 6342 + }, + { + "epoch": 0.5257355988396187, + "grad_norm": 0.24852895736694336, + "learning_rate": 2.3717364276833817e-06, + "loss": 0.53, + "step": 6343 + }, + { + "epoch": 0.5258184832159138, + "grad_norm": 0.28320395946502686, + "learning_rate": 2.3713220058019066e-06, + "loss": 0.5258, + "step": 6344 + }, + { + "epoch": 0.5259013675922088, + "grad_norm": 0.24431675672531128, + "learning_rate": 2.3709075839204314e-06, + "loss": 0.496, + "step": 6345 + }, + { + "epoch": 0.525984251968504, + "grad_norm": 0.25009557604789734, + "learning_rate": 2.370493162038956e-06, + "loss": 0.5008, + "step": 6346 + }, + { + "epoch": 0.526067136344799, + "grad_norm": 0.2650800943374634, + "learning_rate": 2.3700787401574803e-06, + "loss": 0.5165, + "step": 6347 + }, + { + "epoch": 0.5261500207210941, + "grad_norm": 0.2520666718482971, + "learning_rate": 2.369664318276005e-06, + "loss": 0.5412, + "step": 6348 + }, + { + "epoch": 0.5262329050973892, + "grad_norm": 0.2768591642379761, + "learning_rate": 2.36924989639453e-06, + "loss": 0.5055, + "step": 6349 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 0.24288544058799744, + "learning_rate": 2.3688354745130544e-06, + "loss": 0.5045, + "step": 6350 + }, + { + "epoch": 0.5263986738499793, + "grad_norm": 0.26180458068847656, + "learning_rate": 2.368421052631579e-06, + "loss": 0.5634, + "step": 6351 + }, + { + "epoch": 0.5264815582262743, + "grad_norm": 0.2672177255153656, + "learning_rate": 2.3680066307501037e-06, + "loss": 0.5756, + "step": 6352 + }, + { + "epoch": 0.5265644426025694, + "grad_norm": 0.24531570076942444, + "learning_rate": 2.3675922088686286e-06, + "loss": 0.5065, + "step": 6353 + }, + { + "epoch": 0.5266473269788645, + "grad_norm": 0.26691150665283203, + "learning_rate": 2.367177786987153e-06, + "loss": 0.5018, + "step": 6354 + }, + { + "epoch": 0.5267302113551595, + "grad_norm": 0.2566821277141571, + "learning_rate": 2.366763365105678e-06, + "loss": 0.4822, + "step": 6355 + }, + { + "epoch": 0.5268130957314546, + "grad_norm": 0.24459215998649597, + "learning_rate": 2.3663489432242023e-06, + "loss": 0.5114, + "step": 6356 + }, + { + "epoch": 0.5268959801077497, + "grad_norm": 0.259872704744339, + "learning_rate": 2.365934521342727e-06, + "loss": 0.498, + "step": 6357 + }, + { + "epoch": 0.5269788644840447, + "grad_norm": 0.2709864377975464, + "learning_rate": 2.3655200994612516e-06, + "loss": 0.5462, + "step": 6358 + }, + { + "epoch": 0.5270617488603399, + "grad_norm": 0.2575525939464569, + "learning_rate": 2.3651056775797764e-06, + "loss": 0.5457, + "step": 6359 + }, + { + "epoch": 0.5271446332366349, + "grad_norm": 0.2488408088684082, + "learning_rate": 2.364691255698301e-06, + "loss": 0.5181, + "step": 6360 + }, + { + "epoch": 0.52722751761293, + "grad_norm": 0.2466326206922531, + "learning_rate": 2.3642768338168257e-06, + "loss": 0.5327, + "step": 6361 + }, + { + "epoch": 0.5273104019892251, + "grad_norm": 0.24372151494026184, + "learning_rate": 2.3638624119353506e-06, + "loss": 0.5316, + "step": 6362 + }, + { + "epoch": 0.5273932863655201, + "grad_norm": 0.253709614276886, + "learning_rate": 2.363447990053875e-06, + "loss": 0.5009, + "step": 6363 + }, + { + "epoch": 0.5274761707418152, + "grad_norm": 0.2508584260940552, + "learning_rate": 2.3630335681724e-06, + "loss": 0.5182, + "step": 6364 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 0.26674598455429077, + "learning_rate": 2.3626191462909243e-06, + "loss": 0.571, + "step": 6365 + }, + { + "epoch": 0.5276419394944053, + "grad_norm": 0.24264514446258545, + "learning_rate": 2.362204724409449e-06, + "loss": 0.5169, + "step": 6366 + }, + { + "epoch": 0.5277248238707004, + "grad_norm": 0.23980151116847992, + "learning_rate": 2.3617903025279736e-06, + "loss": 0.5083, + "step": 6367 + }, + { + "epoch": 0.5278077082469954, + "grad_norm": 0.2569444179534912, + "learning_rate": 2.3613758806464984e-06, + "loss": 0.5065, + "step": 6368 + }, + { + "epoch": 0.5278905926232905, + "grad_norm": 0.25937119126319885, + "learning_rate": 2.360961458765023e-06, + "loss": 0.5058, + "step": 6369 + }, + { + "epoch": 0.5279734769995856, + "grad_norm": 0.2731715440750122, + "learning_rate": 2.3605470368835477e-06, + "loss": 0.5359, + "step": 6370 + }, + { + "epoch": 0.5280563613758806, + "grad_norm": 0.2567588984966278, + "learning_rate": 2.360132615002072e-06, + "loss": 0.5384, + "step": 6371 + }, + { + "epoch": 0.5281392457521757, + "grad_norm": 0.2555687725543976, + "learning_rate": 2.359718193120597e-06, + "loss": 0.5319, + "step": 6372 + }, + { + "epoch": 0.5282221301284707, + "grad_norm": 0.27566078305244446, + "learning_rate": 2.359303771239122e-06, + "loss": 0.5146, + "step": 6373 + }, + { + "epoch": 0.5283050145047659, + "grad_norm": 0.24831487238407135, + "learning_rate": 2.3588893493576463e-06, + "loss": 0.529, + "step": 6374 + }, + { + "epoch": 0.528387898881061, + "grad_norm": 0.24258820712566376, + "learning_rate": 2.3584749274761707e-06, + "loss": 0.5109, + "step": 6375 + }, + { + "epoch": 0.528470783257356, + "grad_norm": 0.2659376263618469, + "learning_rate": 2.3580605055946956e-06, + "loss": 0.5245, + "step": 6376 + }, + { + "epoch": 0.5285536676336511, + "grad_norm": 0.2636216878890991, + "learning_rate": 2.3576460837132205e-06, + "loss": 0.529, + "step": 6377 + }, + { + "epoch": 0.5286365520099461, + "grad_norm": 0.2664233148097992, + "learning_rate": 2.357231661831745e-06, + "loss": 0.5375, + "step": 6378 + }, + { + "epoch": 0.5287194363862412, + "grad_norm": 0.24140875041484833, + "learning_rate": 2.3568172399502693e-06, + "loss": 0.5119, + "step": 6379 + }, + { + "epoch": 0.5288023207625363, + "grad_norm": 0.2535066604614258, + "learning_rate": 2.356402818068794e-06, + "loss": 0.5464, + "step": 6380 + }, + { + "epoch": 0.5288852051388313, + "grad_norm": 0.27596038579940796, + "learning_rate": 2.355988396187319e-06, + "loss": 0.5179, + "step": 6381 + }, + { + "epoch": 0.5289680895151264, + "grad_norm": 0.25437888503074646, + "learning_rate": 2.3555739743058435e-06, + "loss": 0.4572, + "step": 6382 + }, + { + "epoch": 0.5290509738914214, + "grad_norm": 0.2412635087966919, + "learning_rate": 2.3551595524243683e-06, + "loss": 0.4927, + "step": 6383 + }, + { + "epoch": 0.5291338582677165, + "grad_norm": 0.238821879029274, + "learning_rate": 2.3547451305428927e-06, + "loss": 0.4911, + "step": 6384 + }, + { + "epoch": 0.5292167426440116, + "grad_norm": 0.2707713544368744, + "learning_rate": 2.3543307086614176e-06, + "loss": 0.5217, + "step": 6385 + }, + { + "epoch": 0.5292996270203066, + "grad_norm": 0.26185449957847595, + "learning_rate": 2.3539162867799425e-06, + "loss": 0.5031, + "step": 6386 + }, + { + "epoch": 0.5293825113966018, + "grad_norm": 0.25654590129852295, + "learning_rate": 2.353501864898467e-06, + "loss": 0.513, + "step": 6387 + }, + { + "epoch": 0.5294653957728969, + "grad_norm": 0.25019189715385437, + "learning_rate": 2.3530874430169913e-06, + "loss": 0.5064, + "step": 6388 + }, + { + "epoch": 0.5295482801491919, + "grad_norm": 0.23460635542869568, + "learning_rate": 2.352673021135516e-06, + "loss": 0.5038, + "step": 6389 + }, + { + "epoch": 0.529631164525487, + "grad_norm": 0.24186354875564575, + "learning_rate": 2.352258599254041e-06, + "loss": 0.466, + "step": 6390 + }, + { + "epoch": 0.529714048901782, + "grad_norm": 0.25244656205177307, + "learning_rate": 2.3518441773725655e-06, + "loss": 0.5349, + "step": 6391 + }, + { + "epoch": 0.5297969332780771, + "grad_norm": 0.2693803012371063, + "learning_rate": 2.35142975549109e-06, + "loss": 0.501, + "step": 6392 + }, + { + "epoch": 0.5298798176543722, + "grad_norm": 0.2642437219619751, + "learning_rate": 2.3510153336096147e-06, + "loss": 0.5871, + "step": 6393 + }, + { + "epoch": 0.5299627020306672, + "grad_norm": 0.25743213295936584, + "learning_rate": 2.3506009117281396e-06, + "loss": 0.5415, + "step": 6394 + }, + { + "epoch": 0.5300455864069623, + "grad_norm": 0.2460259199142456, + "learning_rate": 2.350186489846664e-06, + "loss": 0.4462, + "step": 6395 + }, + { + "epoch": 0.5301284707832573, + "grad_norm": 0.2390514463186264, + "learning_rate": 2.3497720679651885e-06, + "loss": 0.4952, + "step": 6396 + }, + { + "epoch": 0.5302113551595524, + "grad_norm": 0.24070106446743011, + "learning_rate": 2.3493576460837133e-06, + "loss": 0.5015, + "step": 6397 + }, + { + "epoch": 0.5302942395358475, + "grad_norm": 0.2496921271085739, + "learning_rate": 2.348943224202238e-06, + "loss": 0.4921, + "step": 6398 + }, + { + "epoch": 0.5303771239121425, + "grad_norm": 0.23851247131824493, + "learning_rate": 2.3485288023207626e-06, + "loss": 0.4866, + "step": 6399 + }, + { + "epoch": 0.5304600082884376, + "grad_norm": 0.26873379945755005, + "learning_rate": 2.3481143804392875e-06, + "loss": 0.5003, + "step": 6400 + }, + { + "epoch": 0.5305428926647326, + "grad_norm": 0.24558179080486298, + "learning_rate": 2.347699958557812e-06, + "loss": 0.4944, + "step": 6401 + }, + { + "epoch": 0.5306257770410278, + "grad_norm": 0.2482336461544037, + "learning_rate": 2.3472855366763368e-06, + "loss": 0.5439, + "step": 6402 + }, + { + "epoch": 0.5307086614173229, + "grad_norm": 0.2447933405637741, + "learning_rate": 2.3468711147948616e-06, + "loss": 0.5321, + "step": 6403 + }, + { + "epoch": 0.5307915457936179, + "grad_norm": 0.2583220899105072, + "learning_rate": 2.346456692913386e-06, + "loss": 0.4839, + "step": 6404 + }, + { + "epoch": 0.530874430169913, + "grad_norm": 0.2625066936016083, + "learning_rate": 2.3460422710319105e-06, + "loss": 0.517, + "step": 6405 + }, + { + "epoch": 0.5309573145462081, + "grad_norm": 0.22728972136974335, + "learning_rate": 2.3456278491504353e-06, + "loss": 0.4695, + "step": 6406 + }, + { + "epoch": 0.5310401989225031, + "grad_norm": 0.2468901127576828, + "learning_rate": 2.34521342726896e-06, + "loss": 0.4855, + "step": 6407 + }, + { + "epoch": 0.5311230832987982, + "grad_norm": 0.266544908285141, + "learning_rate": 2.3447990053874846e-06, + "loss": 0.5505, + "step": 6408 + }, + { + "epoch": 0.5312059676750932, + "grad_norm": 0.25595930218696594, + "learning_rate": 2.344384583506009e-06, + "loss": 0.5133, + "step": 6409 + }, + { + "epoch": 0.5312888520513883, + "grad_norm": 0.2495349794626236, + "learning_rate": 2.343970161624534e-06, + "loss": 0.4945, + "step": 6410 + }, + { + "epoch": 0.5313717364276834, + "grad_norm": 0.23341767489910126, + "learning_rate": 2.3435557397430588e-06, + "loss": 0.4733, + "step": 6411 + }, + { + "epoch": 0.5314546208039784, + "grad_norm": 0.28245553374290466, + "learning_rate": 2.343141317861583e-06, + "loss": 0.5312, + "step": 6412 + }, + { + "epoch": 0.5315375051802735, + "grad_norm": 0.25884687900543213, + "learning_rate": 2.342726895980108e-06, + "loss": 0.508, + "step": 6413 + }, + { + "epoch": 0.5316203895565685, + "grad_norm": 0.23701637983322144, + "learning_rate": 2.3423124740986325e-06, + "loss": 0.4874, + "step": 6414 + }, + { + "epoch": 0.5317032739328637, + "grad_norm": 0.2498830258846283, + "learning_rate": 2.3418980522171573e-06, + "loss": 0.5086, + "step": 6415 + }, + { + "epoch": 0.5317861583091588, + "grad_norm": 0.24463145434856415, + "learning_rate": 2.3414836303356818e-06, + "loss": 0.5173, + "step": 6416 + }, + { + "epoch": 0.5318690426854538, + "grad_norm": 0.2671085298061371, + "learning_rate": 2.3410692084542066e-06, + "loss": 0.4963, + "step": 6417 + }, + { + "epoch": 0.5319519270617489, + "grad_norm": 0.2392653077840805, + "learning_rate": 2.340654786572731e-06, + "loss": 0.5321, + "step": 6418 + }, + { + "epoch": 0.5320348114380439, + "grad_norm": 0.2502838373184204, + "learning_rate": 2.340240364691256e-06, + "loss": 0.4899, + "step": 6419 + }, + { + "epoch": 0.532117695814339, + "grad_norm": 0.23441839218139648, + "learning_rate": 2.3398259428097803e-06, + "loss": 0.4993, + "step": 6420 + }, + { + "epoch": 0.5322005801906341, + "grad_norm": 0.2801746428012848, + "learning_rate": 2.339411520928305e-06, + "loss": 0.5486, + "step": 6421 + }, + { + "epoch": 0.5322834645669291, + "grad_norm": 0.26413431763648987, + "learning_rate": 2.33899709904683e-06, + "loss": 0.5311, + "step": 6422 + }, + { + "epoch": 0.5323663489432242, + "grad_norm": 0.23270761966705322, + "learning_rate": 2.3385826771653545e-06, + "loss": 0.5033, + "step": 6423 + }, + { + "epoch": 0.5324492333195193, + "grad_norm": 0.2547384798526764, + "learning_rate": 2.3381682552838793e-06, + "loss": 0.5126, + "step": 6424 + }, + { + "epoch": 0.5325321176958143, + "grad_norm": 0.26083824038505554, + "learning_rate": 2.3377538334024038e-06, + "loss": 0.5354, + "step": 6425 + }, + { + "epoch": 0.5326150020721094, + "grad_norm": 0.25195205211639404, + "learning_rate": 2.3373394115209286e-06, + "loss": 0.5168, + "step": 6426 + }, + { + "epoch": 0.5326978864484044, + "grad_norm": 0.24227792024612427, + "learning_rate": 2.336924989639453e-06, + "loss": 0.4814, + "step": 6427 + }, + { + "epoch": 0.5327807708246995, + "grad_norm": 0.2606436312198639, + "learning_rate": 2.336510567757978e-06, + "loss": 0.5087, + "step": 6428 + }, + { + "epoch": 0.5328636552009947, + "grad_norm": 0.23782283067703247, + "learning_rate": 2.3360961458765023e-06, + "loss": 0.4931, + "step": 6429 + }, + { + "epoch": 0.5329465395772897, + "grad_norm": 1.042517066001892, + "learning_rate": 2.335681723995027e-06, + "loss": 0.4593, + "step": 6430 + }, + { + "epoch": 0.5330294239535848, + "grad_norm": 0.2633078992366791, + "learning_rate": 2.335267302113552e-06, + "loss": 0.5177, + "step": 6431 + }, + { + "epoch": 0.5331123083298798, + "grad_norm": 0.2669292986392975, + "learning_rate": 2.3348528802320765e-06, + "loss": 0.5331, + "step": 6432 + }, + { + "epoch": 0.5331951927061749, + "grad_norm": 0.23168884217739105, + "learning_rate": 2.334438458350601e-06, + "loss": 0.4666, + "step": 6433 + }, + { + "epoch": 0.53327807708247, + "grad_norm": 0.24455620348453522, + "learning_rate": 2.3340240364691258e-06, + "loss": 0.4865, + "step": 6434 + }, + { + "epoch": 0.533360961458765, + "grad_norm": 0.24641798436641693, + "learning_rate": 2.3336096145876506e-06, + "loss": 0.4906, + "step": 6435 + }, + { + "epoch": 0.5334438458350601, + "grad_norm": 0.24181273579597473, + "learning_rate": 2.333195192706175e-06, + "loss": 0.5032, + "step": 6436 + }, + { + "epoch": 0.5335267302113551, + "grad_norm": 0.2510792016983032, + "learning_rate": 2.3327807708246995e-06, + "loss": 0.4825, + "step": 6437 + }, + { + "epoch": 0.5336096145876502, + "grad_norm": 0.2545083463191986, + "learning_rate": 2.3323663489432243e-06, + "loss": 0.4945, + "step": 6438 + }, + { + "epoch": 0.5336924989639453, + "grad_norm": 0.2604309022426605, + "learning_rate": 2.331951927061749e-06, + "loss": 0.5405, + "step": 6439 + }, + { + "epoch": 0.5337753833402403, + "grad_norm": 0.26018404960632324, + "learning_rate": 2.3315375051802736e-06, + "loss": 0.5393, + "step": 6440 + }, + { + "epoch": 0.5338582677165354, + "grad_norm": 0.25183460116386414, + "learning_rate": 2.331123083298798e-06, + "loss": 0.5511, + "step": 6441 + }, + { + "epoch": 0.5339411520928306, + "grad_norm": 0.2673163115978241, + "learning_rate": 2.330708661417323e-06, + "loss": 0.5345, + "step": 6442 + }, + { + "epoch": 0.5340240364691256, + "grad_norm": 0.26173970103263855, + "learning_rate": 2.3302942395358478e-06, + "loss": 0.5054, + "step": 6443 + }, + { + "epoch": 0.5341069208454207, + "grad_norm": 0.2548843324184418, + "learning_rate": 2.3298798176543726e-06, + "loss": 0.5263, + "step": 6444 + }, + { + "epoch": 0.5341898052217157, + "grad_norm": 0.26117509603500366, + "learning_rate": 2.329465395772897e-06, + "loss": 0.5534, + "step": 6445 + }, + { + "epoch": 0.5342726895980108, + "grad_norm": 0.23365429043769836, + "learning_rate": 2.3290509738914215e-06, + "loss": 0.4932, + "step": 6446 + }, + { + "epoch": 0.5343555739743059, + "grad_norm": 0.24520251154899597, + "learning_rate": 2.3286365520099463e-06, + "loss": 0.4899, + "step": 6447 + }, + { + "epoch": 0.5344384583506009, + "grad_norm": 0.2664850950241089, + "learning_rate": 2.328222130128471e-06, + "loss": 0.5089, + "step": 6448 + }, + { + "epoch": 0.534521342726896, + "grad_norm": 0.2607555687427521, + "learning_rate": 2.3278077082469956e-06, + "loss": 0.5011, + "step": 6449 + }, + { + "epoch": 0.534604227103191, + "grad_norm": 0.2628839313983917, + "learning_rate": 2.32739328636552e-06, + "loss": 0.5474, + "step": 6450 + }, + { + "epoch": 0.5346871114794861, + "grad_norm": 0.2595101594924927, + "learning_rate": 2.326978864484045e-06, + "loss": 0.5166, + "step": 6451 + }, + { + "epoch": 0.5347699958557812, + "grad_norm": 0.24455028772354126, + "learning_rate": 2.3265644426025698e-06, + "loss": 0.4865, + "step": 6452 + }, + { + "epoch": 0.5348528802320762, + "grad_norm": 0.264200896024704, + "learning_rate": 2.326150020721094e-06, + "loss": 0.4932, + "step": 6453 + }, + { + "epoch": 0.5349357646083713, + "grad_norm": 0.2484200894832611, + "learning_rate": 2.3257355988396186e-06, + "loss": 0.5028, + "step": 6454 + }, + { + "epoch": 0.5350186489846664, + "grad_norm": 0.24498198926448822, + "learning_rate": 2.3253211769581435e-06, + "loss": 0.5351, + "step": 6455 + }, + { + "epoch": 0.5351015333609614, + "grad_norm": 0.29188504815101624, + "learning_rate": 2.3249067550766684e-06, + "loss": 0.5199, + "step": 6456 + }, + { + "epoch": 0.5351844177372566, + "grad_norm": 0.24325735867023468, + "learning_rate": 2.3244923331951928e-06, + "loss": 0.4893, + "step": 6457 + }, + { + "epoch": 0.5352673021135516, + "grad_norm": 0.26852038502693176, + "learning_rate": 2.3240779113137176e-06, + "loss": 0.56, + "step": 6458 + }, + { + "epoch": 0.5353501864898467, + "grad_norm": 0.2678771913051605, + "learning_rate": 2.323663489432242e-06, + "loss": 0.5284, + "step": 6459 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 0.2496553510427475, + "learning_rate": 2.323249067550767e-06, + "loss": 0.5178, + "step": 6460 + }, + { + "epoch": 0.5355159552424368, + "grad_norm": 0.2460024505853653, + "learning_rate": 2.3228346456692914e-06, + "loss": 0.4677, + "step": 6461 + }, + { + "epoch": 0.5355988396187319, + "grad_norm": 0.24862638115882874, + "learning_rate": 2.3224202237878162e-06, + "loss": 0.5248, + "step": 6462 + }, + { + "epoch": 0.5356817239950269, + "grad_norm": 0.27347055077552795, + "learning_rate": 2.3220058019063406e-06, + "loss": 0.5472, + "step": 6463 + }, + { + "epoch": 0.535764608371322, + "grad_norm": 0.25867006182670593, + "learning_rate": 2.3215913800248655e-06, + "loss": 0.5392, + "step": 6464 + }, + { + "epoch": 0.5358474927476171, + "grad_norm": 0.26617738604545593, + "learning_rate": 2.3211769581433904e-06, + "loss": 0.5248, + "step": 6465 + }, + { + "epoch": 0.5359303771239121, + "grad_norm": 0.23118197917938232, + "learning_rate": 2.3207625362619148e-06, + "loss": 0.4788, + "step": 6466 + }, + { + "epoch": 0.5360132615002072, + "grad_norm": 0.24537009000778198, + "learning_rate": 2.3203481143804396e-06, + "loss": 0.5117, + "step": 6467 + }, + { + "epoch": 0.5360961458765022, + "grad_norm": 0.2456035017967224, + "learning_rate": 2.319933692498964e-06, + "loss": 0.5134, + "step": 6468 + }, + { + "epoch": 0.5361790302527973, + "grad_norm": 0.24921415746212006, + "learning_rate": 2.319519270617489e-06, + "loss": 0.5161, + "step": 6469 + }, + { + "epoch": 0.5362619146290925, + "grad_norm": 0.2748633623123169, + "learning_rate": 2.3191048487360134e-06, + "loss": 0.5079, + "step": 6470 + }, + { + "epoch": 0.5363447990053875, + "grad_norm": 0.24144750833511353, + "learning_rate": 2.3186904268545382e-06, + "loss": 0.5273, + "step": 6471 + }, + { + "epoch": 0.5364276833816826, + "grad_norm": 0.2632780075073242, + "learning_rate": 2.3182760049730626e-06, + "loss": 0.48, + "step": 6472 + }, + { + "epoch": 0.5365105677579777, + "grad_norm": 0.23515203595161438, + "learning_rate": 2.3178615830915875e-06, + "loss": 0.4925, + "step": 6473 + }, + { + "epoch": 0.5365934521342727, + "grad_norm": 0.25215473771095276, + "learning_rate": 2.317447161210112e-06, + "loss": 0.5063, + "step": 6474 + }, + { + "epoch": 0.5366763365105678, + "grad_norm": 0.2778559625148773, + "learning_rate": 2.317032739328637e-06, + "loss": 0.5331, + "step": 6475 + }, + { + "epoch": 0.5367592208868628, + "grad_norm": 0.2609468102455139, + "learning_rate": 2.3166183174471616e-06, + "loss": 0.5307, + "step": 6476 + }, + { + "epoch": 0.5368421052631579, + "grad_norm": 0.27185723185539246, + "learning_rate": 2.316203895565686e-06, + "loss": 0.5106, + "step": 6477 + }, + { + "epoch": 0.536924989639453, + "grad_norm": 0.2539556920528412, + "learning_rate": 2.3157894736842105e-06, + "loss": 0.5242, + "step": 6478 + }, + { + "epoch": 0.537007874015748, + "grad_norm": 0.24074921011924744, + "learning_rate": 2.3153750518027354e-06, + "loss": 0.4863, + "step": 6479 + }, + { + "epoch": 0.5370907583920431, + "grad_norm": 0.23938608169555664, + "learning_rate": 2.3149606299212602e-06, + "loss": 0.4746, + "step": 6480 + }, + { + "epoch": 0.5371736427683381, + "grad_norm": 0.2394101321697235, + "learning_rate": 2.3145462080397847e-06, + "loss": 0.5246, + "step": 6481 + }, + { + "epoch": 0.5372565271446332, + "grad_norm": 0.23715274035930634, + "learning_rate": 2.3141317861583095e-06, + "loss": 0.5186, + "step": 6482 + }, + { + "epoch": 0.5373394115209283, + "grad_norm": 0.2653084099292755, + "learning_rate": 2.313717364276834e-06, + "loss": 0.504, + "step": 6483 + }, + { + "epoch": 0.5374222958972233, + "grad_norm": 0.2591690123081207, + "learning_rate": 2.313302942395359e-06, + "loss": 0.4913, + "step": 6484 + }, + { + "epoch": 0.5375051802735185, + "grad_norm": 0.2548053562641144, + "learning_rate": 2.3128885205138832e-06, + "loss": 0.5564, + "step": 6485 + }, + { + "epoch": 0.5375880646498135, + "grad_norm": 0.29982855916023254, + "learning_rate": 2.312474098632408e-06, + "loss": 0.5508, + "step": 6486 + }, + { + "epoch": 0.5376709490261086, + "grad_norm": 0.24524138867855072, + "learning_rate": 2.3120596767509325e-06, + "loss": 0.5017, + "step": 6487 + }, + { + "epoch": 0.5377538334024037, + "grad_norm": 0.2610072195529938, + "learning_rate": 2.3116452548694574e-06, + "loss": 0.5486, + "step": 6488 + }, + { + "epoch": 0.5378367177786987, + "grad_norm": 0.27642256021499634, + "learning_rate": 2.3112308329879822e-06, + "loss": 0.5387, + "step": 6489 + }, + { + "epoch": 0.5379196021549938, + "grad_norm": 0.2582828402519226, + "learning_rate": 2.3108164111065067e-06, + "loss": 0.485, + "step": 6490 + }, + { + "epoch": 0.5380024865312889, + "grad_norm": 0.23821255564689636, + "learning_rate": 2.310401989225031e-06, + "loss": 0.4814, + "step": 6491 + }, + { + "epoch": 0.5380853709075839, + "grad_norm": 0.2443426549434662, + "learning_rate": 2.309987567343556e-06, + "loss": 0.4967, + "step": 6492 + }, + { + "epoch": 0.538168255283879, + "grad_norm": 0.2343578040599823, + "learning_rate": 2.309573145462081e-06, + "loss": 0.5049, + "step": 6493 + }, + { + "epoch": 0.538251139660174, + "grad_norm": 0.26309993863105774, + "learning_rate": 2.3091587235806052e-06, + "loss": 0.5612, + "step": 6494 + }, + { + "epoch": 0.5383340240364691, + "grad_norm": 0.2716759741306305, + "learning_rate": 2.3087443016991297e-06, + "loss": 0.5161, + "step": 6495 + }, + { + "epoch": 0.5384169084127642, + "grad_norm": 0.2555462718009949, + "learning_rate": 2.3083298798176545e-06, + "loss": 0.5476, + "step": 6496 + }, + { + "epoch": 0.5384997927890592, + "grad_norm": 0.23208077251911163, + "learning_rate": 2.3079154579361794e-06, + "loss": 0.4671, + "step": 6497 + }, + { + "epoch": 0.5385826771653544, + "grad_norm": 0.25989577174186707, + "learning_rate": 2.307501036054704e-06, + "loss": 0.5556, + "step": 6498 + }, + { + "epoch": 0.5386655615416494, + "grad_norm": 0.24545849859714508, + "learning_rate": 2.3070866141732282e-06, + "loss": 0.4958, + "step": 6499 + }, + { + "epoch": 0.5387484459179445, + "grad_norm": 0.2332368940114975, + "learning_rate": 2.306672192291753e-06, + "loss": 0.5037, + "step": 6500 + }, + { + "epoch": 0.5388313302942396, + "grad_norm": 0.247676819562912, + "learning_rate": 2.306257770410278e-06, + "loss": 0.5639, + "step": 6501 + }, + { + "epoch": 0.5389142146705346, + "grad_norm": 0.2543061375617981, + "learning_rate": 2.305843348528803e-06, + "loss": 0.5101, + "step": 6502 + }, + { + "epoch": 0.5389970990468297, + "grad_norm": 0.23943740129470825, + "learning_rate": 2.3054289266473272e-06, + "loss": 0.5217, + "step": 6503 + }, + { + "epoch": 0.5390799834231247, + "grad_norm": 0.25374290347099304, + "learning_rate": 2.3050145047658517e-06, + "loss": 0.5283, + "step": 6504 + }, + { + "epoch": 0.5391628677994198, + "grad_norm": 0.25433212518692017, + "learning_rate": 2.3046000828843765e-06, + "loss": 0.5349, + "step": 6505 + }, + { + "epoch": 0.5392457521757149, + "grad_norm": 0.2528954744338989, + "learning_rate": 2.3041856610029014e-06, + "loss": 0.5392, + "step": 6506 + }, + { + "epoch": 0.5393286365520099, + "grad_norm": 0.257069855928421, + "learning_rate": 2.303771239121426e-06, + "loss": 0.5338, + "step": 6507 + }, + { + "epoch": 0.539411520928305, + "grad_norm": 0.25785189867019653, + "learning_rate": 2.3033568172399502e-06, + "loss": 0.5266, + "step": 6508 + }, + { + "epoch": 0.5394944053046001, + "grad_norm": 0.238008052110672, + "learning_rate": 2.302942395358475e-06, + "loss": 0.516, + "step": 6509 + }, + { + "epoch": 0.5395772896808951, + "grad_norm": 0.26080599427223206, + "learning_rate": 2.302527973477e-06, + "loss": 0.5406, + "step": 6510 + }, + { + "epoch": 0.5396601740571902, + "grad_norm": 0.2410610169172287, + "learning_rate": 2.3021135515955244e-06, + "loss": 0.4855, + "step": 6511 + }, + { + "epoch": 0.5397430584334852, + "grad_norm": 0.23986878991127014, + "learning_rate": 2.301699129714049e-06, + "loss": 0.4945, + "step": 6512 + }, + { + "epoch": 0.5398259428097804, + "grad_norm": 0.2677222788333893, + "learning_rate": 2.3012847078325737e-06, + "loss": 0.5392, + "step": 6513 + }, + { + "epoch": 0.5399088271860755, + "grad_norm": 0.27108234167099, + "learning_rate": 2.3008702859510985e-06, + "loss": 0.5356, + "step": 6514 + }, + { + "epoch": 0.5399917115623705, + "grad_norm": 0.2542736232280731, + "learning_rate": 2.300455864069623e-06, + "loss": 0.5443, + "step": 6515 + }, + { + "epoch": 0.5400745959386656, + "grad_norm": 0.26492056250572205, + "learning_rate": 2.300041442188148e-06, + "loss": 0.5342, + "step": 6516 + }, + { + "epoch": 0.5401574803149606, + "grad_norm": 0.257731556892395, + "learning_rate": 2.2996270203066722e-06, + "loss": 0.5157, + "step": 6517 + }, + { + "epoch": 0.5402403646912557, + "grad_norm": 0.22746878862380981, + "learning_rate": 2.299212598425197e-06, + "loss": 0.4695, + "step": 6518 + }, + { + "epoch": 0.5403232490675508, + "grad_norm": 0.2542412579059601, + "learning_rate": 2.2987981765437215e-06, + "loss": 0.5346, + "step": 6519 + }, + { + "epoch": 0.5404061334438458, + "grad_norm": 0.2781536877155304, + "learning_rate": 2.2983837546622464e-06, + "loss": 0.5262, + "step": 6520 + }, + { + "epoch": 0.5404890178201409, + "grad_norm": 0.28651830554008484, + "learning_rate": 2.297969332780771e-06, + "loss": 0.4836, + "step": 6521 + }, + { + "epoch": 0.540571902196436, + "grad_norm": 0.26757949590682983, + "learning_rate": 2.2975549108992957e-06, + "loss": 0.5287, + "step": 6522 + }, + { + "epoch": 0.540654786572731, + "grad_norm": 0.2531764805316925, + "learning_rate": 2.2971404890178205e-06, + "loss": 0.5461, + "step": 6523 + }, + { + "epoch": 0.5407376709490261, + "grad_norm": 0.2753967344760895, + "learning_rate": 2.296726067136345e-06, + "loss": 0.5319, + "step": 6524 + }, + { + "epoch": 0.5408205553253211, + "grad_norm": 0.2520001232624054, + "learning_rate": 2.29631164525487e-06, + "loss": 0.5092, + "step": 6525 + }, + { + "epoch": 0.5409034397016163, + "grad_norm": 0.27723753452301025, + "learning_rate": 2.2958972233733942e-06, + "loss": 0.5449, + "step": 6526 + }, + { + "epoch": 0.5409863240779114, + "grad_norm": 0.25658342242240906, + "learning_rate": 2.295482801491919e-06, + "loss": 0.481, + "step": 6527 + }, + { + "epoch": 0.5410692084542064, + "grad_norm": 0.2535136938095093, + "learning_rate": 2.2950683796104435e-06, + "loss": 0.5409, + "step": 6528 + }, + { + "epoch": 0.5411520928305015, + "grad_norm": 0.24627898633480072, + "learning_rate": 2.2946539577289684e-06, + "loss": 0.5135, + "step": 6529 + }, + { + "epoch": 0.5412349772067965, + "grad_norm": 0.2702046036720276, + "learning_rate": 2.294239535847493e-06, + "loss": 0.5121, + "step": 6530 + }, + { + "epoch": 0.5413178615830916, + "grad_norm": 0.24573637545108795, + "learning_rate": 2.2938251139660177e-06, + "loss": 0.5192, + "step": 6531 + }, + { + "epoch": 0.5414007459593867, + "grad_norm": 0.2643126845359802, + "learning_rate": 2.293410692084542e-06, + "loss": 0.5403, + "step": 6532 + }, + { + "epoch": 0.5414836303356817, + "grad_norm": 0.24005448818206787, + "learning_rate": 2.292996270203067e-06, + "loss": 0.4931, + "step": 6533 + }, + { + "epoch": 0.5415665147119768, + "grad_norm": 0.26159682869911194, + "learning_rate": 2.292581848321592e-06, + "loss": 0.5337, + "step": 6534 + }, + { + "epoch": 0.5416493990882718, + "grad_norm": 0.2550058364868164, + "learning_rate": 2.2921674264401163e-06, + "loss": 0.5121, + "step": 6535 + }, + { + "epoch": 0.5417322834645669, + "grad_norm": 0.2611992359161377, + "learning_rate": 2.2917530045586407e-06, + "loss": 0.5435, + "step": 6536 + }, + { + "epoch": 0.541815167840862, + "grad_norm": 0.2520197927951813, + "learning_rate": 2.2913385826771655e-06, + "loss": 0.5087, + "step": 6537 + }, + { + "epoch": 0.541898052217157, + "grad_norm": 0.2501712441444397, + "learning_rate": 2.2909241607956904e-06, + "loss": 0.519, + "step": 6538 + }, + { + "epoch": 0.5419809365934521, + "grad_norm": 0.23599687218666077, + "learning_rate": 2.290509738914215e-06, + "loss": 0.4622, + "step": 6539 + }, + { + "epoch": 0.5420638209697473, + "grad_norm": 0.2405247837305069, + "learning_rate": 2.2900953170327393e-06, + "loss": 0.4976, + "step": 6540 + }, + { + "epoch": 0.5421467053460423, + "grad_norm": 0.24201035499572754, + "learning_rate": 2.289680895151264e-06, + "loss": 0.496, + "step": 6541 + }, + { + "epoch": 0.5422295897223374, + "grad_norm": 0.2473861277103424, + "learning_rate": 2.289266473269789e-06, + "loss": 0.515, + "step": 6542 + }, + { + "epoch": 0.5423124740986324, + "grad_norm": 0.23862788081169128, + "learning_rate": 2.288852051388314e-06, + "loss": 0.4721, + "step": 6543 + }, + { + "epoch": 0.5423953584749275, + "grad_norm": 0.24565549194812775, + "learning_rate": 2.2884376295068383e-06, + "loss": 0.4845, + "step": 6544 + }, + { + "epoch": 0.5424782428512226, + "grad_norm": 0.2502020299434662, + "learning_rate": 2.2880232076253627e-06, + "loss": 0.5084, + "step": 6545 + }, + { + "epoch": 0.5425611272275176, + "grad_norm": 0.2485208511352539, + "learning_rate": 2.2876087857438875e-06, + "loss": 0.5145, + "step": 6546 + }, + { + "epoch": 0.5426440116038127, + "grad_norm": 0.26502999663352966, + "learning_rate": 2.2871943638624124e-06, + "loss": 0.5167, + "step": 6547 + }, + { + "epoch": 0.5427268959801077, + "grad_norm": 0.2360478788614273, + "learning_rate": 2.286779941980937e-06, + "loss": 0.4964, + "step": 6548 + }, + { + "epoch": 0.5428097803564028, + "grad_norm": 0.2452588826417923, + "learning_rate": 2.2863655200994613e-06, + "loss": 0.4834, + "step": 6549 + }, + { + "epoch": 0.5428926647326979, + "grad_norm": 0.257802814245224, + "learning_rate": 2.285951098217986e-06, + "loss": 0.5477, + "step": 6550 + }, + { + "epoch": 0.5429755491089929, + "grad_norm": 0.2413550317287445, + "learning_rate": 2.285536676336511e-06, + "loss": 0.497, + "step": 6551 + }, + { + "epoch": 0.543058433485288, + "grad_norm": 0.270884245634079, + "learning_rate": 2.2851222544550354e-06, + "loss": 0.5686, + "step": 6552 + }, + { + "epoch": 0.543141317861583, + "grad_norm": 0.25023066997528076, + "learning_rate": 2.28470783257356e-06, + "loss": 0.4792, + "step": 6553 + }, + { + "epoch": 0.5432242022378782, + "grad_norm": 0.25632762908935547, + "learning_rate": 2.2842934106920847e-06, + "loss": 0.5297, + "step": 6554 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 0.2588394284248352, + "learning_rate": 2.2838789888106095e-06, + "loss": 0.4911, + "step": 6555 + }, + { + "epoch": 0.5433899709904683, + "grad_norm": 0.24646669626235962, + "learning_rate": 2.283464566929134e-06, + "loss": 0.5151, + "step": 6556 + }, + { + "epoch": 0.5434728553667634, + "grad_norm": 0.2529126703739166, + "learning_rate": 2.2830501450476584e-06, + "loss": 0.4841, + "step": 6557 + }, + { + "epoch": 0.5435557397430585, + "grad_norm": 0.27409106492996216, + "learning_rate": 2.2826357231661833e-06, + "loss": 0.4995, + "step": 6558 + }, + { + "epoch": 0.5436386241193535, + "grad_norm": 0.2473527044057846, + "learning_rate": 2.282221301284708e-06, + "loss": 0.5134, + "step": 6559 + }, + { + "epoch": 0.5437215084956486, + "grad_norm": 0.25761258602142334, + "learning_rate": 2.2818068794032326e-06, + "loss": 0.5179, + "step": 6560 + }, + { + "epoch": 0.5438043928719436, + "grad_norm": 0.25776398181915283, + "learning_rate": 2.2813924575217574e-06, + "loss": 0.534, + "step": 6561 + }, + { + "epoch": 0.5438872772482387, + "grad_norm": 0.2497423142194748, + "learning_rate": 2.280978035640282e-06, + "loss": 0.5002, + "step": 6562 + }, + { + "epoch": 0.5439701616245338, + "grad_norm": 0.24055807292461395, + "learning_rate": 2.2805636137588067e-06, + "loss": 0.5043, + "step": 6563 + }, + { + "epoch": 0.5440530460008288, + "grad_norm": 0.24251294136047363, + "learning_rate": 2.2801491918773316e-06, + "loss": 0.4615, + "step": 6564 + }, + { + "epoch": 0.5441359303771239, + "grad_norm": 0.24967946112155914, + "learning_rate": 2.279734769995856e-06, + "loss": 0.4894, + "step": 6565 + }, + { + "epoch": 0.5442188147534189, + "grad_norm": 0.23459281027317047, + "learning_rate": 2.2793203481143804e-06, + "loss": 0.4954, + "step": 6566 + }, + { + "epoch": 0.544301699129714, + "grad_norm": 0.25285083055496216, + "learning_rate": 2.2789059262329053e-06, + "loss": 0.5251, + "step": 6567 + }, + { + "epoch": 0.5443845835060092, + "grad_norm": 0.26058095693588257, + "learning_rate": 2.27849150435143e-06, + "loss": 0.531, + "step": 6568 + }, + { + "epoch": 0.5444674678823042, + "grad_norm": 0.25042206048965454, + "learning_rate": 2.2780770824699546e-06, + "loss": 0.5193, + "step": 6569 + }, + { + "epoch": 0.5445503522585993, + "grad_norm": 0.2511672079563141, + "learning_rate": 2.277662660588479e-06, + "loss": 0.5136, + "step": 6570 + }, + { + "epoch": 0.5446332366348943, + "grad_norm": 0.2581876516342163, + "learning_rate": 2.277248238707004e-06, + "loss": 0.5468, + "step": 6571 + }, + { + "epoch": 0.5447161210111894, + "grad_norm": 0.2791345417499542, + "learning_rate": 2.2768338168255287e-06, + "loss": 0.4915, + "step": 6572 + }, + { + "epoch": 0.5447990053874845, + "grad_norm": 0.25077468156814575, + "learning_rate": 2.276419394944053e-06, + "loss": 0.5306, + "step": 6573 + }, + { + "epoch": 0.5448818897637795, + "grad_norm": 0.23396171629428864, + "learning_rate": 2.276004973062578e-06, + "loss": 0.4901, + "step": 6574 + }, + { + "epoch": 0.5449647741400746, + "grad_norm": 0.23536935448646545, + "learning_rate": 2.2755905511811024e-06, + "loss": 0.4595, + "step": 6575 + }, + { + "epoch": 0.5450476585163697, + "grad_norm": 0.24590708315372467, + "learning_rate": 2.2751761292996273e-06, + "loss": 0.5205, + "step": 6576 + }, + { + "epoch": 0.5451305428926647, + "grad_norm": 0.2386498749256134, + "learning_rate": 2.2747617074181517e-06, + "loss": 0.5095, + "step": 6577 + }, + { + "epoch": 0.5452134272689598, + "grad_norm": 0.25398457050323486, + "learning_rate": 2.2743472855366766e-06, + "loss": 0.5088, + "step": 6578 + }, + { + "epoch": 0.5452963116452548, + "grad_norm": 0.24864931404590607, + "learning_rate": 2.273932863655201e-06, + "loss": 0.4979, + "step": 6579 + }, + { + "epoch": 0.5453791960215499, + "grad_norm": 0.2437008023262024, + "learning_rate": 2.273518441773726e-06, + "loss": 0.4861, + "step": 6580 + }, + { + "epoch": 0.545462080397845, + "grad_norm": 0.25029250979423523, + "learning_rate": 2.2731040198922507e-06, + "loss": 0.5352, + "step": 6581 + }, + { + "epoch": 0.54554496477414, + "grad_norm": 0.25604164600372314, + "learning_rate": 2.272689598010775e-06, + "loss": 0.5164, + "step": 6582 + }, + { + "epoch": 0.5456278491504352, + "grad_norm": 0.26274481415748596, + "learning_rate": 2.2722751761293e-06, + "loss": 0.5633, + "step": 6583 + }, + { + "epoch": 0.5457107335267302, + "grad_norm": 0.236623153090477, + "learning_rate": 2.2718607542478244e-06, + "loss": 0.4917, + "step": 6584 + }, + { + "epoch": 0.5457936179030253, + "grad_norm": 0.2482939511537552, + "learning_rate": 2.2714463323663493e-06, + "loss": 0.4812, + "step": 6585 + }, + { + "epoch": 0.5458765022793204, + "grad_norm": 0.2705908715724945, + "learning_rate": 2.2710319104848737e-06, + "loss": 0.4911, + "step": 6586 + }, + { + "epoch": 0.5459593866556154, + "grad_norm": 0.26237836480140686, + "learning_rate": 2.2706174886033986e-06, + "loss": 0.5014, + "step": 6587 + }, + { + "epoch": 0.5460422710319105, + "grad_norm": 0.2577623426914215, + "learning_rate": 2.270203066721923e-06, + "loss": 0.5179, + "step": 6588 + }, + { + "epoch": 0.5461251554082055, + "grad_norm": 0.2478286176919937, + "learning_rate": 2.269788644840448e-06, + "loss": 0.5345, + "step": 6589 + }, + { + "epoch": 0.5462080397845006, + "grad_norm": 0.25043806433677673, + "learning_rate": 2.2693742229589723e-06, + "loss": 0.5197, + "step": 6590 + }, + { + "epoch": 0.5462909241607957, + "grad_norm": 0.24193818867206573, + "learning_rate": 2.268959801077497e-06, + "loss": 0.5384, + "step": 6591 + }, + { + "epoch": 0.5463738085370907, + "grad_norm": 0.2517133057117462, + "learning_rate": 2.268545379196022e-06, + "loss": 0.524, + "step": 6592 + }, + { + "epoch": 0.5464566929133858, + "grad_norm": 0.267214298248291, + "learning_rate": 2.2681309573145464e-06, + "loss": 0.5175, + "step": 6593 + }, + { + "epoch": 0.546539577289681, + "grad_norm": 0.25527647137641907, + "learning_rate": 2.267716535433071e-06, + "loss": 0.5452, + "step": 6594 + }, + { + "epoch": 0.546622461665976, + "grad_norm": 0.25162872672080994, + "learning_rate": 2.2673021135515957e-06, + "loss": 0.5411, + "step": 6595 + }, + { + "epoch": 0.5467053460422711, + "grad_norm": 0.24190975725650787, + "learning_rate": 2.2668876916701206e-06, + "loss": 0.4988, + "step": 6596 + }, + { + "epoch": 0.5467882304185661, + "grad_norm": 0.2561761736869812, + "learning_rate": 2.266473269788645e-06, + "loss": 0.5114, + "step": 6597 + }, + { + "epoch": 0.5468711147948612, + "grad_norm": 0.25706586241722107, + "learning_rate": 2.2660588479071694e-06, + "loss": 0.5071, + "step": 6598 + }, + { + "epoch": 0.5469539991711563, + "grad_norm": 0.24239616096019745, + "learning_rate": 2.2656444260256943e-06, + "loss": 0.4938, + "step": 6599 + }, + { + "epoch": 0.5470368835474513, + "grad_norm": 0.2396056205034256, + "learning_rate": 2.265230004144219e-06, + "loss": 0.4685, + "step": 6600 + }, + { + "epoch": 0.5471197679237464, + "grad_norm": 0.2696729898452759, + "learning_rate": 2.2648155822627436e-06, + "loss": 0.524, + "step": 6601 + }, + { + "epoch": 0.5472026523000414, + "grad_norm": 0.26619845628738403, + "learning_rate": 2.2644011603812684e-06, + "loss": 0.524, + "step": 6602 + }, + { + "epoch": 0.5472855366763365, + "grad_norm": 0.2642883360385895, + "learning_rate": 2.263986738499793e-06, + "loss": 0.5186, + "step": 6603 + }, + { + "epoch": 0.5473684210526316, + "grad_norm": 0.24857720732688904, + "learning_rate": 2.2635723166183177e-06, + "loss": 0.5034, + "step": 6604 + }, + { + "epoch": 0.5474513054289266, + "grad_norm": 0.2582820951938629, + "learning_rate": 2.2631578947368426e-06, + "loss": 0.5408, + "step": 6605 + }, + { + "epoch": 0.5475341898052217, + "grad_norm": 0.24397948384284973, + "learning_rate": 2.262743472855367e-06, + "loss": 0.5244, + "step": 6606 + }, + { + "epoch": 0.5476170741815168, + "grad_norm": 0.24887827038764954, + "learning_rate": 2.2623290509738914e-06, + "loss": 0.4728, + "step": 6607 + }, + { + "epoch": 0.5476999585578118, + "grad_norm": 0.26025259494781494, + "learning_rate": 2.2619146290924163e-06, + "loss": 0.5362, + "step": 6608 + }, + { + "epoch": 0.547782842934107, + "grad_norm": 0.24786938726902008, + "learning_rate": 2.261500207210941e-06, + "loss": 0.4969, + "step": 6609 + }, + { + "epoch": 0.547865727310402, + "grad_norm": 0.23713509738445282, + "learning_rate": 2.2610857853294656e-06, + "loss": 0.5122, + "step": 6610 + }, + { + "epoch": 0.5479486116866971, + "grad_norm": 0.25092899799346924, + "learning_rate": 2.26067136344799e-06, + "loss": 0.5538, + "step": 6611 + }, + { + "epoch": 0.5480314960629922, + "grad_norm": 0.25763678550720215, + "learning_rate": 2.260256941566515e-06, + "loss": 0.4548, + "step": 6612 + }, + { + "epoch": 0.5481143804392872, + "grad_norm": 0.23442333936691284, + "learning_rate": 2.2598425196850397e-06, + "loss": 0.4788, + "step": 6613 + }, + { + "epoch": 0.5481972648155823, + "grad_norm": 0.2464684247970581, + "learning_rate": 2.259428097803564e-06, + "loss": 0.5208, + "step": 6614 + }, + { + "epoch": 0.5482801491918773, + "grad_norm": 0.26804471015930176, + "learning_rate": 2.2590136759220886e-06, + "loss": 0.5412, + "step": 6615 + }, + { + "epoch": 0.5483630335681724, + "grad_norm": 0.2600443661212921, + "learning_rate": 2.2585992540406134e-06, + "loss": 0.5618, + "step": 6616 + }, + { + "epoch": 0.5484459179444675, + "grad_norm": 0.24212680757045746, + "learning_rate": 2.2581848321591383e-06, + "loss": 0.4882, + "step": 6617 + }, + { + "epoch": 0.5485288023207625, + "grad_norm": 0.26214906573295593, + "learning_rate": 2.2577704102776627e-06, + "loss": 0.5362, + "step": 6618 + }, + { + "epoch": 0.5486116866970576, + "grad_norm": 0.24785733222961426, + "learning_rate": 2.2573559883961876e-06, + "loss": 0.5145, + "step": 6619 + }, + { + "epoch": 0.5486945710733526, + "grad_norm": 0.24866034090518951, + "learning_rate": 2.256941566514712e-06, + "loss": 0.5016, + "step": 6620 + }, + { + "epoch": 0.5487774554496477, + "grad_norm": 0.2405577450990677, + "learning_rate": 2.256527144633237e-06, + "loss": 0.4807, + "step": 6621 + }, + { + "epoch": 0.5488603398259428, + "grad_norm": 0.23233672976493835, + "learning_rate": 2.2561127227517617e-06, + "loss": 0.462, + "step": 6622 + }, + { + "epoch": 0.5489432242022378, + "grad_norm": 0.24285991489887238, + "learning_rate": 2.255698300870286e-06, + "loss": 0.5396, + "step": 6623 + }, + { + "epoch": 0.549026108578533, + "grad_norm": 0.24776673316955566, + "learning_rate": 2.2552838789888106e-06, + "loss": 0.5123, + "step": 6624 + }, + { + "epoch": 0.5491089929548281, + "grad_norm": 0.25986745953559875, + "learning_rate": 2.2548694571073354e-06, + "loss": 0.5201, + "step": 6625 + }, + { + "epoch": 0.5491918773311231, + "grad_norm": 0.26079463958740234, + "learning_rate": 2.2544550352258603e-06, + "loss": 0.514, + "step": 6626 + }, + { + "epoch": 0.5492747617074182, + "grad_norm": 0.2788945734500885, + "learning_rate": 2.2540406133443847e-06, + "loss": 0.5762, + "step": 6627 + }, + { + "epoch": 0.5493576460837132, + "grad_norm": 0.26654958724975586, + "learning_rate": 2.2536261914629096e-06, + "loss": 0.5556, + "step": 6628 + }, + { + "epoch": 0.5494405304600083, + "grad_norm": 0.2418651431798935, + "learning_rate": 2.253211769581434e-06, + "loss": 0.5075, + "step": 6629 + }, + { + "epoch": 0.5495234148363034, + "grad_norm": 0.2530539333820343, + "learning_rate": 2.252797347699959e-06, + "loss": 0.5058, + "step": 6630 + }, + { + "epoch": 0.5496062992125984, + "grad_norm": 0.23838302493095398, + "learning_rate": 2.2523829258184833e-06, + "loss": 0.4864, + "step": 6631 + }, + { + "epoch": 0.5496891835888935, + "grad_norm": 0.22494280338287354, + "learning_rate": 2.251968503937008e-06, + "loss": 0.4385, + "step": 6632 + }, + { + "epoch": 0.5497720679651885, + "grad_norm": 0.2569532096385956, + "learning_rate": 2.2515540820555326e-06, + "loss": 0.5069, + "step": 6633 + }, + { + "epoch": 0.5498549523414836, + "grad_norm": 0.264695942401886, + "learning_rate": 2.2511396601740574e-06, + "loss": 0.5575, + "step": 6634 + }, + { + "epoch": 0.5499378367177787, + "grad_norm": 0.23193007707595825, + "learning_rate": 2.250725238292582e-06, + "loss": 0.4887, + "step": 6635 + }, + { + "epoch": 0.5500207210940737, + "grad_norm": 0.2540712058544159, + "learning_rate": 2.2503108164111067e-06, + "loss": 0.5206, + "step": 6636 + }, + { + "epoch": 0.5501036054703689, + "grad_norm": 0.2629149258136749, + "learning_rate": 2.249896394529631e-06, + "loss": 0.4967, + "step": 6637 + }, + { + "epoch": 0.5501864898466639, + "grad_norm": 0.25593721866607666, + "learning_rate": 2.249481972648156e-06, + "loss": 0.4978, + "step": 6638 + }, + { + "epoch": 0.550269374222959, + "grad_norm": 0.2597472667694092, + "learning_rate": 2.2490675507666805e-06, + "loss": 0.5243, + "step": 6639 + }, + { + "epoch": 0.5503522585992541, + "grad_norm": 0.2476757913827896, + "learning_rate": 2.2486531288852053e-06, + "loss": 0.5215, + "step": 6640 + }, + { + "epoch": 0.5504351429755491, + "grad_norm": 0.2531794607639313, + "learning_rate": 2.24823870700373e-06, + "loss": 0.4713, + "step": 6641 + }, + { + "epoch": 0.5505180273518442, + "grad_norm": 0.2635590732097626, + "learning_rate": 2.2478242851222546e-06, + "loss": 0.5172, + "step": 6642 + }, + { + "epoch": 0.5506009117281393, + "grad_norm": 0.25735560059547424, + "learning_rate": 2.2474098632407794e-06, + "loss": 0.5297, + "step": 6643 + }, + { + "epoch": 0.5506837961044343, + "grad_norm": 0.2500936985015869, + "learning_rate": 2.246995441359304e-06, + "loss": 0.5229, + "step": 6644 + }, + { + "epoch": 0.5507666804807294, + "grad_norm": 0.25684598088264465, + "learning_rate": 2.2465810194778287e-06, + "loss": 0.4893, + "step": 6645 + }, + { + "epoch": 0.5508495648570244, + "grad_norm": 0.24249882996082306, + "learning_rate": 2.246166597596353e-06, + "loss": 0.525, + "step": 6646 + }, + { + "epoch": 0.5509324492333195, + "grad_norm": 0.2668401300907135, + "learning_rate": 2.245752175714878e-06, + "loss": 0.4673, + "step": 6647 + }, + { + "epoch": 0.5510153336096146, + "grad_norm": 0.2504402697086334, + "learning_rate": 2.2453377538334025e-06, + "loss": 0.4883, + "step": 6648 + }, + { + "epoch": 0.5510982179859096, + "grad_norm": 0.2634369730949402, + "learning_rate": 2.2449233319519273e-06, + "loss": 0.527, + "step": 6649 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 0.25065648555755615, + "learning_rate": 2.244508910070452e-06, + "loss": 0.5433, + "step": 6650 + }, + { + "epoch": 0.5512639867384997, + "grad_norm": 0.26149362325668335, + "learning_rate": 2.2440944881889766e-06, + "loss": 0.5066, + "step": 6651 + }, + { + "epoch": 0.5513468711147949, + "grad_norm": 0.25088149309158325, + "learning_rate": 2.243680066307501e-06, + "loss": 0.5183, + "step": 6652 + }, + { + "epoch": 0.55142975549109, + "grad_norm": 0.2456623762845993, + "learning_rate": 2.243265644426026e-06, + "loss": 0.5045, + "step": 6653 + }, + { + "epoch": 0.551512639867385, + "grad_norm": 0.253814160823822, + "learning_rate": 2.2428512225445507e-06, + "loss": 0.5167, + "step": 6654 + }, + { + "epoch": 0.5515955242436801, + "grad_norm": 0.24434365332126617, + "learning_rate": 2.242436800663075e-06, + "loss": 0.4992, + "step": 6655 + }, + { + "epoch": 0.5516784086199751, + "grad_norm": 0.295291006565094, + "learning_rate": 2.2420223787815996e-06, + "loss": 0.5401, + "step": 6656 + }, + { + "epoch": 0.5517612929962702, + "grad_norm": 0.2473120242357254, + "learning_rate": 2.2416079569001245e-06, + "loss": 0.5095, + "step": 6657 + }, + { + "epoch": 0.5518441773725653, + "grad_norm": 0.24909988045692444, + "learning_rate": 2.2411935350186493e-06, + "loss": 0.5023, + "step": 6658 + }, + { + "epoch": 0.5519270617488603, + "grad_norm": 0.2601225972175598, + "learning_rate": 2.2407791131371737e-06, + "loss": 0.5399, + "step": 6659 + }, + { + "epoch": 0.5520099461251554, + "grad_norm": 0.2561967372894287, + "learning_rate": 2.240364691255698e-06, + "loss": 0.5043, + "step": 6660 + }, + { + "epoch": 0.5520928305014505, + "grad_norm": 0.2355106621980667, + "learning_rate": 2.239950269374223e-06, + "loss": 0.4644, + "step": 6661 + }, + { + "epoch": 0.5521757148777455, + "grad_norm": 0.22099344432353973, + "learning_rate": 2.239535847492748e-06, + "loss": 0.5116, + "step": 6662 + }, + { + "epoch": 0.5522585992540406, + "grad_norm": 0.24902327358722687, + "learning_rate": 2.2391214256112727e-06, + "loss": 0.5139, + "step": 6663 + }, + { + "epoch": 0.5523414836303356, + "grad_norm": 0.2827756702899933, + "learning_rate": 2.238707003729797e-06, + "loss": 0.5348, + "step": 6664 + }, + { + "epoch": 0.5524243680066308, + "grad_norm": 0.25690025091171265, + "learning_rate": 2.2382925818483216e-06, + "loss": 0.5489, + "step": 6665 + }, + { + "epoch": 0.5525072523829259, + "grad_norm": 0.2549784779548645, + "learning_rate": 2.2378781599668465e-06, + "loss": 0.5218, + "step": 6666 + }, + { + "epoch": 0.5525901367592209, + "grad_norm": 0.2609732747077942, + "learning_rate": 2.2374637380853713e-06, + "loss": 0.5164, + "step": 6667 + }, + { + "epoch": 0.552673021135516, + "grad_norm": 0.25055378675460815, + "learning_rate": 2.2370493162038958e-06, + "loss": 0.5107, + "step": 6668 + }, + { + "epoch": 0.552755905511811, + "grad_norm": 0.2701563537120819, + "learning_rate": 2.23663489432242e-06, + "loss": 0.5603, + "step": 6669 + }, + { + "epoch": 0.5528387898881061, + "grad_norm": 0.26581987738609314, + "learning_rate": 2.236220472440945e-06, + "loss": 0.5285, + "step": 6670 + }, + { + "epoch": 0.5529216742644012, + "grad_norm": 0.27082082629203796, + "learning_rate": 2.23580605055947e-06, + "loss": 0.5079, + "step": 6671 + }, + { + "epoch": 0.5530045586406962, + "grad_norm": 0.2907712757587433, + "learning_rate": 2.2353916286779943e-06, + "loss": 0.5435, + "step": 6672 + }, + { + "epoch": 0.5530874430169913, + "grad_norm": 0.25652575492858887, + "learning_rate": 2.2349772067965188e-06, + "loss": 0.5535, + "step": 6673 + }, + { + "epoch": 0.5531703273932864, + "grad_norm": 0.2547234892845154, + "learning_rate": 2.2345627849150436e-06, + "loss": 0.5273, + "step": 6674 + }, + { + "epoch": 0.5532532117695814, + "grad_norm": 0.27957451343536377, + "learning_rate": 2.2341483630335685e-06, + "loss": 0.5353, + "step": 6675 + }, + { + "epoch": 0.5533360961458765, + "grad_norm": 0.2562316954135895, + "learning_rate": 2.233733941152093e-06, + "loss": 0.525, + "step": 6676 + }, + { + "epoch": 0.5534189805221715, + "grad_norm": 0.27274441719055176, + "learning_rate": 2.2333195192706178e-06, + "loss": 0.5221, + "step": 6677 + }, + { + "epoch": 0.5535018648984666, + "grad_norm": 0.26940515637397766, + "learning_rate": 2.232905097389142e-06, + "loss": 0.5667, + "step": 6678 + }, + { + "epoch": 0.5535847492747618, + "grad_norm": 0.2552153170108795, + "learning_rate": 2.232490675507667e-06, + "loss": 0.5526, + "step": 6679 + }, + { + "epoch": 0.5536676336510568, + "grad_norm": 0.26357507705688477, + "learning_rate": 2.2320762536261915e-06, + "loss": 0.5381, + "step": 6680 + }, + { + "epoch": 0.5537505180273519, + "grad_norm": 0.26159876585006714, + "learning_rate": 2.2316618317447163e-06, + "loss": 0.5006, + "step": 6681 + }, + { + "epoch": 0.5538334024036469, + "grad_norm": 0.24495220184326172, + "learning_rate": 2.2312474098632408e-06, + "loss": 0.5271, + "step": 6682 + }, + { + "epoch": 0.553916286779942, + "grad_norm": 0.24605438113212585, + "learning_rate": 2.2308329879817656e-06, + "loss": 0.4799, + "step": 6683 + }, + { + "epoch": 0.5539991711562371, + "grad_norm": 0.24976782500743866, + "learning_rate": 2.2304185661002905e-06, + "loss": 0.5271, + "step": 6684 + }, + { + "epoch": 0.5540820555325321, + "grad_norm": 0.23402699828147888, + "learning_rate": 2.230004144218815e-06, + "loss": 0.5059, + "step": 6685 + }, + { + "epoch": 0.5541649399088272, + "grad_norm": 0.2661688029766083, + "learning_rate": 2.2295897223373398e-06, + "loss": 0.5013, + "step": 6686 + }, + { + "epoch": 0.5542478242851222, + "grad_norm": 0.23779591917991638, + "learning_rate": 2.229175300455864e-06, + "loss": 0.4784, + "step": 6687 + }, + { + "epoch": 0.5543307086614173, + "grad_norm": 0.2522019147872925, + "learning_rate": 2.228760878574389e-06, + "loss": 0.4968, + "step": 6688 + }, + { + "epoch": 0.5544135930377124, + "grad_norm": 0.2501700520515442, + "learning_rate": 2.2283464566929135e-06, + "loss": 0.48, + "step": 6689 + }, + { + "epoch": 0.5544964774140074, + "grad_norm": 0.25089114904403687, + "learning_rate": 2.2279320348114383e-06, + "loss": 0.5364, + "step": 6690 + }, + { + "epoch": 0.5545793617903025, + "grad_norm": 0.25853243470191956, + "learning_rate": 2.2275176129299628e-06, + "loss": 0.5261, + "step": 6691 + }, + { + "epoch": 0.5546622461665977, + "grad_norm": 0.24486172199249268, + "learning_rate": 2.2271031910484876e-06, + "loss": 0.5133, + "step": 6692 + }, + { + "epoch": 0.5547451305428927, + "grad_norm": 0.2460934966802597, + "learning_rate": 2.226688769167012e-06, + "loss": 0.4998, + "step": 6693 + }, + { + "epoch": 0.5548280149191878, + "grad_norm": 0.2383737862110138, + "learning_rate": 2.226274347285537e-06, + "loss": 0.5279, + "step": 6694 + }, + { + "epoch": 0.5549108992954828, + "grad_norm": 0.34418246150016785, + "learning_rate": 2.2258599254040618e-06, + "loss": 0.5392, + "step": 6695 + }, + { + "epoch": 0.5549937836717779, + "grad_norm": 0.2585318684577942, + "learning_rate": 2.225445503522586e-06, + "loss": 0.5212, + "step": 6696 + }, + { + "epoch": 0.555076668048073, + "grad_norm": 0.26318106055259705, + "learning_rate": 2.2250310816411106e-06, + "loss": 0.5204, + "step": 6697 + }, + { + "epoch": 0.555159552424368, + "grad_norm": 0.2504889965057373, + "learning_rate": 2.2246166597596355e-06, + "loss": 0.4904, + "step": 6698 + }, + { + "epoch": 0.5552424368006631, + "grad_norm": 0.2442457228899002, + "learning_rate": 2.2242022378781603e-06, + "loss": 0.5104, + "step": 6699 + }, + { + "epoch": 0.5553253211769581, + "grad_norm": 0.26003116369247437, + "learning_rate": 2.2237878159966848e-06, + "loss": 0.5088, + "step": 6700 + }, + { + "epoch": 0.5554082055532532, + "grad_norm": 0.2596370577812195, + "learning_rate": 2.2233733941152096e-06, + "loss": 0.4865, + "step": 6701 + }, + { + "epoch": 0.5554910899295483, + "grad_norm": 0.24386043846607208, + "learning_rate": 2.222958972233734e-06, + "loss": 0.5141, + "step": 6702 + }, + { + "epoch": 0.5555739743058433, + "grad_norm": 0.22959300875663757, + "learning_rate": 2.222544550352259e-06, + "loss": 0.5069, + "step": 6703 + }, + { + "epoch": 0.5556568586821384, + "grad_norm": 0.25850918889045715, + "learning_rate": 2.2221301284707838e-06, + "loss": 0.5241, + "step": 6704 + }, + { + "epoch": 0.5557397430584334, + "grad_norm": 0.25218072533607483, + "learning_rate": 2.221715706589308e-06, + "loss": 0.5245, + "step": 6705 + }, + { + "epoch": 0.5558226274347285, + "grad_norm": 0.2580578029155731, + "learning_rate": 2.2213012847078326e-06, + "loss": 0.5377, + "step": 6706 + }, + { + "epoch": 0.5559055118110237, + "grad_norm": 0.24447354674339294, + "learning_rate": 2.2208868628263575e-06, + "loss": 0.5215, + "step": 6707 + }, + { + "epoch": 0.5559883961873187, + "grad_norm": 0.2602294385433197, + "learning_rate": 2.2204724409448823e-06, + "loss": 0.5599, + "step": 6708 + }, + { + "epoch": 0.5560712805636138, + "grad_norm": 0.2630845904350281, + "learning_rate": 2.2200580190634068e-06, + "loss": 0.5037, + "step": 6709 + }, + { + "epoch": 0.5561541649399089, + "grad_norm": 0.26405468583106995, + "learning_rate": 2.219643597181931e-06, + "loss": 0.5089, + "step": 6710 + }, + { + "epoch": 0.5562370493162039, + "grad_norm": 0.2528949975967407, + "learning_rate": 2.219229175300456e-06, + "loss": 0.5051, + "step": 6711 + }, + { + "epoch": 0.556319933692499, + "grad_norm": 0.2421298772096634, + "learning_rate": 2.218814753418981e-06, + "loss": 0.4863, + "step": 6712 + }, + { + "epoch": 0.556402818068794, + "grad_norm": 0.24853503704071045, + "learning_rate": 2.2184003315375053e-06, + "loss": 0.5272, + "step": 6713 + }, + { + "epoch": 0.5564857024450891, + "grad_norm": 0.2547926902770996, + "learning_rate": 2.2179859096560298e-06, + "loss": 0.4924, + "step": 6714 + }, + { + "epoch": 0.5565685868213842, + "grad_norm": 0.25918227434158325, + "learning_rate": 2.2175714877745546e-06, + "loss": 0.5198, + "step": 6715 + }, + { + "epoch": 0.5566514711976792, + "grad_norm": 0.24321429431438446, + "learning_rate": 2.2171570658930795e-06, + "loss": 0.4863, + "step": 6716 + }, + { + "epoch": 0.5567343555739743, + "grad_norm": 0.2434886395931244, + "learning_rate": 2.216742644011604e-06, + "loss": 0.5038, + "step": 6717 + }, + { + "epoch": 0.5568172399502693, + "grad_norm": 0.23510506749153137, + "learning_rate": 2.2163282221301284e-06, + "loss": 0.491, + "step": 6718 + }, + { + "epoch": 0.5569001243265644, + "grad_norm": 0.24735766649246216, + "learning_rate": 2.215913800248653e-06, + "loss": 0.5273, + "step": 6719 + }, + { + "epoch": 0.5569830087028595, + "grad_norm": 0.2380368709564209, + "learning_rate": 2.215499378367178e-06, + "loss": 0.4515, + "step": 6720 + }, + { + "epoch": 0.5570658930791546, + "grad_norm": 0.26386696100234985, + "learning_rate": 2.215084956485703e-06, + "loss": 0.5522, + "step": 6721 + }, + { + "epoch": 0.5571487774554497, + "grad_norm": 0.2418750524520874, + "learning_rate": 2.2146705346042273e-06, + "loss": 0.5271, + "step": 6722 + }, + { + "epoch": 0.5572316618317447, + "grad_norm": 0.24561430513858795, + "learning_rate": 2.2142561127227518e-06, + "loss": 0.4644, + "step": 6723 + }, + { + "epoch": 0.5573145462080398, + "grad_norm": 0.2461211234331131, + "learning_rate": 2.2138416908412766e-06, + "loss": 0.4898, + "step": 6724 + }, + { + "epoch": 0.5573974305843349, + "grad_norm": 0.27242475748062134, + "learning_rate": 2.2134272689598015e-06, + "loss": 0.5524, + "step": 6725 + }, + { + "epoch": 0.5574803149606299, + "grad_norm": 0.24661262333393097, + "learning_rate": 2.213012847078326e-06, + "loss": 0.4969, + "step": 6726 + }, + { + "epoch": 0.557563199336925, + "grad_norm": 0.25293391942977905, + "learning_rate": 2.2125984251968504e-06, + "loss": 0.515, + "step": 6727 + }, + { + "epoch": 0.5576460837132201, + "grad_norm": 0.25692564249038696, + "learning_rate": 2.212184003315375e-06, + "loss": 0.5263, + "step": 6728 + }, + { + "epoch": 0.5577289680895151, + "grad_norm": 0.24624674022197723, + "learning_rate": 2.2117695814339e-06, + "loss": 0.4814, + "step": 6729 + }, + { + "epoch": 0.5578118524658102, + "grad_norm": 0.2488211840391159, + "learning_rate": 2.2113551595524245e-06, + "loss": 0.5084, + "step": 6730 + }, + { + "epoch": 0.5578947368421052, + "grad_norm": 0.25877895951271057, + "learning_rate": 2.210940737670949e-06, + "loss": 0.5167, + "step": 6731 + }, + { + "epoch": 0.5579776212184003, + "grad_norm": 0.247158482670784, + "learning_rate": 2.2105263157894738e-06, + "loss": 0.5086, + "step": 6732 + }, + { + "epoch": 0.5580605055946954, + "grad_norm": 0.2448057234287262, + "learning_rate": 2.2101118939079986e-06, + "loss": 0.5151, + "step": 6733 + }, + { + "epoch": 0.5581433899709904, + "grad_norm": 0.23616185784339905, + "learning_rate": 2.209697472026523e-06, + "loss": 0.4773, + "step": 6734 + }, + { + "epoch": 0.5582262743472856, + "grad_norm": 0.24789026379585266, + "learning_rate": 2.209283050145048e-06, + "loss": 0.4877, + "step": 6735 + }, + { + "epoch": 0.5583091587235806, + "grad_norm": 0.29094547033309937, + "learning_rate": 2.2088686282635724e-06, + "loss": 0.5466, + "step": 6736 + }, + { + "epoch": 0.5583920430998757, + "grad_norm": 0.2526368796825409, + "learning_rate": 2.2084542063820972e-06, + "loss": 0.4986, + "step": 6737 + }, + { + "epoch": 0.5584749274761708, + "grad_norm": 0.2480211853981018, + "learning_rate": 2.2080397845006216e-06, + "loss": 0.533, + "step": 6738 + }, + { + "epoch": 0.5585578118524658, + "grad_norm": 0.2625378966331482, + "learning_rate": 2.2076253626191465e-06, + "loss": 0.463, + "step": 6739 + }, + { + "epoch": 0.5586406962287609, + "grad_norm": 0.28003379702568054, + "learning_rate": 2.207210940737671e-06, + "loss": 0.5793, + "step": 6740 + }, + { + "epoch": 0.5587235806050559, + "grad_norm": 0.25161269307136536, + "learning_rate": 2.206796518856196e-06, + "loss": 0.5552, + "step": 6741 + }, + { + "epoch": 0.558806464981351, + "grad_norm": 0.25574639439582825, + "learning_rate": 2.2063820969747206e-06, + "loss": 0.5923, + "step": 6742 + }, + { + "epoch": 0.5588893493576461, + "grad_norm": 0.34086763858795166, + "learning_rate": 2.205967675093245e-06, + "loss": 0.5248, + "step": 6743 + }, + { + "epoch": 0.5589722337339411, + "grad_norm": 0.26053422689437866, + "learning_rate": 2.20555325321177e-06, + "loss": 0.5168, + "step": 6744 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 0.2463839203119278, + "learning_rate": 2.2051388313302944e-06, + "loss": 0.4523, + "step": 6745 + }, + { + "epoch": 0.5591380024865313, + "grad_norm": 0.25365376472473145, + "learning_rate": 2.2047244094488192e-06, + "loss": 0.4856, + "step": 6746 + }, + { + "epoch": 0.5592208868628263, + "grad_norm": 0.26249104738235474, + "learning_rate": 2.2043099875673436e-06, + "loss": 0.5242, + "step": 6747 + }, + { + "epoch": 0.5593037712391214, + "grad_norm": 0.2581215798854828, + "learning_rate": 2.2038955656858685e-06, + "loss": 0.5165, + "step": 6748 + }, + { + "epoch": 0.5593866556154165, + "grad_norm": 0.2643652558326721, + "learning_rate": 2.203481143804393e-06, + "loss": 0.5134, + "step": 6749 + }, + { + "epoch": 0.5594695399917116, + "grad_norm": 0.2512471079826355, + "learning_rate": 2.203066721922918e-06, + "loss": 0.5135, + "step": 6750 + }, + { + "epoch": 0.5595524243680067, + "grad_norm": 0.24923647940158844, + "learning_rate": 2.2026523000414422e-06, + "loss": 0.5131, + "step": 6751 + }, + { + "epoch": 0.5596353087443017, + "grad_norm": 0.2682216763496399, + "learning_rate": 2.202237878159967e-06, + "loss": 0.4941, + "step": 6752 + }, + { + "epoch": 0.5597181931205968, + "grad_norm": 0.26558762788772583, + "learning_rate": 2.201823456278492e-06, + "loss": 0.5333, + "step": 6753 + }, + { + "epoch": 0.5598010774968918, + "grad_norm": 0.26479247212409973, + "learning_rate": 2.2014090343970164e-06, + "loss": 0.5231, + "step": 6754 + }, + { + "epoch": 0.5598839618731869, + "grad_norm": 0.27479779720306396, + "learning_rate": 2.200994612515541e-06, + "loss": 0.5327, + "step": 6755 + }, + { + "epoch": 0.559966846249482, + "grad_norm": 0.25493863224983215, + "learning_rate": 2.2005801906340657e-06, + "loss": 0.4957, + "step": 6756 + }, + { + "epoch": 0.560049730625777, + "grad_norm": 0.25611984729766846, + "learning_rate": 2.2001657687525905e-06, + "loss": 0.5112, + "step": 6757 + }, + { + "epoch": 0.5601326150020721, + "grad_norm": 0.23430822789669037, + "learning_rate": 2.199751346871115e-06, + "loss": 0.4817, + "step": 6758 + }, + { + "epoch": 0.5602154993783672, + "grad_norm": 0.2409360557794571, + "learning_rate": 2.1993369249896394e-06, + "loss": 0.4888, + "step": 6759 + }, + { + "epoch": 0.5602983837546622, + "grad_norm": 0.24902597069740295, + "learning_rate": 2.1989225031081642e-06, + "loss": 0.5345, + "step": 6760 + }, + { + "epoch": 0.5603812681309573, + "grad_norm": 0.24342364072799683, + "learning_rate": 2.198508081226689e-06, + "loss": 0.4884, + "step": 6761 + }, + { + "epoch": 0.5604641525072523, + "grad_norm": 0.27023157477378845, + "learning_rate": 2.198093659345214e-06, + "loss": 0.498, + "step": 6762 + }, + { + "epoch": 0.5605470368835475, + "grad_norm": 0.24152563512325287, + "learning_rate": 2.1976792374637384e-06, + "loss": 0.561, + "step": 6763 + }, + { + "epoch": 0.5606299212598426, + "grad_norm": 0.2502158284187317, + "learning_rate": 2.197264815582263e-06, + "loss": 0.4694, + "step": 6764 + }, + { + "epoch": 0.5607128056361376, + "grad_norm": 0.26024189591407776, + "learning_rate": 2.1968503937007877e-06, + "loss": 0.4852, + "step": 6765 + }, + { + "epoch": 0.5607956900124327, + "grad_norm": 0.2368878275156021, + "learning_rate": 2.1964359718193125e-06, + "loss": 0.4924, + "step": 6766 + }, + { + "epoch": 0.5608785743887277, + "grad_norm": 0.2454586625099182, + "learning_rate": 2.196021549937837e-06, + "loss": 0.5143, + "step": 6767 + }, + { + "epoch": 0.5609614587650228, + "grad_norm": 0.2631664574146271, + "learning_rate": 2.1956071280563614e-06, + "loss": 0.5198, + "step": 6768 + }, + { + "epoch": 0.5610443431413179, + "grad_norm": 0.28291255235671997, + "learning_rate": 2.1951927061748862e-06, + "loss": 0.5204, + "step": 6769 + }, + { + "epoch": 0.5611272275176129, + "grad_norm": 0.25925248861312866, + "learning_rate": 2.194778284293411e-06, + "loss": 0.5437, + "step": 6770 + }, + { + "epoch": 0.561210111893908, + "grad_norm": 0.2535804510116577, + "learning_rate": 2.1943638624119355e-06, + "loss": 0.4765, + "step": 6771 + }, + { + "epoch": 0.561292996270203, + "grad_norm": 0.2360614389181137, + "learning_rate": 2.19394944053046e-06, + "loss": 0.4636, + "step": 6772 + }, + { + "epoch": 0.5613758806464981, + "grad_norm": 0.2677358090877533, + "learning_rate": 2.193535018648985e-06, + "loss": 0.5219, + "step": 6773 + }, + { + "epoch": 0.5614587650227932, + "grad_norm": 0.25463342666625977, + "learning_rate": 2.1931205967675097e-06, + "loss": 0.5362, + "step": 6774 + }, + { + "epoch": 0.5615416493990882, + "grad_norm": 0.2519337832927704, + "learning_rate": 2.192706174886034e-06, + "loss": 0.5181, + "step": 6775 + }, + { + "epoch": 0.5616245337753833, + "grad_norm": 0.2742956578731537, + "learning_rate": 2.1922917530045585e-06, + "loss": 0.5188, + "step": 6776 + }, + { + "epoch": 0.5617074181516785, + "grad_norm": 0.2522757053375244, + "learning_rate": 2.1918773311230834e-06, + "loss": 0.5341, + "step": 6777 + }, + { + "epoch": 0.5617903025279735, + "grad_norm": 0.2416883260011673, + "learning_rate": 2.1914629092416082e-06, + "loss": 0.4975, + "step": 6778 + }, + { + "epoch": 0.5618731869042686, + "grad_norm": 0.29266440868377686, + "learning_rate": 2.1910484873601327e-06, + "loss": 0.5214, + "step": 6779 + }, + { + "epoch": 0.5619560712805636, + "grad_norm": 0.26967450976371765, + "learning_rate": 2.1906340654786575e-06, + "loss": 0.6286, + "step": 6780 + }, + { + "epoch": 0.5620389556568587, + "grad_norm": 0.24212948977947235, + "learning_rate": 2.190219643597182e-06, + "loss": 0.4824, + "step": 6781 + }, + { + "epoch": 0.5621218400331538, + "grad_norm": 0.24513229727745056, + "learning_rate": 2.189805221715707e-06, + "loss": 0.4991, + "step": 6782 + }, + { + "epoch": 0.5622047244094488, + "grad_norm": 0.2748834192752838, + "learning_rate": 2.1893907998342317e-06, + "loss": 0.5547, + "step": 6783 + }, + { + "epoch": 0.5622876087857439, + "grad_norm": 0.274343878030777, + "learning_rate": 2.188976377952756e-06, + "loss": 0.5653, + "step": 6784 + }, + { + "epoch": 0.5623704931620389, + "grad_norm": 0.24547961354255676, + "learning_rate": 2.1885619560712805e-06, + "loss": 0.5109, + "step": 6785 + }, + { + "epoch": 0.562453377538334, + "grad_norm": 0.24585209786891937, + "learning_rate": 2.1881475341898054e-06, + "loss": 0.4764, + "step": 6786 + }, + { + "epoch": 0.5625362619146291, + "grad_norm": 0.2633073329925537, + "learning_rate": 2.1877331123083302e-06, + "loss": 0.5022, + "step": 6787 + }, + { + "epoch": 0.5626191462909241, + "grad_norm": 0.26757538318634033, + "learning_rate": 2.1873186904268547e-06, + "loss": 0.539, + "step": 6788 + }, + { + "epoch": 0.5627020306672192, + "grad_norm": 0.2609325051307678, + "learning_rate": 2.186904268545379e-06, + "loss": 0.5312, + "step": 6789 + }, + { + "epoch": 0.5627849150435142, + "grad_norm": 0.23535577952861786, + "learning_rate": 2.186489846663904e-06, + "loss": 0.4833, + "step": 6790 + }, + { + "epoch": 0.5628677994198094, + "grad_norm": 0.2535238265991211, + "learning_rate": 2.186075424782429e-06, + "loss": 0.5265, + "step": 6791 + }, + { + "epoch": 0.5629506837961045, + "grad_norm": 0.2781060039997101, + "learning_rate": 2.1856610029009532e-06, + "loss": 0.5258, + "step": 6792 + }, + { + "epoch": 0.5630335681723995, + "grad_norm": 0.26238733530044556, + "learning_rate": 2.185246581019478e-06, + "loss": 0.5331, + "step": 6793 + }, + { + "epoch": 0.5631164525486946, + "grad_norm": 0.23170161247253418, + "learning_rate": 2.1848321591380025e-06, + "loss": 0.4866, + "step": 6794 + }, + { + "epoch": 0.5631993369249897, + "grad_norm": 0.28719040751457214, + "learning_rate": 2.1844177372565274e-06, + "loss": 0.4896, + "step": 6795 + }, + { + "epoch": 0.5632822213012847, + "grad_norm": 0.25430524349212646, + "learning_rate": 2.184003315375052e-06, + "loss": 0.5283, + "step": 6796 + }, + { + "epoch": 0.5633651056775798, + "grad_norm": 0.2664030194282532, + "learning_rate": 2.1835888934935767e-06, + "loss": 0.522, + "step": 6797 + }, + { + "epoch": 0.5634479900538748, + "grad_norm": 0.2651091516017914, + "learning_rate": 2.183174471612101e-06, + "loss": 0.5184, + "step": 6798 + }, + { + "epoch": 0.5635308744301699, + "grad_norm": 0.2563745975494385, + "learning_rate": 2.182760049730626e-06, + "loss": 0.4956, + "step": 6799 + }, + { + "epoch": 0.563613758806465, + "grad_norm": 0.26024138927459717, + "learning_rate": 2.182345627849151e-06, + "loss": 0.4975, + "step": 6800 + }, + { + "epoch": 0.56369664318276, + "grad_norm": 0.2537502646446228, + "learning_rate": 2.1819312059676752e-06, + "loss": 0.5161, + "step": 6801 + }, + { + "epoch": 0.5637795275590551, + "grad_norm": 0.272898405790329, + "learning_rate": 2.1815167840862e-06, + "loss": 0.5311, + "step": 6802 + }, + { + "epoch": 0.5638624119353501, + "grad_norm": 0.2569980323314667, + "learning_rate": 2.1811023622047245e-06, + "loss": 0.5019, + "step": 6803 + }, + { + "epoch": 0.5639452963116452, + "grad_norm": 0.24839839339256287, + "learning_rate": 2.1806879403232494e-06, + "loss": 0.4864, + "step": 6804 + }, + { + "epoch": 0.5640281806879404, + "grad_norm": 0.2477736622095108, + "learning_rate": 2.180273518441774e-06, + "loss": 0.5092, + "step": 6805 + }, + { + "epoch": 0.5641110650642354, + "grad_norm": 0.24689464271068573, + "learning_rate": 2.1798590965602987e-06, + "loss": 0.5135, + "step": 6806 + }, + { + "epoch": 0.5641939494405305, + "grad_norm": 0.25871995091438293, + "learning_rate": 2.179444674678823e-06, + "loss": 0.4806, + "step": 6807 + }, + { + "epoch": 0.5642768338168255, + "grad_norm": 0.24663349986076355, + "learning_rate": 2.179030252797348e-06, + "loss": 0.5366, + "step": 6808 + }, + { + "epoch": 0.5643597181931206, + "grad_norm": 0.2565981149673462, + "learning_rate": 2.1786158309158724e-06, + "loss": 0.5383, + "step": 6809 + }, + { + "epoch": 0.5644426025694157, + "grad_norm": 0.24843023717403412, + "learning_rate": 2.1782014090343973e-06, + "loss": 0.497, + "step": 6810 + }, + { + "epoch": 0.5645254869457107, + "grad_norm": 0.24689508974552155, + "learning_rate": 2.177786987152922e-06, + "loss": 0.469, + "step": 6811 + }, + { + "epoch": 0.5646083713220058, + "grad_norm": 0.2394951581954956, + "learning_rate": 2.1773725652714465e-06, + "loss": 0.4829, + "step": 6812 + }, + { + "epoch": 0.5646912556983009, + "grad_norm": 0.2677682638168335, + "learning_rate": 2.176958143389971e-06, + "loss": 0.5337, + "step": 6813 + }, + { + "epoch": 0.5647741400745959, + "grad_norm": 0.23785725235939026, + "learning_rate": 2.176543721508496e-06, + "loss": 0.4731, + "step": 6814 + }, + { + "epoch": 0.564857024450891, + "grad_norm": 0.26679790019989014, + "learning_rate": 2.1761292996270207e-06, + "loss": 0.5047, + "step": 6815 + }, + { + "epoch": 0.564939908827186, + "grad_norm": 0.2449958771467209, + "learning_rate": 2.175714877745545e-06, + "loss": 0.4957, + "step": 6816 + }, + { + "epoch": 0.5650227932034811, + "grad_norm": 0.24972158670425415, + "learning_rate": 2.1753004558640695e-06, + "loss": 0.5234, + "step": 6817 + }, + { + "epoch": 0.5651056775797763, + "grad_norm": 0.24558044970035553, + "learning_rate": 2.1748860339825944e-06, + "loss": 0.5078, + "step": 6818 + }, + { + "epoch": 0.5651885619560713, + "grad_norm": 0.24016351997852325, + "learning_rate": 2.1744716121011193e-06, + "loss": 0.4969, + "step": 6819 + }, + { + "epoch": 0.5652714463323664, + "grad_norm": 0.26924651861190796, + "learning_rate": 2.1740571902196437e-06, + "loss": 0.5324, + "step": 6820 + }, + { + "epoch": 0.5653543307086614, + "grad_norm": 0.2732982635498047, + "learning_rate": 2.1736427683381685e-06, + "loss": 0.5443, + "step": 6821 + }, + { + "epoch": 0.5654372150849565, + "grad_norm": 0.2457607239484787, + "learning_rate": 2.173228346456693e-06, + "loss": 0.4781, + "step": 6822 + }, + { + "epoch": 0.5655200994612516, + "grad_norm": 0.24153544008731842, + "learning_rate": 2.172813924575218e-06, + "loss": 0.4706, + "step": 6823 + }, + { + "epoch": 0.5656029838375466, + "grad_norm": 0.27105608582496643, + "learning_rate": 2.1723995026937427e-06, + "loss": 0.5712, + "step": 6824 + }, + { + "epoch": 0.5656858682138417, + "grad_norm": 0.25592291355133057, + "learning_rate": 2.171985080812267e-06, + "loss": 0.5493, + "step": 6825 + }, + { + "epoch": 0.5657687525901368, + "grad_norm": 0.2603011429309845, + "learning_rate": 2.1715706589307915e-06, + "loss": 0.5066, + "step": 6826 + }, + { + "epoch": 0.5658516369664318, + "grad_norm": 0.23842434585094452, + "learning_rate": 2.1711562370493164e-06, + "loss": 0.4958, + "step": 6827 + }, + { + "epoch": 0.5659345213427269, + "grad_norm": 0.2782936990261078, + "learning_rate": 2.1707418151678413e-06, + "loss": 0.553, + "step": 6828 + }, + { + "epoch": 0.5660174057190219, + "grad_norm": 0.23288938403129578, + "learning_rate": 2.1703273932863657e-06, + "loss": 0.4857, + "step": 6829 + }, + { + "epoch": 0.566100290095317, + "grad_norm": 0.27536720037460327, + "learning_rate": 2.16991297140489e-06, + "loss": 0.5773, + "step": 6830 + }, + { + "epoch": 0.5661831744716121, + "grad_norm": 0.23298627138137817, + "learning_rate": 2.169498549523415e-06, + "loss": 0.506, + "step": 6831 + }, + { + "epoch": 0.5662660588479071, + "grad_norm": 0.24248434603214264, + "learning_rate": 2.16908412764194e-06, + "loss": 0.5051, + "step": 6832 + }, + { + "epoch": 0.5663489432242023, + "grad_norm": 0.24425236880779266, + "learning_rate": 2.1686697057604643e-06, + "loss": 0.4904, + "step": 6833 + }, + { + "epoch": 0.5664318276004973, + "grad_norm": 0.26388466358184814, + "learning_rate": 2.1682552838789887e-06, + "loss": 0.5388, + "step": 6834 + }, + { + "epoch": 0.5665147119767924, + "grad_norm": 0.30168116092681885, + "learning_rate": 2.1678408619975136e-06, + "loss": 0.4787, + "step": 6835 + }, + { + "epoch": 0.5665975963530875, + "grad_norm": 0.24056224524974823, + "learning_rate": 2.1674264401160384e-06, + "loss": 0.5366, + "step": 6836 + }, + { + "epoch": 0.5666804807293825, + "grad_norm": 0.250581294298172, + "learning_rate": 2.167012018234563e-06, + "loss": 0.5098, + "step": 6837 + }, + { + "epoch": 0.5667633651056776, + "grad_norm": 0.26606622338294983, + "learning_rate": 2.1665975963530877e-06, + "loss": 0.5117, + "step": 6838 + }, + { + "epoch": 0.5668462494819726, + "grad_norm": 0.24313655495643616, + "learning_rate": 2.166183174471612e-06, + "loss": 0.5291, + "step": 6839 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 0.24502582848072052, + "learning_rate": 2.165768752590137e-06, + "loss": 0.4997, + "step": 6840 + }, + { + "epoch": 0.5670120182345628, + "grad_norm": 0.26168525218963623, + "learning_rate": 2.165354330708662e-06, + "loss": 0.5031, + "step": 6841 + }, + { + "epoch": 0.5670949026108578, + "grad_norm": 0.28097787499427795, + "learning_rate": 2.1649399088271863e-06, + "loss": 0.5407, + "step": 6842 + }, + { + "epoch": 0.5671777869871529, + "grad_norm": 0.23830267786979675, + "learning_rate": 2.1645254869457107e-06, + "loss": 0.4688, + "step": 6843 + }, + { + "epoch": 0.567260671363448, + "grad_norm": 0.24095578491687775, + "learning_rate": 2.1641110650642356e-06, + "loss": 0.4776, + "step": 6844 + }, + { + "epoch": 0.567343555739743, + "grad_norm": 0.25086846947669983, + "learning_rate": 2.1636966431827604e-06, + "loss": 0.4805, + "step": 6845 + }, + { + "epoch": 0.5674264401160382, + "grad_norm": 0.27203264832496643, + "learning_rate": 2.163282221301285e-06, + "loss": 0.4878, + "step": 6846 + }, + { + "epoch": 0.5675093244923332, + "grad_norm": 0.24917611479759216, + "learning_rate": 2.1628677994198097e-06, + "loss": 0.53, + "step": 6847 + }, + { + "epoch": 0.5675922088686283, + "grad_norm": 0.2567592263221741, + "learning_rate": 2.162453377538334e-06, + "loss": 0.5126, + "step": 6848 + }, + { + "epoch": 0.5676750932449234, + "grad_norm": 0.2687549889087677, + "learning_rate": 2.162038955656859e-06, + "loss": 0.5328, + "step": 6849 + }, + { + "epoch": 0.5677579776212184, + "grad_norm": 0.2636517882347107, + "learning_rate": 2.1616245337753834e-06, + "loss": 0.5359, + "step": 6850 + }, + { + "epoch": 0.5678408619975135, + "grad_norm": 0.23475773632526398, + "learning_rate": 2.1612101118939083e-06, + "loss": 0.4751, + "step": 6851 + }, + { + "epoch": 0.5679237463738085, + "grad_norm": 0.26145997643470764, + "learning_rate": 2.1607956900124327e-06, + "loss": 0.4986, + "step": 6852 + }, + { + "epoch": 0.5680066307501036, + "grad_norm": 0.2344672530889511, + "learning_rate": 2.1603812681309576e-06, + "loss": 0.4882, + "step": 6853 + }, + { + "epoch": 0.5680895151263987, + "grad_norm": 0.25856712460517883, + "learning_rate": 2.159966846249482e-06, + "loss": 0.5081, + "step": 6854 + }, + { + "epoch": 0.5681723995026937, + "grad_norm": 0.2712509036064148, + "learning_rate": 2.159552424368007e-06, + "loss": 0.5107, + "step": 6855 + }, + { + "epoch": 0.5682552838789888, + "grad_norm": 0.2448880523443222, + "learning_rate": 2.1591380024865317e-06, + "loss": 0.4895, + "step": 6856 + }, + { + "epoch": 0.5683381682552838, + "grad_norm": 0.2586429715156555, + "learning_rate": 2.158723580605056e-06, + "loss": 0.5014, + "step": 6857 + }, + { + "epoch": 0.5684210526315789, + "grad_norm": 0.23744970560073853, + "learning_rate": 2.1583091587235806e-06, + "loss": 0.5214, + "step": 6858 + }, + { + "epoch": 0.568503937007874, + "grad_norm": 0.24999180436134338, + "learning_rate": 2.1578947368421054e-06, + "loss": 0.5363, + "step": 6859 + }, + { + "epoch": 0.568586821384169, + "grad_norm": 0.25269779562950134, + "learning_rate": 2.1574803149606303e-06, + "loss": 0.5184, + "step": 6860 + }, + { + "epoch": 0.5686697057604642, + "grad_norm": 0.22791919112205505, + "learning_rate": 2.1570658930791547e-06, + "loss": 0.491, + "step": 6861 + }, + { + "epoch": 0.5687525901367593, + "grad_norm": 0.25028613209724426, + "learning_rate": 2.1566514711976796e-06, + "loss": 0.4775, + "step": 6862 + }, + { + "epoch": 0.5688354745130543, + "grad_norm": 0.23463541269302368, + "learning_rate": 2.156237049316204e-06, + "loss": 0.5088, + "step": 6863 + }, + { + "epoch": 0.5689183588893494, + "grad_norm": 0.2559487521648407, + "learning_rate": 2.155822627434729e-06, + "loss": 0.5128, + "step": 6864 + }, + { + "epoch": 0.5690012432656444, + "grad_norm": 0.24276772141456604, + "learning_rate": 2.1554082055532533e-06, + "loss": 0.4868, + "step": 6865 + }, + { + "epoch": 0.5690841276419395, + "grad_norm": 0.2374277114868164, + "learning_rate": 2.154993783671778e-06, + "loss": 0.4932, + "step": 6866 + }, + { + "epoch": 0.5691670120182346, + "grad_norm": 0.25064802169799805, + "learning_rate": 2.1545793617903026e-06, + "loss": 0.5269, + "step": 6867 + }, + { + "epoch": 0.5692498963945296, + "grad_norm": 0.2636088728904724, + "learning_rate": 2.1541649399088274e-06, + "loss": 0.5369, + "step": 6868 + }, + { + "epoch": 0.5693327807708247, + "grad_norm": 0.25316721200942993, + "learning_rate": 2.1537505180273523e-06, + "loss": 0.5623, + "step": 6869 + }, + { + "epoch": 0.5694156651471197, + "grad_norm": 0.23277218639850616, + "learning_rate": 2.1533360961458767e-06, + "loss": 0.4962, + "step": 6870 + }, + { + "epoch": 0.5694985495234148, + "grad_norm": 0.2382570505142212, + "learning_rate": 2.152921674264401e-06, + "loss": 0.4811, + "step": 6871 + }, + { + "epoch": 0.5695814338997099, + "grad_norm": 0.25456398725509644, + "learning_rate": 2.152507252382926e-06, + "loss": 0.5224, + "step": 6872 + }, + { + "epoch": 0.5696643182760049, + "grad_norm": 0.2352830022573471, + "learning_rate": 2.152092830501451e-06, + "loss": 0.5352, + "step": 6873 + }, + { + "epoch": 0.5697472026523, + "grad_norm": 0.3195996582508087, + "learning_rate": 2.1516784086199753e-06, + "loss": 0.4926, + "step": 6874 + }, + { + "epoch": 0.5698300870285951, + "grad_norm": 0.25869014859199524, + "learning_rate": 2.1512639867384997e-06, + "loss": 0.5027, + "step": 6875 + }, + { + "epoch": 0.5699129714048902, + "grad_norm": 0.25786271691322327, + "learning_rate": 2.1508495648570246e-06, + "loss": 0.5064, + "step": 6876 + }, + { + "epoch": 0.5699958557811853, + "grad_norm": 0.23170273005962372, + "learning_rate": 2.1504351429755494e-06, + "loss": 0.4651, + "step": 6877 + }, + { + "epoch": 0.5700787401574803, + "grad_norm": 0.2402847409248352, + "learning_rate": 2.150020721094074e-06, + "loss": 0.474, + "step": 6878 + }, + { + "epoch": 0.5701616245337754, + "grad_norm": 0.2535620331764221, + "learning_rate": 2.1496062992125983e-06, + "loss": 0.4883, + "step": 6879 + }, + { + "epoch": 0.5702445089100705, + "grad_norm": 0.2515906095504761, + "learning_rate": 2.149191877331123e-06, + "loss": 0.5198, + "step": 6880 + }, + { + "epoch": 0.5703273932863655, + "grad_norm": 0.248002290725708, + "learning_rate": 2.148777455449648e-06, + "loss": 0.4926, + "step": 6881 + }, + { + "epoch": 0.5704102776626606, + "grad_norm": 0.25074461102485657, + "learning_rate": 2.148363033568173e-06, + "loss": 0.5219, + "step": 6882 + }, + { + "epoch": 0.5704931620389556, + "grad_norm": 0.24167846143245697, + "learning_rate": 2.1479486116866973e-06, + "loss": 0.4846, + "step": 6883 + }, + { + "epoch": 0.5705760464152507, + "grad_norm": 0.24150928854942322, + "learning_rate": 2.1475341898052217e-06, + "loss": 0.4971, + "step": 6884 + }, + { + "epoch": 0.5706589307915458, + "grad_norm": 0.25882285833358765, + "learning_rate": 2.1471197679237466e-06, + "loss": 0.5564, + "step": 6885 + }, + { + "epoch": 0.5707418151678408, + "grad_norm": 0.26763129234313965, + "learning_rate": 2.1467053460422714e-06, + "loss": 0.5297, + "step": 6886 + }, + { + "epoch": 0.570824699544136, + "grad_norm": 0.2488892823457718, + "learning_rate": 2.146290924160796e-06, + "loss": 0.5425, + "step": 6887 + }, + { + "epoch": 0.570907583920431, + "grad_norm": 0.25771936774253845, + "learning_rate": 2.1458765022793203e-06, + "loss": 0.4873, + "step": 6888 + }, + { + "epoch": 0.5709904682967261, + "grad_norm": 0.27319231629371643, + "learning_rate": 2.145462080397845e-06, + "loss": 0.4759, + "step": 6889 + }, + { + "epoch": 0.5710733526730212, + "grad_norm": 0.2465440183877945, + "learning_rate": 2.14504765851637e-06, + "loss": 0.5173, + "step": 6890 + }, + { + "epoch": 0.5711562370493162, + "grad_norm": 0.25317203998565674, + "learning_rate": 2.1446332366348944e-06, + "loss": 0.5089, + "step": 6891 + }, + { + "epoch": 0.5712391214256113, + "grad_norm": 0.28206726908683777, + "learning_rate": 2.144218814753419e-06, + "loss": 0.5547, + "step": 6892 + }, + { + "epoch": 0.5713220058019063, + "grad_norm": 0.22236795723438263, + "learning_rate": 2.1438043928719437e-06, + "loss": 0.4575, + "step": 6893 + }, + { + "epoch": 0.5714048901782014, + "grad_norm": 0.26360049843788147, + "learning_rate": 2.1433899709904686e-06, + "loss": 0.5071, + "step": 6894 + }, + { + "epoch": 0.5714877745544965, + "grad_norm": 0.24831523001194, + "learning_rate": 2.142975549108993e-06, + "loss": 0.4892, + "step": 6895 + }, + { + "epoch": 0.5715706589307915, + "grad_norm": 0.2780512571334839, + "learning_rate": 2.142561127227518e-06, + "loss": 0.5179, + "step": 6896 + }, + { + "epoch": 0.5716535433070866, + "grad_norm": 0.22670027613639832, + "learning_rate": 2.1421467053460423e-06, + "loss": 0.4526, + "step": 6897 + }, + { + "epoch": 0.5717364276833817, + "grad_norm": 0.2650861442089081, + "learning_rate": 2.141732283464567e-06, + "loss": 0.5604, + "step": 6898 + }, + { + "epoch": 0.5718193120596767, + "grad_norm": 0.25674375891685486, + "learning_rate": 2.1413178615830916e-06, + "loss": 0.455, + "step": 6899 + }, + { + "epoch": 0.5719021964359718, + "grad_norm": 0.25056350231170654, + "learning_rate": 2.1409034397016164e-06, + "loss": 0.5504, + "step": 6900 + }, + { + "epoch": 0.5719850808122668, + "grad_norm": 0.2588407099246979, + "learning_rate": 2.140489017820141e-06, + "loss": 0.5423, + "step": 6901 + }, + { + "epoch": 0.572067965188562, + "grad_norm": 0.24767525494098663, + "learning_rate": 2.1400745959386657e-06, + "loss": 0.5207, + "step": 6902 + }, + { + "epoch": 0.5721508495648571, + "grad_norm": 0.2453669160604477, + "learning_rate": 2.1396601740571906e-06, + "loss": 0.5088, + "step": 6903 + }, + { + "epoch": 0.5722337339411521, + "grad_norm": 0.26973190903663635, + "learning_rate": 2.139245752175715e-06, + "loss": 0.5401, + "step": 6904 + }, + { + "epoch": 0.5723166183174472, + "grad_norm": 0.2471945881843567, + "learning_rate": 2.13883133029424e-06, + "loss": 0.5345, + "step": 6905 + }, + { + "epoch": 0.5723995026937422, + "grad_norm": 0.23530882596969604, + "learning_rate": 2.1384169084127643e-06, + "loss": 0.4944, + "step": 6906 + }, + { + "epoch": 0.5724823870700373, + "grad_norm": 0.24947302043437958, + "learning_rate": 2.138002486531289e-06, + "loss": 0.5359, + "step": 6907 + }, + { + "epoch": 0.5725652714463324, + "grad_norm": 0.2509220838546753, + "learning_rate": 2.1375880646498136e-06, + "loss": 0.5075, + "step": 6908 + }, + { + "epoch": 0.5726481558226274, + "grad_norm": 0.23490571975708008, + "learning_rate": 2.1371736427683384e-06, + "loss": 0.5121, + "step": 6909 + }, + { + "epoch": 0.5727310401989225, + "grad_norm": 0.24267177283763885, + "learning_rate": 2.136759220886863e-06, + "loss": 0.5024, + "step": 6910 + }, + { + "epoch": 0.5728139245752176, + "grad_norm": 0.25396257638931274, + "learning_rate": 2.1363447990053877e-06, + "loss": 0.4955, + "step": 6911 + }, + { + "epoch": 0.5728968089515126, + "grad_norm": 0.25273582339286804, + "learning_rate": 2.135930377123912e-06, + "loss": 0.4966, + "step": 6912 + }, + { + "epoch": 0.5729796933278077, + "grad_norm": 0.3058982491493225, + "learning_rate": 2.135515955242437e-06, + "loss": 0.4921, + "step": 6913 + }, + { + "epoch": 0.5730625777041027, + "grad_norm": 0.25896719098091125, + "learning_rate": 2.135101533360962e-06, + "loss": 0.5375, + "step": 6914 + }, + { + "epoch": 0.5731454620803978, + "grad_norm": 0.25453031063079834, + "learning_rate": 2.1346871114794863e-06, + "loss": 0.4932, + "step": 6915 + }, + { + "epoch": 0.573228346456693, + "grad_norm": 0.26555195450782776, + "learning_rate": 2.1342726895980107e-06, + "loss": 0.472, + "step": 6916 + }, + { + "epoch": 0.573311230832988, + "grad_norm": 0.25462326407432556, + "learning_rate": 2.1338582677165356e-06, + "loss": 0.5206, + "step": 6917 + }, + { + "epoch": 0.5733941152092831, + "grad_norm": 0.2751496732234955, + "learning_rate": 2.1334438458350605e-06, + "loss": 0.5475, + "step": 6918 + }, + { + "epoch": 0.5734769995855781, + "grad_norm": 0.25081750750541687, + "learning_rate": 2.133029423953585e-06, + "loss": 0.4747, + "step": 6919 + }, + { + "epoch": 0.5735598839618732, + "grad_norm": 0.23758797347545624, + "learning_rate": 2.1326150020721097e-06, + "loss": 0.4933, + "step": 6920 + }, + { + "epoch": 0.5736427683381683, + "grad_norm": 0.24933770298957825, + "learning_rate": 2.132200580190634e-06, + "loss": 0.4944, + "step": 6921 + }, + { + "epoch": 0.5737256527144633, + "grad_norm": 0.24576669931411743, + "learning_rate": 2.131786158309159e-06, + "loss": 0.4971, + "step": 6922 + }, + { + "epoch": 0.5738085370907584, + "grad_norm": 0.22193902730941772, + "learning_rate": 2.131371736427684e-06, + "loss": 0.4692, + "step": 6923 + }, + { + "epoch": 0.5738914214670534, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.1309573145462083e-06, + "loss": 0.4699, + "step": 6924 + }, + { + "epoch": 0.5739743058433485, + "grad_norm": 0.2476954311132431, + "learning_rate": 2.1305428926647327e-06, + "loss": 0.5195, + "step": 6925 + }, + { + "epoch": 0.5740571902196436, + "grad_norm": 0.2659852206707001, + "learning_rate": 2.1301284707832576e-06, + "loss": 0.5434, + "step": 6926 + }, + { + "epoch": 0.5741400745959386, + "grad_norm": 0.25771036744117737, + "learning_rate": 2.1297140489017825e-06, + "loss": 0.5089, + "step": 6927 + }, + { + "epoch": 0.5742229589722337, + "grad_norm": 0.25762423872947693, + "learning_rate": 2.129299627020307e-06, + "loss": 0.4807, + "step": 6928 + }, + { + "epoch": 0.5743058433485289, + "grad_norm": 0.2667299211025238, + "learning_rate": 2.1288852051388313e-06, + "loss": 0.5516, + "step": 6929 + }, + { + "epoch": 0.5743887277248239, + "grad_norm": 0.24360893666744232, + "learning_rate": 2.128470783257356e-06, + "loss": 0.4742, + "step": 6930 + }, + { + "epoch": 0.574471612101119, + "grad_norm": 0.23657257854938507, + "learning_rate": 2.128056361375881e-06, + "loss": 0.5112, + "step": 6931 + }, + { + "epoch": 0.574554496477414, + "grad_norm": 0.2701961398124695, + "learning_rate": 2.1276419394944055e-06, + "loss": 0.5303, + "step": 6932 + }, + { + "epoch": 0.5746373808537091, + "grad_norm": 0.25101447105407715, + "learning_rate": 2.12722751761293e-06, + "loss": 0.4999, + "step": 6933 + }, + { + "epoch": 0.5747202652300042, + "grad_norm": 0.24873068928718567, + "learning_rate": 2.1268130957314547e-06, + "loss": 0.4924, + "step": 6934 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 0.24996604025363922, + "learning_rate": 2.1263986738499796e-06, + "loss": 0.5298, + "step": 6935 + }, + { + "epoch": 0.5748860339825943, + "grad_norm": 0.24461808800697327, + "learning_rate": 2.125984251968504e-06, + "loss": 0.51, + "step": 6936 + }, + { + "epoch": 0.5749689183588893, + "grad_norm": 0.2643933594226837, + "learning_rate": 2.1255698300870285e-06, + "loss": 0.5264, + "step": 6937 + }, + { + "epoch": 0.5750518027351844, + "grad_norm": 0.23806890845298767, + "learning_rate": 2.1251554082055533e-06, + "loss": 0.5027, + "step": 6938 + }, + { + "epoch": 0.5751346871114795, + "grad_norm": 0.27218151092529297, + "learning_rate": 2.124740986324078e-06, + "loss": 0.5318, + "step": 6939 + }, + { + "epoch": 0.5752175714877745, + "grad_norm": 0.22350987792015076, + "learning_rate": 2.124326564442603e-06, + "loss": 0.4888, + "step": 6940 + }, + { + "epoch": 0.5753004558640696, + "grad_norm": 0.2323259711265564, + "learning_rate": 2.1239121425611275e-06, + "loss": 0.4729, + "step": 6941 + }, + { + "epoch": 0.5753833402403646, + "grad_norm": 0.24104352295398712, + "learning_rate": 2.123497720679652e-06, + "loss": 0.523, + "step": 6942 + }, + { + "epoch": 0.5754662246166597, + "grad_norm": 0.2431877702474594, + "learning_rate": 2.1230832987981768e-06, + "loss": 0.5062, + "step": 6943 + }, + { + "epoch": 0.5755491089929549, + "grad_norm": 0.26327648758888245, + "learning_rate": 2.1226688769167016e-06, + "loss": 0.529, + "step": 6944 + }, + { + "epoch": 0.5756319933692499, + "grad_norm": 0.2389914095401764, + "learning_rate": 2.122254455035226e-06, + "loss": 0.493, + "step": 6945 + }, + { + "epoch": 0.575714877745545, + "grad_norm": 0.2684669494628906, + "learning_rate": 2.1218400331537505e-06, + "loss": 0.5167, + "step": 6946 + }, + { + "epoch": 0.5757977621218401, + "grad_norm": 0.24491667747497559, + "learning_rate": 2.1214256112722753e-06, + "loss": 0.4982, + "step": 6947 + }, + { + "epoch": 0.5758806464981351, + "grad_norm": 0.26056501269340515, + "learning_rate": 2.1210111893908e-06, + "loss": 0.4786, + "step": 6948 + }, + { + "epoch": 0.5759635308744302, + "grad_norm": 0.24958321452140808, + "learning_rate": 2.1205967675093246e-06, + "loss": 0.5268, + "step": 6949 + }, + { + "epoch": 0.5760464152507252, + "grad_norm": 0.26304423809051514, + "learning_rate": 2.120182345627849e-06, + "loss": 0.5052, + "step": 6950 + }, + { + "epoch": 0.5761292996270203, + "grad_norm": 0.24252867698669434, + "learning_rate": 2.119767923746374e-06, + "loss": 0.492, + "step": 6951 + }, + { + "epoch": 0.5762121840033154, + "grad_norm": 0.24240742623806, + "learning_rate": 2.1193535018648988e-06, + "loss": 0.4901, + "step": 6952 + }, + { + "epoch": 0.5762950683796104, + "grad_norm": 0.233645498752594, + "learning_rate": 2.118939079983423e-06, + "loss": 0.4705, + "step": 6953 + }, + { + "epoch": 0.5763779527559055, + "grad_norm": 0.2472175806760788, + "learning_rate": 2.118524658101948e-06, + "loss": 0.5278, + "step": 6954 + }, + { + "epoch": 0.5764608371322005, + "grad_norm": 0.2415953278541565, + "learning_rate": 2.1181102362204725e-06, + "loss": 0.5276, + "step": 6955 + }, + { + "epoch": 0.5765437215084956, + "grad_norm": 0.2640228271484375, + "learning_rate": 2.1176958143389973e-06, + "loss": 0.525, + "step": 6956 + }, + { + "epoch": 0.5766266058847908, + "grad_norm": 0.24906033277511597, + "learning_rate": 2.1172813924575218e-06, + "loss": 0.5293, + "step": 6957 + }, + { + "epoch": 0.5767094902610858, + "grad_norm": 0.25417330861091614, + "learning_rate": 2.1168669705760466e-06, + "loss": 0.5645, + "step": 6958 + }, + { + "epoch": 0.5767923746373809, + "grad_norm": 0.544574499130249, + "learning_rate": 2.116452548694571e-06, + "loss": 0.547, + "step": 6959 + }, + { + "epoch": 0.5768752590136759, + "grad_norm": 0.26363080739974976, + "learning_rate": 2.116038126813096e-06, + "loss": 0.5433, + "step": 6960 + }, + { + "epoch": 0.576958143389971, + "grad_norm": 0.27027764916419983, + "learning_rate": 2.1156237049316208e-06, + "loss": 0.5273, + "step": 6961 + }, + { + "epoch": 0.5770410277662661, + "grad_norm": 0.2637671232223511, + "learning_rate": 2.115209283050145e-06, + "loss": 0.5328, + "step": 6962 + }, + { + "epoch": 0.5771239121425611, + "grad_norm": 0.2623881697654724, + "learning_rate": 2.11479486116867e-06, + "loss": 0.5152, + "step": 6963 + }, + { + "epoch": 0.5772067965188562, + "grad_norm": 0.2632167637348175, + "learning_rate": 2.1143804392871945e-06, + "loss": 0.554, + "step": 6964 + }, + { + "epoch": 0.5772896808951513, + "grad_norm": 0.2562211751937866, + "learning_rate": 2.1139660174057193e-06, + "loss": 0.5432, + "step": 6965 + }, + { + "epoch": 0.5773725652714463, + "grad_norm": 0.2768903076648712, + "learning_rate": 2.1135515955242438e-06, + "loss": 0.5348, + "step": 6966 + }, + { + "epoch": 0.5774554496477414, + "grad_norm": 0.24414454400539398, + "learning_rate": 2.1131371736427686e-06, + "loss": 0.4963, + "step": 6967 + }, + { + "epoch": 0.5775383340240364, + "grad_norm": 0.2897086441516876, + "learning_rate": 2.112722751761293e-06, + "loss": 0.5251, + "step": 6968 + }, + { + "epoch": 0.5776212184003315, + "grad_norm": 0.24225042760372162, + "learning_rate": 2.112308329879818e-06, + "loss": 0.4822, + "step": 6969 + }, + { + "epoch": 0.5777041027766266, + "grad_norm": 0.25282052159309387, + "learning_rate": 2.1118939079983423e-06, + "loss": 0.5048, + "step": 6970 + }, + { + "epoch": 0.5777869871529216, + "grad_norm": 0.25212106108665466, + "learning_rate": 2.111479486116867e-06, + "loss": 0.5221, + "step": 6971 + }, + { + "epoch": 0.5778698715292168, + "grad_norm": 0.26850950717926025, + "learning_rate": 2.111065064235392e-06, + "loss": 0.5506, + "step": 6972 + }, + { + "epoch": 0.5779527559055118, + "grad_norm": 0.2614234387874603, + "learning_rate": 2.1106506423539165e-06, + "loss": 0.528, + "step": 6973 + }, + { + "epoch": 0.5780356402818069, + "grad_norm": 0.2378188967704773, + "learning_rate": 2.110236220472441e-06, + "loss": 0.4752, + "step": 6974 + }, + { + "epoch": 0.578118524658102, + "grad_norm": 0.24433471262454987, + "learning_rate": 2.1098217985909658e-06, + "loss": 0.4895, + "step": 6975 + }, + { + "epoch": 0.578201409034397, + "grad_norm": 0.24690131843090057, + "learning_rate": 2.1094073767094906e-06, + "loss": 0.4804, + "step": 6976 + }, + { + "epoch": 0.5782842934106921, + "grad_norm": 0.26826491951942444, + "learning_rate": 2.108992954828015e-06, + "loss": 0.5246, + "step": 6977 + }, + { + "epoch": 0.5783671777869872, + "grad_norm": 0.2733611762523651, + "learning_rate": 2.1085785329465395e-06, + "loss": 0.5604, + "step": 6978 + }, + { + "epoch": 0.5784500621632822, + "grad_norm": 0.24956448376178741, + "learning_rate": 2.1081641110650643e-06, + "loss": 0.4614, + "step": 6979 + }, + { + "epoch": 0.5785329465395773, + "grad_norm": 0.2566247880458832, + "learning_rate": 2.107749689183589e-06, + "loss": 0.5032, + "step": 6980 + }, + { + "epoch": 0.5786158309158723, + "grad_norm": 0.27115607261657715, + "learning_rate": 2.107335267302114e-06, + "loss": 0.5004, + "step": 6981 + }, + { + "epoch": 0.5786987152921674, + "grad_norm": 0.24391800165176392, + "learning_rate": 2.1069208454206385e-06, + "loss": 0.4709, + "step": 6982 + }, + { + "epoch": 0.5787815996684625, + "grad_norm": 0.2551482617855072, + "learning_rate": 2.106506423539163e-06, + "loss": 0.5087, + "step": 6983 + }, + { + "epoch": 0.5788644840447575, + "grad_norm": 0.23404806852340698, + "learning_rate": 2.1060920016576878e-06, + "loss": 0.5133, + "step": 6984 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 0.24805274605751038, + "learning_rate": 2.1056775797762126e-06, + "loss": 0.4992, + "step": 6985 + }, + { + "epoch": 0.5790302527973477, + "grad_norm": 0.2542927861213684, + "learning_rate": 2.105263157894737e-06, + "loss": 0.5034, + "step": 6986 + }, + { + "epoch": 0.5791131371736428, + "grad_norm": 0.2653771638870239, + "learning_rate": 2.1048487360132615e-06, + "loss": 0.5226, + "step": 6987 + }, + { + "epoch": 0.5791960215499379, + "grad_norm": 0.2601239085197449, + "learning_rate": 2.1044343141317863e-06, + "loss": 0.533, + "step": 6988 + }, + { + "epoch": 0.5792789059262329, + "grad_norm": 0.25498005747795105, + "learning_rate": 2.104019892250311e-06, + "loss": 0.5106, + "step": 6989 + }, + { + "epoch": 0.579361790302528, + "grad_norm": 0.2512105703353882, + "learning_rate": 2.1036054703688356e-06, + "loss": 0.5309, + "step": 6990 + }, + { + "epoch": 0.579444674678823, + "grad_norm": 0.2515394687652588, + "learning_rate": 2.10319104848736e-06, + "loss": 0.5062, + "step": 6991 + }, + { + "epoch": 0.5795275590551181, + "grad_norm": 0.23252178728580475, + "learning_rate": 2.102776626605885e-06, + "loss": 0.4622, + "step": 6992 + }, + { + "epoch": 0.5796104434314132, + "grad_norm": 0.23711036145687103, + "learning_rate": 2.1023622047244098e-06, + "loss": 0.4955, + "step": 6993 + }, + { + "epoch": 0.5796933278077082, + "grad_norm": 0.24118472635746002, + "learning_rate": 2.101947782842934e-06, + "loss": 0.5359, + "step": 6994 + }, + { + "epoch": 0.5797762121840033, + "grad_norm": 0.235325425863266, + "learning_rate": 2.1015333609614586e-06, + "loss": 0.5113, + "step": 6995 + }, + { + "epoch": 0.5798590965602984, + "grad_norm": 0.22876925766468048, + "learning_rate": 2.1011189390799835e-06, + "loss": 0.4717, + "step": 6996 + }, + { + "epoch": 0.5799419809365934, + "grad_norm": 0.23370322585105896, + "learning_rate": 2.1007045171985083e-06, + "loss": 0.4895, + "step": 6997 + }, + { + "epoch": 0.5800248653128885, + "grad_norm": 0.2535211443901062, + "learning_rate": 2.1002900953170328e-06, + "loss": 0.5168, + "step": 6998 + }, + { + "epoch": 0.5801077496891835, + "grad_norm": 0.25341692566871643, + "learning_rate": 2.0998756734355576e-06, + "loss": 0.5049, + "step": 6999 + }, + { + "epoch": 0.5801906340654787, + "grad_norm": 0.26144468784332275, + "learning_rate": 2.099461251554082e-06, + "loss": 0.5314, + "step": 7000 + }, + { + "epoch": 0.5802735184417738, + "grad_norm": 0.25136107206344604, + "learning_rate": 2.099046829672607e-06, + "loss": 0.5552, + "step": 7001 + }, + { + "epoch": 0.5803564028180688, + "grad_norm": 0.2441081702709198, + "learning_rate": 2.0986324077911318e-06, + "loss": 0.5212, + "step": 7002 + }, + { + "epoch": 0.5804392871943639, + "grad_norm": 0.2558124363422394, + "learning_rate": 2.098217985909656e-06, + "loss": 0.5268, + "step": 7003 + }, + { + "epoch": 0.5805221715706589, + "grad_norm": 0.29374974966049194, + "learning_rate": 2.0978035640281806e-06, + "loss": 0.5533, + "step": 7004 + }, + { + "epoch": 0.580605055946954, + "grad_norm": 0.24382349848747253, + "learning_rate": 2.0973891421467055e-06, + "loss": 0.5273, + "step": 7005 + }, + { + "epoch": 0.5806879403232491, + "grad_norm": 0.25861698389053345, + "learning_rate": 2.0969747202652304e-06, + "loss": 0.5291, + "step": 7006 + }, + { + "epoch": 0.5807708246995441, + "grad_norm": 0.26489460468292236, + "learning_rate": 2.0965602983837548e-06, + "loss": 0.5378, + "step": 7007 + }, + { + "epoch": 0.5808537090758392, + "grad_norm": 0.2305721938610077, + "learning_rate": 2.0961458765022792e-06, + "loss": 0.4785, + "step": 7008 + }, + { + "epoch": 0.5809365934521342, + "grad_norm": 0.2587321996688843, + "learning_rate": 2.095731454620804e-06, + "loss": 0.5268, + "step": 7009 + }, + { + "epoch": 0.5810194778284293, + "grad_norm": 0.25018784403800964, + "learning_rate": 2.095317032739329e-06, + "loss": 0.5023, + "step": 7010 + }, + { + "epoch": 0.5811023622047244, + "grad_norm": 0.24712339043617249, + "learning_rate": 2.0949026108578534e-06, + "loss": 0.5032, + "step": 7011 + }, + { + "epoch": 0.5811852465810194, + "grad_norm": 0.2324785590171814, + "learning_rate": 2.0944881889763782e-06, + "loss": 0.509, + "step": 7012 + }, + { + "epoch": 0.5812681309573146, + "grad_norm": 0.24056555330753326, + "learning_rate": 2.0940737670949026e-06, + "loss": 0.5132, + "step": 7013 + }, + { + "epoch": 0.5813510153336097, + "grad_norm": 0.28391996026039124, + "learning_rate": 2.0936593452134275e-06, + "loss": 0.5009, + "step": 7014 + }, + { + "epoch": 0.5814338997099047, + "grad_norm": 0.23484179377555847, + "learning_rate": 2.093244923331952e-06, + "loss": 0.4897, + "step": 7015 + }, + { + "epoch": 0.5815167840861998, + "grad_norm": 0.3120536506175995, + "learning_rate": 2.092830501450477e-06, + "loss": 0.4926, + "step": 7016 + }, + { + "epoch": 0.5815996684624948, + "grad_norm": 0.2796797752380371, + "learning_rate": 2.0924160795690012e-06, + "loss": 0.5198, + "step": 7017 + }, + { + "epoch": 0.5816825528387899, + "grad_norm": 0.25453701615333557, + "learning_rate": 2.092001657687526e-06, + "loss": 0.4945, + "step": 7018 + }, + { + "epoch": 0.581765437215085, + "grad_norm": 0.27903154492378235, + "learning_rate": 2.091587235806051e-06, + "loss": 0.4968, + "step": 7019 + }, + { + "epoch": 0.58184832159138, + "grad_norm": 0.261905699968338, + "learning_rate": 2.0911728139245754e-06, + "loss": 0.5233, + "step": 7020 + }, + { + "epoch": 0.5819312059676751, + "grad_norm": 0.23436205089092255, + "learning_rate": 2.0907583920431002e-06, + "loss": 0.4773, + "step": 7021 + }, + { + "epoch": 0.5820140903439701, + "grad_norm": 0.2452559471130371, + "learning_rate": 2.0903439701616247e-06, + "loss": 0.4907, + "step": 7022 + }, + { + "epoch": 0.5820969747202652, + "grad_norm": 0.24142327904701233, + "learning_rate": 2.0899295482801495e-06, + "loss": 0.4851, + "step": 7023 + }, + { + "epoch": 0.5821798590965603, + "grad_norm": 0.24704895913600922, + "learning_rate": 2.089515126398674e-06, + "loss": 0.5126, + "step": 7024 + }, + { + "epoch": 0.5822627434728553, + "grad_norm": 0.2448217272758484, + "learning_rate": 2.089100704517199e-06, + "loss": 0.4602, + "step": 7025 + }, + { + "epoch": 0.5823456278491504, + "grad_norm": 0.24494588375091553, + "learning_rate": 2.0886862826357232e-06, + "loss": 0.4915, + "step": 7026 + }, + { + "epoch": 0.5824285122254454, + "grad_norm": 0.28574055433273315, + "learning_rate": 2.088271860754248e-06, + "loss": 0.5485, + "step": 7027 + }, + { + "epoch": 0.5825113966017406, + "grad_norm": 0.2651892304420471, + "learning_rate": 2.0878574388727725e-06, + "loss": 0.4915, + "step": 7028 + }, + { + "epoch": 0.5825942809780357, + "grad_norm": 0.25330835580825806, + "learning_rate": 2.0874430169912974e-06, + "loss": 0.4962, + "step": 7029 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 0.2751433849334717, + "learning_rate": 2.0870285951098222e-06, + "loss": 0.5396, + "step": 7030 + }, + { + "epoch": 0.5827600497306258, + "grad_norm": 0.2511467933654785, + "learning_rate": 2.0866141732283467e-06, + "loss": 0.4635, + "step": 7031 + }, + { + "epoch": 0.5828429341069209, + "grad_norm": 0.2671818137168884, + "learning_rate": 2.086199751346871e-06, + "loss": 0.5064, + "step": 7032 + }, + { + "epoch": 0.5829258184832159, + "grad_norm": 0.2497912347316742, + "learning_rate": 2.085785329465396e-06, + "loss": 0.5185, + "step": 7033 + }, + { + "epoch": 0.583008702859511, + "grad_norm": 0.2587089538574219, + "learning_rate": 2.085370907583921e-06, + "loss": 0.5345, + "step": 7034 + }, + { + "epoch": 0.583091587235806, + "grad_norm": 0.28195324540138245, + "learning_rate": 2.0849564857024452e-06, + "loss": 0.5341, + "step": 7035 + }, + { + "epoch": 0.5831744716121011, + "grad_norm": 0.2471335232257843, + "learning_rate": 2.0845420638209697e-06, + "loss": 0.4857, + "step": 7036 + }, + { + "epoch": 0.5832573559883962, + "grad_norm": 0.2673737108707428, + "learning_rate": 2.0841276419394945e-06, + "loss": 0.4723, + "step": 7037 + }, + { + "epoch": 0.5833402403646912, + "grad_norm": 0.252214252948761, + "learning_rate": 2.0837132200580194e-06, + "loss": 0.5516, + "step": 7038 + }, + { + "epoch": 0.5834231247409863, + "grad_norm": 0.24657896161079407, + "learning_rate": 2.083298798176544e-06, + "loss": 0.481, + "step": 7039 + }, + { + "epoch": 0.5835060091172813, + "grad_norm": 0.2771572768688202, + "learning_rate": 2.0828843762950687e-06, + "loss": 0.5429, + "step": 7040 + }, + { + "epoch": 0.5835888934935765, + "grad_norm": 0.2735159695148468, + "learning_rate": 2.082469954413593e-06, + "loss": 0.5184, + "step": 7041 + }, + { + "epoch": 0.5836717778698716, + "grad_norm": 0.2555929124355316, + "learning_rate": 2.082055532532118e-06, + "loss": 0.5176, + "step": 7042 + }, + { + "epoch": 0.5837546622461666, + "grad_norm": 0.2489946335554123, + "learning_rate": 2.081641110650643e-06, + "loss": 0.5208, + "step": 7043 + }, + { + "epoch": 0.5838375466224617, + "grad_norm": 0.24368712306022644, + "learning_rate": 2.0812266887691672e-06, + "loss": 0.4853, + "step": 7044 + }, + { + "epoch": 0.5839204309987568, + "grad_norm": 0.2540227174758911, + "learning_rate": 2.0808122668876917e-06, + "loss": 0.5299, + "step": 7045 + }, + { + "epoch": 0.5840033153750518, + "grad_norm": 0.2428118735551834, + "learning_rate": 2.0803978450062165e-06, + "loss": 0.5101, + "step": 7046 + }, + { + "epoch": 0.5840861997513469, + "grad_norm": 0.23730070888996124, + "learning_rate": 2.0799834231247414e-06, + "loss": 0.4985, + "step": 7047 + }, + { + "epoch": 0.5841690841276419, + "grad_norm": 0.2544708549976349, + "learning_rate": 2.079569001243266e-06, + "loss": 0.5332, + "step": 7048 + }, + { + "epoch": 0.584251968503937, + "grad_norm": 0.24635224044322968, + "learning_rate": 2.0791545793617902e-06, + "loss": 0.5407, + "step": 7049 + }, + { + "epoch": 0.5843348528802321, + "grad_norm": 0.2336777001619339, + "learning_rate": 2.078740157480315e-06, + "loss": 0.4701, + "step": 7050 + }, + { + "epoch": 0.5844177372565271, + "grad_norm": 0.25895237922668457, + "learning_rate": 2.07832573559884e-06, + "loss": 0.5073, + "step": 7051 + }, + { + "epoch": 0.5845006216328222, + "grad_norm": 0.2759750485420227, + "learning_rate": 2.0779113137173644e-06, + "loss": 0.5132, + "step": 7052 + }, + { + "epoch": 0.5845835060091172, + "grad_norm": 0.2508476674556732, + "learning_rate": 2.077496891835889e-06, + "loss": 0.5114, + "step": 7053 + }, + { + "epoch": 0.5846663903854123, + "grad_norm": 0.2519690692424774, + "learning_rate": 2.0770824699544137e-06, + "loss": 0.5247, + "step": 7054 + }, + { + "epoch": 0.5847492747617075, + "grad_norm": 0.25541624426841736, + "learning_rate": 2.0766680480729385e-06, + "loss": 0.5291, + "step": 7055 + }, + { + "epoch": 0.5848321591380025, + "grad_norm": 0.23479880392551422, + "learning_rate": 2.076253626191463e-06, + "loss": 0.5049, + "step": 7056 + }, + { + "epoch": 0.5849150435142976, + "grad_norm": 0.2396572232246399, + "learning_rate": 2.075839204309988e-06, + "loss": 0.4554, + "step": 7057 + }, + { + "epoch": 0.5849979278905926, + "grad_norm": 0.2508334815502167, + "learning_rate": 2.0754247824285122e-06, + "loss": 0.4851, + "step": 7058 + }, + { + "epoch": 0.5850808122668877, + "grad_norm": 0.2513599097728729, + "learning_rate": 2.075010360547037e-06, + "loss": 0.4938, + "step": 7059 + }, + { + "epoch": 0.5851636966431828, + "grad_norm": 0.2525906264781952, + "learning_rate": 2.074595938665562e-06, + "loss": 0.4715, + "step": 7060 + }, + { + "epoch": 0.5852465810194778, + "grad_norm": 0.2502525746822357, + "learning_rate": 2.0741815167840864e-06, + "loss": 0.5147, + "step": 7061 + }, + { + "epoch": 0.5853294653957729, + "grad_norm": 0.27871906757354736, + "learning_rate": 2.073767094902611e-06, + "loss": 0.5267, + "step": 7062 + }, + { + "epoch": 0.585412349772068, + "grad_norm": 0.25349000096321106, + "learning_rate": 2.0733526730211357e-06, + "loss": 0.5156, + "step": 7063 + }, + { + "epoch": 0.585495234148363, + "grad_norm": 0.28002166748046875, + "learning_rate": 2.0729382511396605e-06, + "loss": 0.5196, + "step": 7064 + }, + { + "epoch": 0.5855781185246581, + "grad_norm": 0.23989444971084595, + "learning_rate": 2.072523829258185e-06, + "loss": 0.4864, + "step": 7065 + }, + { + "epoch": 0.5856610029009531, + "grad_norm": 0.2547285258769989, + "learning_rate": 2.07210940737671e-06, + "loss": 0.5212, + "step": 7066 + }, + { + "epoch": 0.5857438872772482, + "grad_norm": 0.2565279006958008, + "learning_rate": 2.0716949854952342e-06, + "loss": 0.5118, + "step": 7067 + }, + { + "epoch": 0.5858267716535434, + "grad_norm": 0.22572758793830872, + "learning_rate": 2.071280563613759e-06, + "loss": 0.5072, + "step": 7068 + }, + { + "epoch": 0.5859096560298384, + "grad_norm": 0.24741727113723755, + "learning_rate": 2.0708661417322835e-06, + "loss": 0.4875, + "step": 7069 + }, + { + "epoch": 0.5859925404061335, + "grad_norm": 0.24452349543571472, + "learning_rate": 2.0704517198508084e-06, + "loss": 0.5319, + "step": 7070 + }, + { + "epoch": 0.5860754247824285, + "grad_norm": 0.24936793744564056, + "learning_rate": 2.070037297969333e-06, + "loss": 0.4805, + "step": 7071 + }, + { + "epoch": 0.5861583091587236, + "grad_norm": 0.2534197270870209, + "learning_rate": 2.0696228760878577e-06, + "loss": 0.536, + "step": 7072 + }, + { + "epoch": 0.5862411935350187, + "grad_norm": 0.2716324031352997, + "learning_rate": 2.069208454206382e-06, + "loss": 0.5634, + "step": 7073 + }, + { + "epoch": 0.5863240779113137, + "grad_norm": 0.25334855914115906, + "learning_rate": 2.068794032324907e-06, + "loss": 0.5179, + "step": 7074 + }, + { + "epoch": 0.5864069622876088, + "grad_norm": 0.2360331118106842, + "learning_rate": 2.068379610443432e-06, + "loss": 0.4939, + "step": 7075 + }, + { + "epoch": 0.5864898466639038, + "grad_norm": 0.2439807504415512, + "learning_rate": 2.0679651885619562e-06, + "loss": 0.4795, + "step": 7076 + }, + { + "epoch": 0.5865727310401989, + "grad_norm": 0.24327127635478973, + "learning_rate": 2.0675507666804807e-06, + "loss": 0.5132, + "step": 7077 + }, + { + "epoch": 0.586655615416494, + "grad_norm": 0.26541149616241455, + "learning_rate": 2.0671363447990055e-06, + "loss": 0.5464, + "step": 7078 + }, + { + "epoch": 0.586738499792789, + "grad_norm": 0.2763189375400543, + "learning_rate": 2.0667219229175304e-06, + "loss": 0.5622, + "step": 7079 + }, + { + "epoch": 0.5868213841690841, + "grad_norm": 0.26275181770324707, + "learning_rate": 2.066307501036055e-06, + "loss": 0.5285, + "step": 7080 + }, + { + "epoch": 0.5869042685453792, + "grad_norm": 0.24810324609279633, + "learning_rate": 2.0658930791545797e-06, + "loss": 0.5075, + "step": 7081 + }, + { + "epoch": 0.5869871529216742, + "grad_norm": 0.2494848519563675, + "learning_rate": 2.065478657273104e-06, + "loss": 0.5465, + "step": 7082 + }, + { + "epoch": 0.5870700372979694, + "grad_norm": 0.2554476261138916, + "learning_rate": 2.065064235391629e-06, + "loss": 0.494, + "step": 7083 + }, + { + "epoch": 0.5871529216742644, + "grad_norm": 0.23384955525398254, + "learning_rate": 2.0646498135101534e-06, + "loss": 0.4642, + "step": 7084 + }, + { + "epoch": 0.5872358060505595, + "grad_norm": 0.26285260915756226, + "learning_rate": 2.0642353916286783e-06, + "loss": 0.5312, + "step": 7085 + }, + { + "epoch": 0.5873186904268546, + "grad_norm": 0.22776228189468384, + "learning_rate": 2.0638209697472027e-06, + "loss": 0.4862, + "step": 7086 + }, + { + "epoch": 0.5874015748031496, + "grad_norm": 0.25582006573677063, + "learning_rate": 2.0634065478657275e-06, + "loss": 0.5215, + "step": 7087 + }, + { + "epoch": 0.5874844591794447, + "grad_norm": 0.2890077233314514, + "learning_rate": 2.0629921259842524e-06, + "loss": 0.4709, + "step": 7088 + }, + { + "epoch": 0.5875673435557397, + "grad_norm": 0.2580912113189697, + "learning_rate": 2.062577704102777e-06, + "loss": 0.5462, + "step": 7089 + }, + { + "epoch": 0.5876502279320348, + "grad_norm": 0.23824095726013184, + "learning_rate": 2.0621632822213013e-06, + "loss": 0.4723, + "step": 7090 + }, + { + "epoch": 0.5877331123083299, + "grad_norm": 0.25006887316703796, + "learning_rate": 2.061748860339826e-06, + "loss": 0.5109, + "step": 7091 + }, + { + "epoch": 0.5878159966846249, + "grad_norm": 0.2542170286178589, + "learning_rate": 2.061334438458351e-06, + "loss": 0.5481, + "step": 7092 + }, + { + "epoch": 0.58789888106092, + "grad_norm": 0.266501784324646, + "learning_rate": 2.0609200165768754e-06, + "loss": 0.5029, + "step": 7093 + }, + { + "epoch": 0.587981765437215, + "grad_norm": 0.22816888988018036, + "learning_rate": 2.0605055946954e-06, + "loss": 0.4759, + "step": 7094 + }, + { + "epoch": 0.5880646498135101, + "grad_norm": 0.2662350833415985, + "learning_rate": 2.0600911728139247e-06, + "loss": 0.5458, + "step": 7095 + }, + { + "epoch": 0.5881475341898053, + "grad_norm": 0.24926501512527466, + "learning_rate": 2.0596767509324495e-06, + "loss": 0.5019, + "step": 7096 + }, + { + "epoch": 0.5882304185661003, + "grad_norm": 0.2516896426677704, + "learning_rate": 2.059262329050974e-06, + "loss": 0.5102, + "step": 7097 + }, + { + "epoch": 0.5883133029423954, + "grad_norm": 0.27426382899284363, + "learning_rate": 2.0588479071694984e-06, + "loss": 0.4924, + "step": 7098 + }, + { + "epoch": 0.5883961873186905, + "grad_norm": 0.24528458714485168, + "learning_rate": 2.0584334852880233e-06, + "loss": 0.5054, + "step": 7099 + }, + { + "epoch": 0.5884790716949855, + "grad_norm": 0.23953644931316376, + "learning_rate": 2.058019063406548e-06, + "loss": 0.5052, + "step": 7100 + }, + { + "epoch": 0.5885619560712806, + "grad_norm": 0.2551461160182953, + "learning_rate": 2.057604641525073e-06, + "loss": 0.5142, + "step": 7101 + }, + { + "epoch": 0.5886448404475756, + "grad_norm": 0.23514710366725922, + "learning_rate": 2.0571902196435974e-06, + "loss": 0.4916, + "step": 7102 + }, + { + "epoch": 0.5887277248238707, + "grad_norm": 0.23652450740337372, + "learning_rate": 2.056775797762122e-06, + "loss": 0.4909, + "step": 7103 + }, + { + "epoch": 0.5888106092001658, + "grad_norm": 0.2377953827381134, + "learning_rate": 2.0563613758806467e-06, + "loss": 0.4739, + "step": 7104 + }, + { + "epoch": 0.5888934935764608, + "grad_norm": 0.24432353675365448, + "learning_rate": 2.0559469539991715e-06, + "loss": 0.5045, + "step": 7105 + }, + { + "epoch": 0.5889763779527559, + "grad_norm": 0.25374001264572144, + "learning_rate": 2.055532532117696e-06, + "loss": 0.4898, + "step": 7106 + }, + { + "epoch": 0.5890592623290509, + "grad_norm": 0.2532568871974945, + "learning_rate": 2.0551181102362204e-06, + "loss": 0.5458, + "step": 7107 + }, + { + "epoch": 0.589142146705346, + "grad_norm": 0.24246250092983246, + "learning_rate": 2.0547036883547453e-06, + "loss": 0.4827, + "step": 7108 + }, + { + "epoch": 0.5892250310816411, + "grad_norm": 0.2488243281841278, + "learning_rate": 2.05428926647327e-06, + "loss": 0.4864, + "step": 7109 + }, + { + "epoch": 0.5893079154579361, + "grad_norm": 0.25663700699806213, + "learning_rate": 2.0538748445917946e-06, + "loss": 0.5444, + "step": 7110 + }, + { + "epoch": 0.5893907998342313, + "grad_norm": 0.2334025800228119, + "learning_rate": 2.053460422710319e-06, + "loss": 0.5338, + "step": 7111 + }, + { + "epoch": 0.5894736842105263, + "grad_norm": 0.25552383065223694, + "learning_rate": 2.053046000828844e-06, + "loss": 0.4698, + "step": 7112 + }, + { + "epoch": 0.5895565685868214, + "grad_norm": 0.24929475784301758, + "learning_rate": 2.0526315789473687e-06, + "loss": 0.5404, + "step": 7113 + }, + { + "epoch": 0.5896394529631165, + "grad_norm": 0.2369585782289505, + "learning_rate": 2.052217157065893e-06, + "loss": 0.5314, + "step": 7114 + }, + { + "epoch": 0.5897223373394115, + "grad_norm": 0.24345029890537262, + "learning_rate": 2.051802735184418e-06, + "loss": 0.4573, + "step": 7115 + }, + { + "epoch": 0.5898052217157066, + "grad_norm": 0.22991551458835602, + "learning_rate": 2.0513883133029424e-06, + "loss": 0.4668, + "step": 7116 + }, + { + "epoch": 0.5898881060920017, + "grad_norm": 0.23583443462848663, + "learning_rate": 2.0509738914214673e-06, + "loss": 0.4732, + "step": 7117 + }, + { + "epoch": 0.5899709904682967, + "grad_norm": 0.25125980377197266, + "learning_rate": 2.0505594695399917e-06, + "loss": 0.5436, + "step": 7118 + }, + { + "epoch": 0.5900538748445918, + "grad_norm": 0.2282799631357193, + "learning_rate": 2.0501450476585166e-06, + "loss": 0.4451, + "step": 7119 + }, + { + "epoch": 0.5901367592208868, + "grad_norm": 0.2500872313976288, + "learning_rate": 2.049730625777041e-06, + "loss": 0.5275, + "step": 7120 + }, + { + "epoch": 0.5902196435971819, + "grad_norm": 0.2458445131778717, + "learning_rate": 2.049316203895566e-06, + "loss": 0.5101, + "step": 7121 + }, + { + "epoch": 0.590302527973477, + "grad_norm": 0.23951749503612518, + "learning_rate": 2.0489017820140907e-06, + "loss": 0.4734, + "step": 7122 + }, + { + "epoch": 0.590385412349772, + "grad_norm": 0.24508202075958252, + "learning_rate": 2.048487360132615e-06, + "loss": 0.5089, + "step": 7123 + }, + { + "epoch": 0.5904682967260672, + "grad_norm": 0.4416239857673645, + "learning_rate": 2.04807293825114e-06, + "loss": 0.4745, + "step": 7124 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 0.2517060935497284, + "learning_rate": 2.0476585163696644e-06, + "loss": 0.5045, + "step": 7125 + }, + { + "epoch": 0.5906340654786573, + "grad_norm": 0.2561684548854828, + "learning_rate": 2.0472440944881893e-06, + "loss": 0.5132, + "step": 7126 + }, + { + "epoch": 0.5907169498549524, + "grad_norm": 0.23915694653987885, + "learning_rate": 2.0468296726067137e-06, + "loss": 0.5002, + "step": 7127 + }, + { + "epoch": 0.5907998342312474, + "grad_norm": 0.2521812319755554, + "learning_rate": 2.0464152507252386e-06, + "loss": 0.4791, + "step": 7128 + }, + { + "epoch": 0.5908827186075425, + "grad_norm": 0.27208375930786133, + "learning_rate": 2.046000828843763e-06, + "loss": 0.546, + "step": 7129 + }, + { + "epoch": 0.5909656029838376, + "grad_norm": 0.2670110762119293, + "learning_rate": 2.045586406962288e-06, + "loss": 0.5093, + "step": 7130 + }, + { + "epoch": 0.5910484873601326, + "grad_norm": 0.2515048384666443, + "learning_rate": 2.0451719850808123e-06, + "loss": 0.5182, + "step": 7131 + }, + { + "epoch": 0.5911313717364277, + "grad_norm": 0.2619059681892395, + "learning_rate": 2.044757563199337e-06, + "loss": 0.5163, + "step": 7132 + }, + { + "epoch": 0.5912142561127227, + "grad_norm": 0.27125146985054016, + "learning_rate": 2.044343141317862e-06, + "loss": 0.5507, + "step": 7133 + }, + { + "epoch": 0.5912971404890178, + "grad_norm": 0.25460928678512573, + "learning_rate": 2.0439287194363864e-06, + "loss": 0.5275, + "step": 7134 + }, + { + "epoch": 0.5913800248653129, + "grad_norm": 0.25562402606010437, + "learning_rate": 2.043514297554911e-06, + "loss": 0.5464, + "step": 7135 + }, + { + "epoch": 0.5914629092416079, + "grad_norm": 0.24707679450511932, + "learning_rate": 2.0430998756734357e-06, + "loss": 0.5006, + "step": 7136 + }, + { + "epoch": 0.591545793617903, + "grad_norm": 0.23694053292274475, + "learning_rate": 2.0426854537919606e-06, + "loss": 0.5208, + "step": 7137 + }, + { + "epoch": 0.591628677994198, + "grad_norm": 0.2629997730255127, + "learning_rate": 2.042271031910485e-06, + "loss": 0.5078, + "step": 7138 + }, + { + "epoch": 0.5917115623704932, + "grad_norm": 0.24393396079540253, + "learning_rate": 2.04185661002901e-06, + "loss": 0.5337, + "step": 7139 + }, + { + "epoch": 0.5917944467467883, + "grad_norm": 0.2364320009946823, + "learning_rate": 2.0414421881475343e-06, + "loss": 0.5128, + "step": 7140 + }, + { + "epoch": 0.5918773311230833, + "grad_norm": 0.23367220163345337, + "learning_rate": 2.041027766266059e-06, + "loss": 0.4792, + "step": 7141 + }, + { + "epoch": 0.5919602154993784, + "grad_norm": 0.26399385929107666, + "learning_rate": 2.040613344384584e-06, + "loss": 0.521, + "step": 7142 + }, + { + "epoch": 0.5920430998756734, + "grad_norm": 0.25298821926116943, + "learning_rate": 2.0401989225031084e-06, + "loss": 0.5338, + "step": 7143 + }, + { + "epoch": 0.5921259842519685, + "grad_norm": 0.26767560839653015, + "learning_rate": 2.039784500621633e-06, + "loss": 0.5426, + "step": 7144 + }, + { + "epoch": 0.5922088686282636, + "grad_norm": 0.22602687776088715, + "learning_rate": 2.0393700787401577e-06, + "loss": 0.4724, + "step": 7145 + }, + { + "epoch": 0.5922917530045586, + "grad_norm": 0.23066291213035583, + "learning_rate": 2.0389556568586826e-06, + "loss": 0.4902, + "step": 7146 + }, + { + "epoch": 0.5923746373808537, + "grad_norm": 0.2534584403038025, + "learning_rate": 2.038541234977207e-06, + "loss": 0.5592, + "step": 7147 + }, + { + "epoch": 0.5924575217571488, + "grad_norm": 0.25279468297958374, + "learning_rate": 2.0381268130957314e-06, + "loss": 0.503, + "step": 7148 + }, + { + "epoch": 0.5925404061334438, + "grad_norm": 0.24566665291786194, + "learning_rate": 2.0377123912142563e-06, + "loss": 0.5325, + "step": 7149 + }, + { + "epoch": 0.5926232905097389, + "grad_norm": 0.25009801983833313, + "learning_rate": 2.037297969332781e-06, + "loss": 0.5265, + "step": 7150 + }, + { + "epoch": 0.5927061748860339, + "grad_norm": 0.2644427716732025, + "learning_rate": 2.0368835474513056e-06, + "loss": 0.5027, + "step": 7151 + }, + { + "epoch": 0.592789059262329, + "grad_norm": 0.2641938626766205, + "learning_rate": 2.03646912556983e-06, + "loss": 0.498, + "step": 7152 + }, + { + "epoch": 0.5928719436386242, + "grad_norm": 0.2512242794036865, + "learning_rate": 2.036054703688355e-06, + "loss": 0.5033, + "step": 7153 + }, + { + "epoch": 0.5929548280149192, + "grad_norm": 0.23715555667877197, + "learning_rate": 2.0356402818068797e-06, + "loss": 0.5027, + "step": 7154 + }, + { + "epoch": 0.5930377123912143, + "grad_norm": 0.2728380858898163, + "learning_rate": 2.035225859925404e-06, + "loss": 0.5844, + "step": 7155 + }, + { + "epoch": 0.5931205967675093, + "grad_norm": 0.254942923784256, + "learning_rate": 2.0348114380439286e-06, + "loss": 0.4933, + "step": 7156 + }, + { + "epoch": 0.5932034811438044, + "grad_norm": 0.24977999925613403, + "learning_rate": 2.0343970161624534e-06, + "loss": 0.4812, + "step": 7157 + }, + { + "epoch": 0.5932863655200995, + "grad_norm": 0.2468833327293396, + "learning_rate": 2.0339825942809783e-06, + "loss": 0.5002, + "step": 7158 + }, + { + "epoch": 0.5933692498963945, + "grad_norm": 0.25220105051994324, + "learning_rate": 2.033568172399503e-06, + "loss": 0.4933, + "step": 7159 + }, + { + "epoch": 0.5934521342726896, + "grad_norm": 0.2557823359966278, + "learning_rate": 2.0331537505180276e-06, + "loss": 0.5186, + "step": 7160 + }, + { + "epoch": 0.5935350186489846, + "grad_norm": 0.24152763187885284, + "learning_rate": 2.032739328636552e-06, + "loss": 0.5078, + "step": 7161 + }, + { + "epoch": 0.5936179030252797, + "grad_norm": 0.2746153473854065, + "learning_rate": 2.032324906755077e-06, + "loss": 0.5779, + "step": 7162 + }, + { + "epoch": 0.5937007874015748, + "grad_norm": 0.24602672457695007, + "learning_rate": 2.0319104848736017e-06, + "loss": 0.4935, + "step": 7163 + }, + { + "epoch": 0.5937836717778698, + "grad_norm": 0.2525305449962616, + "learning_rate": 2.031496062992126e-06, + "loss": 0.5156, + "step": 7164 + }, + { + "epoch": 0.593866556154165, + "grad_norm": 0.2548738420009613, + "learning_rate": 2.0310816411106506e-06, + "loss": 0.5458, + "step": 7165 + }, + { + "epoch": 0.5939494405304601, + "grad_norm": 0.2462504804134369, + "learning_rate": 2.0306672192291754e-06, + "loss": 0.516, + "step": 7166 + }, + { + "epoch": 0.5940323249067551, + "grad_norm": 0.23627308011054993, + "learning_rate": 2.0302527973477003e-06, + "loss": 0.4872, + "step": 7167 + }, + { + "epoch": 0.5941152092830502, + "grad_norm": 0.24238628149032593, + "learning_rate": 2.0298383754662247e-06, + "loss": 0.5349, + "step": 7168 + }, + { + "epoch": 0.5941980936593452, + "grad_norm": 0.2517055869102478, + "learning_rate": 2.029423953584749e-06, + "loss": 0.5261, + "step": 7169 + }, + { + "epoch": 0.5942809780356403, + "grad_norm": 0.23596324026584625, + "learning_rate": 2.029009531703274e-06, + "loss": 0.5013, + "step": 7170 + }, + { + "epoch": 0.5943638624119354, + "grad_norm": 0.2644835412502289, + "learning_rate": 2.028595109821799e-06, + "loss": 0.5469, + "step": 7171 + }, + { + "epoch": 0.5944467467882304, + "grad_norm": 0.23013754189014435, + "learning_rate": 2.0281806879403233e-06, + "loss": 0.5085, + "step": 7172 + }, + { + "epoch": 0.5945296311645255, + "grad_norm": 0.26325729489326477, + "learning_rate": 2.027766266058848e-06, + "loss": 0.5032, + "step": 7173 + }, + { + "epoch": 0.5946125155408205, + "grad_norm": 0.25173231959342957, + "learning_rate": 2.0273518441773726e-06, + "loss": 0.4737, + "step": 7174 + }, + { + "epoch": 0.5946953999171156, + "grad_norm": 0.2677830457687378, + "learning_rate": 2.0269374222958974e-06, + "loss": 0.5437, + "step": 7175 + }, + { + "epoch": 0.5947782842934107, + "grad_norm": 0.25447332859039307, + "learning_rate": 2.026523000414422e-06, + "loss": 0.5207, + "step": 7176 + }, + { + "epoch": 0.5948611686697057, + "grad_norm": 0.23260942101478577, + "learning_rate": 2.0261085785329467e-06, + "loss": 0.4815, + "step": 7177 + }, + { + "epoch": 0.5949440530460008, + "grad_norm": 0.23927584290504456, + "learning_rate": 2.025694156651471e-06, + "loss": 0.4722, + "step": 7178 + }, + { + "epoch": 0.5950269374222958, + "grad_norm": 0.24127982556819916, + "learning_rate": 2.025279734769996e-06, + "loss": 0.5051, + "step": 7179 + }, + { + "epoch": 0.595109821798591, + "grad_norm": 0.25573161244392395, + "learning_rate": 2.024865312888521e-06, + "loss": 0.5315, + "step": 7180 + }, + { + "epoch": 0.5951927061748861, + "grad_norm": 0.2412177324295044, + "learning_rate": 2.0244508910070453e-06, + "loss": 0.475, + "step": 7181 + }, + { + "epoch": 0.5952755905511811, + "grad_norm": 0.25961607694625854, + "learning_rate": 2.02403646912557e-06, + "loss": 0.5219, + "step": 7182 + }, + { + "epoch": 0.5953584749274762, + "grad_norm": 0.22433798015117645, + "learning_rate": 2.0236220472440946e-06, + "loss": 0.4416, + "step": 7183 + }, + { + "epoch": 0.5954413593037713, + "grad_norm": 0.23862871527671814, + "learning_rate": 2.0232076253626194e-06, + "loss": 0.5005, + "step": 7184 + }, + { + "epoch": 0.5955242436800663, + "grad_norm": 0.2506169378757477, + "learning_rate": 2.022793203481144e-06, + "loss": 0.5067, + "step": 7185 + }, + { + "epoch": 0.5956071280563614, + "grad_norm": 0.24941295385360718, + "learning_rate": 2.0223787815996687e-06, + "loss": 0.5172, + "step": 7186 + }, + { + "epoch": 0.5956900124326564, + "grad_norm": 0.2580548822879791, + "learning_rate": 2.021964359718193e-06, + "loss": 0.4673, + "step": 7187 + }, + { + "epoch": 0.5957728968089515, + "grad_norm": 0.256923109292984, + "learning_rate": 2.021549937836718e-06, + "loss": 0.5292, + "step": 7188 + }, + { + "epoch": 0.5958557811852466, + "grad_norm": 0.22918663918972015, + "learning_rate": 2.0211355159552425e-06, + "loss": 0.5023, + "step": 7189 + }, + { + "epoch": 0.5959386655615416, + "grad_norm": 0.2430020570755005, + "learning_rate": 2.0207210940737673e-06, + "loss": 0.5438, + "step": 7190 + }, + { + "epoch": 0.5960215499378367, + "grad_norm": 0.2416132092475891, + "learning_rate": 2.020306672192292e-06, + "loss": 0.4302, + "step": 7191 + }, + { + "epoch": 0.5961044343141317, + "grad_norm": 0.2606355547904968, + "learning_rate": 2.0198922503108166e-06, + "loss": 0.5242, + "step": 7192 + }, + { + "epoch": 0.5961873186904268, + "grad_norm": 0.2469773292541504, + "learning_rate": 2.019477828429341e-06, + "loss": 0.5146, + "step": 7193 + }, + { + "epoch": 0.596270203066722, + "grad_norm": 0.23135803639888763, + "learning_rate": 2.019063406547866e-06, + "loss": 0.4753, + "step": 7194 + }, + { + "epoch": 0.596353087443017, + "grad_norm": 0.24678289890289307, + "learning_rate": 2.0186489846663907e-06, + "loss": 0.502, + "step": 7195 + }, + { + "epoch": 0.5964359718193121, + "grad_norm": 0.2497129887342453, + "learning_rate": 2.018234562784915e-06, + "loss": 0.55, + "step": 7196 + }, + { + "epoch": 0.5965188561956072, + "grad_norm": 0.23194198310375214, + "learning_rate": 2.0178201409034396e-06, + "loss": 0.4564, + "step": 7197 + }, + { + "epoch": 0.5966017405719022, + "grad_norm": 0.24830985069274902, + "learning_rate": 2.0174057190219645e-06, + "loss": 0.4716, + "step": 7198 + }, + { + "epoch": 0.5966846249481973, + "grad_norm": 0.2368018478155136, + "learning_rate": 2.0169912971404893e-06, + "loss": 0.4961, + "step": 7199 + }, + { + "epoch": 0.5967675093244923, + "grad_norm": 0.27191850543022156, + "learning_rate": 2.016576875259014e-06, + "loss": 0.5404, + "step": 7200 + }, + { + "epoch": 0.5968503937007874, + "grad_norm": 0.24760980904102325, + "learning_rate": 2.0161624533775386e-06, + "loss": 0.4944, + "step": 7201 + }, + { + "epoch": 0.5969332780770825, + "grad_norm": 0.24840272963047028, + "learning_rate": 2.015748031496063e-06, + "loss": 0.5179, + "step": 7202 + }, + { + "epoch": 0.5970161624533775, + "grad_norm": 0.23617996275424957, + "learning_rate": 2.015333609614588e-06, + "loss": 0.4847, + "step": 7203 + }, + { + "epoch": 0.5970990468296726, + "grad_norm": 0.2423424869775772, + "learning_rate": 2.0149191877331127e-06, + "loss": 0.5059, + "step": 7204 + }, + { + "epoch": 0.5971819312059676, + "grad_norm": 0.26880550384521484, + "learning_rate": 2.014504765851637e-06, + "loss": 0.5232, + "step": 7205 + }, + { + "epoch": 0.5972648155822627, + "grad_norm": 0.26396694779396057, + "learning_rate": 2.0140903439701616e-06, + "loss": 0.5177, + "step": 7206 + }, + { + "epoch": 0.5973476999585579, + "grad_norm": 0.24279367923736572, + "learning_rate": 2.0136759220886865e-06, + "loss": 0.4492, + "step": 7207 + }, + { + "epoch": 0.5974305843348529, + "grad_norm": 0.2182420939207077, + "learning_rate": 2.0132615002072113e-06, + "loss": 0.4559, + "step": 7208 + }, + { + "epoch": 0.597513468711148, + "grad_norm": 0.2548601031303406, + "learning_rate": 2.0128470783257357e-06, + "loss": 0.5278, + "step": 7209 + }, + { + "epoch": 0.597596353087443, + "grad_norm": 0.25321441888809204, + "learning_rate": 2.01243265644426e-06, + "loss": 0.5048, + "step": 7210 + }, + { + "epoch": 0.5976792374637381, + "grad_norm": 0.24233996868133545, + "learning_rate": 2.012018234562785e-06, + "loss": 0.4744, + "step": 7211 + }, + { + "epoch": 0.5977621218400332, + "grad_norm": 0.26476359367370605, + "learning_rate": 2.01160381268131e-06, + "loss": 0.5272, + "step": 7212 + }, + { + "epoch": 0.5978450062163282, + "grad_norm": 0.24120981991291046, + "learning_rate": 2.0111893907998343e-06, + "loss": 0.4854, + "step": 7213 + }, + { + "epoch": 0.5979278905926233, + "grad_norm": 0.259945273399353, + "learning_rate": 2.0107749689183588e-06, + "loss": 0.5317, + "step": 7214 + }, + { + "epoch": 0.5980107749689184, + "grad_norm": 0.2367905229330063, + "learning_rate": 2.0103605470368836e-06, + "loss": 0.5267, + "step": 7215 + }, + { + "epoch": 0.5980936593452134, + "grad_norm": 0.26913875341415405, + "learning_rate": 2.0099461251554085e-06, + "loss": 0.492, + "step": 7216 + }, + { + "epoch": 0.5981765437215085, + "grad_norm": 0.2668052613735199, + "learning_rate": 2.009531703273933e-06, + "loss": 0.4966, + "step": 7217 + }, + { + "epoch": 0.5982594280978035, + "grad_norm": 0.2524052858352661, + "learning_rate": 2.0091172813924578e-06, + "loss": 0.5207, + "step": 7218 + }, + { + "epoch": 0.5983423124740986, + "grad_norm": 0.23523323237895966, + "learning_rate": 2.008702859510982e-06, + "loss": 0.5148, + "step": 7219 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 0.25379422307014465, + "learning_rate": 2.008288437629507e-06, + "loss": 0.4997, + "step": 7220 + }, + { + "epoch": 0.5985080812266887, + "grad_norm": 0.2646865248680115, + "learning_rate": 2.007874015748032e-06, + "loss": 0.5112, + "step": 7221 + }, + { + "epoch": 0.5985909656029839, + "grad_norm": 0.24566394090652466, + "learning_rate": 2.0074595938665563e-06, + "loss": 0.528, + "step": 7222 + }, + { + "epoch": 0.5986738499792789, + "grad_norm": 0.25786566734313965, + "learning_rate": 2.0070451719850808e-06, + "loss": 0.516, + "step": 7223 + }, + { + "epoch": 0.598756734355574, + "grad_norm": 0.23459851741790771, + "learning_rate": 2.0066307501036056e-06, + "loss": 0.4908, + "step": 7224 + }, + { + "epoch": 0.5988396187318691, + "grad_norm": 0.3547213077545166, + "learning_rate": 2.0062163282221305e-06, + "loss": 0.5895, + "step": 7225 + }, + { + "epoch": 0.5989225031081641, + "grad_norm": 0.2473573237657547, + "learning_rate": 2.005801906340655e-06, + "loss": 0.4905, + "step": 7226 + }, + { + "epoch": 0.5990053874844592, + "grad_norm": 0.24339185655117035, + "learning_rate": 2.0053874844591798e-06, + "loss": 0.5124, + "step": 7227 + }, + { + "epoch": 0.5990882718607542, + "grad_norm": 0.24199272692203522, + "learning_rate": 2.004973062577704e-06, + "loss": 0.5052, + "step": 7228 + }, + { + "epoch": 0.5991711562370493, + "grad_norm": 0.2280384749174118, + "learning_rate": 2.004558640696229e-06, + "loss": 0.4535, + "step": 7229 + }, + { + "epoch": 0.5992540406133444, + "grad_norm": 0.2599889934062958, + "learning_rate": 2.0041442188147535e-06, + "loss": 0.5303, + "step": 7230 + }, + { + "epoch": 0.5993369249896394, + "grad_norm": 0.24235686659812927, + "learning_rate": 2.0037297969332783e-06, + "loss": 0.44, + "step": 7231 + }, + { + "epoch": 0.5994198093659345, + "grad_norm": 0.2694646120071411, + "learning_rate": 2.0033153750518028e-06, + "loss": 0.5709, + "step": 7232 + }, + { + "epoch": 0.5995026937422296, + "grad_norm": 0.24260428547859192, + "learning_rate": 2.0029009531703276e-06, + "loss": 0.4875, + "step": 7233 + }, + { + "epoch": 0.5995855781185246, + "grad_norm": 0.24749921262264252, + "learning_rate": 2.002486531288852e-06, + "loss": 0.5085, + "step": 7234 + }, + { + "epoch": 0.5996684624948198, + "grad_norm": 0.23846492171287537, + "learning_rate": 2.002072109407377e-06, + "loss": 0.5146, + "step": 7235 + }, + { + "epoch": 0.5997513468711148, + "grad_norm": 0.24476668238639832, + "learning_rate": 2.0016576875259013e-06, + "loss": 0.4801, + "step": 7236 + }, + { + "epoch": 0.5998342312474099, + "grad_norm": 0.2375708967447281, + "learning_rate": 2.001243265644426e-06, + "loss": 0.4823, + "step": 7237 + }, + { + "epoch": 0.599917115623705, + "grad_norm": 0.29655495285987854, + "learning_rate": 2.000828843762951e-06, + "loss": 0.5719, + "step": 7238 + }, + { + "epoch": 0.6, + "grad_norm": 0.23830047249794006, + "learning_rate": 2.0004144218814755e-06, + "loss": 0.4655, + "step": 7239 + }, + { + "epoch": 0.6000828843762951, + "grad_norm": 0.24259479343891144, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.487, + "step": 7240 + }, + { + "epoch": 0.6001657687525901, + "grad_norm": 0.24634137749671936, + "learning_rate": 1.9995855781185248e-06, + "loss": 0.5186, + "step": 7241 + }, + { + "epoch": 0.6002486531288852, + "grad_norm": 0.2844352424144745, + "learning_rate": 1.9991711562370496e-06, + "loss": 0.5213, + "step": 7242 + }, + { + "epoch": 0.6003315375051803, + "grad_norm": 0.2541354298591614, + "learning_rate": 1.998756734355574e-06, + "loss": 0.5214, + "step": 7243 + }, + { + "epoch": 0.6004144218814753, + "grad_norm": 0.2775161862373352, + "learning_rate": 1.998342312474099e-06, + "loss": 0.5435, + "step": 7244 + }, + { + "epoch": 0.6004973062577704, + "grad_norm": 0.24239714443683624, + "learning_rate": 1.9979278905926233e-06, + "loss": 0.5392, + "step": 7245 + }, + { + "epoch": 0.6005801906340654, + "grad_norm": 0.23084254562854767, + "learning_rate": 1.997513468711148e-06, + "loss": 0.4863, + "step": 7246 + }, + { + "epoch": 0.6006630750103605, + "grad_norm": 0.2523568570613861, + "learning_rate": 1.9970990468296726e-06, + "loss": 0.5215, + "step": 7247 + }, + { + "epoch": 0.6007459593866556, + "grad_norm": 0.273552268743515, + "learning_rate": 1.9966846249481975e-06, + "loss": 0.5105, + "step": 7248 + }, + { + "epoch": 0.6008288437629506, + "grad_norm": 0.24874143302440643, + "learning_rate": 1.9962702030667223e-06, + "loss": 0.5428, + "step": 7249 + }, + { + "epoch": 0.6009117281392458, + "grad_norm": 0.24535632133483887, + "learning_rate": 1.9958557811852468e-06, + "loss": 0.4932, + "step": 7250 + }, + { + "epoch": 0.6009946125155409, + "grad_norm": 0.2434893399477005, + "learning_rate": 1.995441359303771e-06, + "loss": 0.5089, + "step": 7251 + }, + { + "epoch": 0.6010774968918359, + "grad_norm": 0.23096027970314026, + "learning_rate": 1.995026937422296e-06, + "loss": 0.5146, + "step": 7252 + }, + { + "epoch": 0.601160381268131, + "grad_norm": 0.2682402729988098, + "learning_rate": 1.994612515540821e-06, + "loss": 0.5275, + "step": 7253 + }, + { + "epoch": 0.601243265644426, + "grad_norm": 0.24476107954978943, + "learning_rate": 1.9941980936593453e-06, + "loss": 0.5022, + "step": 7254 + }, + { + "epoch": 0.6013261500207211, + "grad_norm": 0.25766220688819885, + "learning_rate": 1.9937836717778698e-06, + "loss": 0.5305, + "step": 7255 + }, + { + "epoch": 0.6014090343970162, + "grad_norm": 0.2475586235523224, + "learning_rate": 1.9933692498963946e-06, + "loss": 0.5468, + "step": 7256 + }, + { + "epoch": 0.6014919187733112, + "grad_norm": 0.2656251788139343, + "learning_rate": 1.9929548280149195e-06, + "loss": 0.4649, + "step": 7257 + }, + { + "epoch": 0.6015748031496063, + "grad_norm": 0.2491811364889145, + "learning_rate": 1.992540406133444e-06, + "loss": 0.5405, + "step": 7258 + }, + { + "epoch": 0.6016576875259013, + "grad_norm": 0.2699982523918152, + "learning_rate": 1.9921259842519688e-06, + "loss": 0.5067, + "step": 7259 + }, + { + "epoch": 0.6017405719021964, + "grad_norm": 0.2554823160171509, + "learning_rate": 1.991711562370493e-06, + "loss": 0.469, + "step": 7260 + }, + { + "epoch": 0.6018234562784915, + "grad_norm": 0.24127066135406494, + "learning_rate": 1.991297140489018e-06, + "loss": 0.4983, + "step": 7261 + }, + { + "epoch": 0.6019063406547865, + "grad_norm": 0.253755658864975, + "learning_rate": 1.990882718607543e-06, + "loss": 0.5068, + "step": 7262 + }, + { + "epoch": 0.6019892250310817, + "grad_norm": 0.2487199604511261, + "learning_rate": 1.9904682967260673e-06, + "loss": 0.5098, + "step": 7263 + }, + { + "epoch": 0.6020721094073767, + "grad_norm": 0.25411364436149597, + "learning_rate": 1.9900538748445918e-06, + "loss": 0.4969, + "step": 7264 + }, + { + "epoch": 0.6021549937836718, + "grad_norm": 0.24313510954380035, + "learning_rate": 1.9896394529631166e-06, + "loss": 0.5386, + "step": 7265 + }, + { + "epoch": 0.6022378781599669, + "grad_norm": 0.24139322340488434, + "learning_rate": 1.9892250310816415e-06, + "loss": 0.5057, + "step": 7266 + }, + { + "epoch": 0.6023207625362619, + "grad_norm": 0.2517094016075134, + "learning_rate": 1.988810609200166e-06, + "loss": 0.4983, + "step": 7267 + }, + { + "epoch": 0.602403646912557, + "grad_norm": 0.2521829307079315, + "learning_rate": 1.9883961873186904e-06, + "loss": 0.4599, + "step": 7268 + }, + { + "epoch": 0.6024865312888521, + "grad_norm": 0.25147804617881775, + "learning_rate": 1.987981765437215e-06, + "loss": 0.4694, + "step": 7269 + }, + { + "epoch": 0.6025694156651471, + "grad_norm": 0.2431638091802597, + "learning_rate": 1.98756734355574e-06, + "loss": 0.5157, + "step": 7270 + }, + { + "epoch": 0.6026523000414422, + "grad_norm": 0.24006785452365875, + "learning_rate": 1.9871529216742645e-06, + "loss": 0.491, + "step": 7271 + }, + { + "epoch": 0.6027351844177372, + "grad_norm": 0.24439837038516998, + "learning_rate": 1.986738499792789e-06, + "loss": 0.4803, + "step": 7272 + }, + { + "epoch": 0.6028180687940323, + "grad_norm": 0.2519010603427887, + "learning_rate": 1.9863240779113138e-06, + "loss": 0.4958, + "step": 7273 + }, + { + "epoch": 0.6029009531703274, + "grad_norm": 0.3182460367679596, + "learning_rate": 1.9859096560298386e-06, + "loss": 0.5081, + "step": 7274 + }, + { + "epoch": 0.6029838375466224, + "grad_norm": 0.26510176062583923, + "learning_rate": 1.985495234148363e-06, + "loss": 0.5844, + "step": 7275 + }, + { + "epoch": 0.6030667219229175, + "grad_norm": 0.27085205912590027, + "learning_rate": 1.985080812266888e-06, + "loss": 0.5236, + "step": 7276 + }, + { + "epoch": 0.6031496062992125, + "grad_norm": 0.26251110434532166, + "learning_rate": 1.9846663903854124e-06, + "loss": 0.4848, + "step": 7277 + }, + { + "epoch": 0.6032324906755077, + "grad_norm": 0.24869200587272644, + "learning_rate": 1.984251968503937e-06, + "loss": 0.5307, + "step": 7278 + }, + { + "epoch": 0.6033153750518028, + "grad_norm": 0.2676868140697479, + "learning_rate": 1.983837546622462e-06, + "loss": 0.5229, + "step": 7279 + }, + { + "epoch": 0.6033982594280978, + "grad_norm": 0.2655852437019348, + "learning_rate": 1.9834231247409865e-06, + "loss": 0.5507, + "step": 7280 + }, + { + "epoch": 0.6034811438043929, + "grad_norm": 0.2519152760505676, + "learning_rate": 1.983008702859511e-06, + "loss": 0.4584, + "step": 7281 + }, + { + "epoch": 0.603564028180688, + "grad_norm": 0.2447376847267151, + "learning_rate": 1.9825942809780358e-06, + "loss": 0.5168, + "step": 7282 + }, + { + "epoch": 0.603646912556983, + "grad_norm": 0.24475589394569397, + "learning_rate": 1.9821798590965606e-06, + "loss": 0.5089, + "step": 7283 + }, + { + "epoch": 0.6037297969332781, + "grad_norm": 0.2440909594297409, + "learning_rate": 1.981765437215085e-06, + "loss": 0.4832, + "step": 7284 + }, + { + "epoch": 0.6038126813095731, + "grad_norm": 0.30234843492507935, + "learning_rate": 1.98135101533361e-06, + "loss": 0.5411, + "step": 7285 + }, + { + "epoch": 0.6038955656858682, + "grad_norm": 0.22577033936977386, + "learning_rate": 1.9809365934521344e-06, + "loss": 0.4956, + "step": 7286 + }, + { + "epoch": 0.6039784500621633, + "grad_norm": 0.24123620986938477, + "learning_rate": 1.9805221715706592e-06, + "loss": 0.4965, + "step": 7287 + }, + { + "epoch": 0.6040613344384583, + "grad_norm": 0.24383677542209625, + "learning_rate": 1.9801077496891836e-06, + "loss": 0.4748, + "step": 7288 + }, + { + "epoch": 0.6041442188147534, + "grad_norm": 0.27585357427597046, + "learning_rate": 1.9796933278077085e-06, + "loss": 0.5328, + "step": 7289 + }, + { + "epoch": 0.6042271031910484, + "grad_norm": 0.26693394780158997, + "learning_rate": 1.979278905926233e-06, + "loss": 0.5103, + "step": 7290 + }, + { + "epoch": 0.6043099875673436, + "grad_norm": 0.2339395135641098, + "learning_rate": 1.978864484044758e-06, + "loss": 0.4621, + "step": 7291 + }, + { + "epoch": 0.6043928719436387, + "grad_norm": 0.25768768787384033, + "learning_rate": 1.9784500621632822e-06, + "loss": 0.5309, + "step": 7292 + }, + { + "epoch": 0.6044757563199337, + "grad_norm": 0.24905137717723846, + "learning_rate": 1.978035640281807e-06, + "loss": 0.4902, + "step": 7293 + }, + { + "epoch": 0.6045586406962288, + "grad_norm": 0.24553972482681274, + "learning_rate": 1.977621218400332e-06, + "loss": 0.4861, + "step": 7294 + }, + { + "epoch": 0.6046415250725238, + "grad_norm": 0.2717408239841461, + "learning_rate": 1.9772067965188564e-06, + "loss": 0.5005, + "step": 7295 + }, + { + "epoch": 0.6047244094488189, + "grad_norm": 0.26099878549575806, + "learning_rate": 1.976792374637381e-06, + "loss": 0.5261, + "step": 7296 + }, + { + "epoch": 0.604807293825114, + "grad_norm": 0.23488135635852814, + "learning_rate": 1.9763779527559057e-06, + "loss": 0.4724, + "step": 7297 + }, + { + "epoch": 0.604890178201409, + "grad_norm": 0.26371946930885315, + "learning_rate": 1.9759635308744305e-06, + "loss": 0.5092, + "step": 7298 + }, + { + "epoch": 0.6049730625777041, + "grad_norm": 0.24398347735404968, + "learning_rate": 1.975549108992955e-06, + "loss": 0.513, + "step": 7299 + }, + { + "epoch": 0.6050559469539992, + "grad_norm": 0.26201242208480835, + "learning_rate": 1.97513468711148e-06, + "loss": 0.5102, + "step": 7300 + }, + { + "epoch": 0.6051388313302942, + "grad_norm": 0.24725355207920074, + "learning_rate": 1.9747202652300042e-06, + "loss": 0.5278, + "step": 7301 + }, + { + "epoch": 0.6052217157065893, + "grad_norm": 0.24889881908893585, + "learning_rate": 1.974305843348529e-06, + "loss": 0.4949, + "step": 7302 + }, + { + "epoch": 0.6053046000828843, + "grad_norm": 0.23751646280288696, + "learning_rate": 1.973891421467054e-06, + "loss": 0.4979, + "step": 7303 + }, + { + "epoch": 0.6053874844591794, + "grad_norm": 0.24814149737358093, + "learning_rate": 1.9734769995855784e-06, + "loss": 0.4878, + "step": 7304 + }, + { + "epoch": 0.6054703688354746, + "grad_norm": 0.2643147110939026, + "learning_rate": 1.973062577704103e-06, + "loss": 0.5201, + "step": 7305 + }, + { + "epoch": 0.6055532532117696, + "grad_norm": 0.27950504422187805, + "learning_rate": 1.9726481558226277e-06, + "loss": 0.5001, + "step": 7306 + }, + { + "epoch": 0.6056361375880647, + "grad_norm": 0.23282644152641296, + "learning_rate": 1.9722337339411525e-06, + "loss": 0.4756, + "step": 7307 + }, + { + "epoch": 0.6057190219643597, + "grad_norm": 0.2604025900363922, + "learning_rate": 1.971819312059677e-06, + "loss": 0.4873, + "step": 7308 + }, + { + "epoch": 0.6058019063406548, + "grad_norm": 0.23811350762844086, + "learning_rate": 1.9714048901782014e-06, + "loss": 0.4914, + "step": 7309 + }, + { + "epoch": 0.6058847907169499, + "grad_norm": 0.23760077357292175, + "learning_rate": 1.9709904682967262e-06, + "loss": 0.4789, + "step": 7310 + }, + { + "epoch": 0.6059676750932449, + "grad_norm": 0.24992050230503082, + "learning_rate": 1.970576046415251e-06, + "loss": 0.4998, + "step": 7311 + }, + { + "epoch": 0.60605055946954, + "grad_norm": 0.22765006124973297, + "learning_rate": 1.9701616245337755e-06, + "loss": 0.4543, + "step": 7312 + }, + { + "epoch": 0.606133443845835, + "grad_norm": 0.26817891001701355, + "learning_rate": 1.9697472026523e-06, + "loss": 0.4761, + "step": 7313 + }, + { + "epoch": 0.6062163282221301, + "grad_norm": 0.2507833242416382, + "learning_rate": 1.969332780770825e-06, + "loss": 0.4974, + "step": 7314 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 0.23670829832553864, + "learning_rate": 1.9689183588893497e-06, + "loss": 0.4986, + "step": 7315 + }, + { + "epoch": 0.6063820969747202, + "grad_norm": 0.25621798634529114, + "learning_rate": 1.968503937007874e-06, + "loss": 0.502, + "step": 7316 + }, + { + "epoch": 0.6064649813510153, + "grad_norm": 0.25322964787483215, + "learning_rate": 1.9680895151263985e-06, + "loss": 0.5626, + "step": 7317 + }, + { + "epoch": 0.6065478657273105, + "grad_norm": 0.24855974316596985, + "learning_rate": 1.9676750932449234e-06, + "loss": 0.4926, + "step": 7318 + }, + { + "epoch": 0.6066307501036055, + "grad_norm": 0.24284352362155914, + "learning_rate": 1.9672606713634482e-06, + "loss": 0.4924, + "step": 7319 + }, + { + "epoch": 0.6067136344799006, + "grad_norm": 0.2741515338420868, + "learning_rate": 1.966846249481973e-06, + "loss": 0.4916, + "step": 7320 + }, + { + "epoch": 0.6067965188561956, + "grad_norm": 0.26332640647888184, + "learning_rate": 1.9664318276004975e-06, + "loss": 0.4922, + "step": 7321 + }, + { + "epoch": 0.6068794032324907, + "grad_norm": 0.24537773430347443, + "learning_rate": 1.966017405719022e-06, + "loss": 0.5066, + "step": 7322 + }, + { + "epoch": 0.6069622876087858, + "grad_norm": 0.2610885798931122, + "learning_rate": 1.965602983837547e-06, + "loss": 0.5258, + "step": 7323 + }, + { + "epoch": 0.6070451719850808, + "grad_norm": 0.24114835262298584, + "learning_rate": 1.9651885619560717e-06, + "loss": 0.511, + "step": 7324 + }, + { + "epoch": 0.6071280563613759, + "grad_norm": 0.27025753259658813, + "learning_rate": 1.964774140074596e-06, + "loss": 0.5455, + "step": 7325 + }, + { + "epoch": 0.6072109407376709, + "grad_norm": 0.24671214818954468, + "learning_rate": 1.9643597181931205e-06, + "loss": 0.5226, + "step": 7326 + }, + { + "epoch": 0.607293825113966, + "grad_norm": 0.2493586242198944, + "learning_rate": 1.9639452963116454e-06, + "loss": 0.5315, + "step": 7327 + }, + { + "epoch": 0.6073767094902611, + "grad_norm": 0.2350810170173645, + "learning_rate": 1.9635308744301702e-06, + "loss": 0.4985, + "step": 7328 + }, + { + "epoch": 0.6074595938665561, + "grad_norm": 0.24529248476028442, + "learning_rate": 1.9631164525486947e-06, + "loss": 0.5124, + "step": 7329 + }, + { + "epoch": 0.6075424782428512, + "grad_norm": 0.2517112195491791, + "learning_rate": 1.962702030667219e-06, + "loss": 0.5027, + "step": 7330 + }, + { + "epoch": 0.6076253626191462, + "grad_norm": 0.2617841958999634, + "learning_rate": 1.962287608785744e-06, + "loss": 0.5194, + "step": 7331 + }, + { + "epoch": 0.6077082469954413, + "grad_norm": 0.25392720103263855, + "learning_rate": 1.961873186904269e-06, + "loss": 0.4695, + "step": 7332 + }, + { + "epoch": 0.6077911313717365, + "grad_norm": 0.26062217354774475, + "learning_rate": 1.9614587650227932e-06, + "loss": 0.4778, + "step": 7333 + }, + { + "epoch": 0.6078740157480315, + "grad_norm": 0.274837851524353, + "learning_rate": 1.961044343141318e-06, + "loss": 0.5873, + "step": 7334 + }, + { + "epoch": 0.6079569001243266, + "grad_norm": 0.23879998922348022, + "learning_rate": 1.9606299212598425e-06, + "loss": 0.4749, + "step": 7335 + }, + { + "epoch": 0.6080397845006217, + "grad_norm": 0.2575720548629761, + "learning_rate": 1.9602154993783674e-06, + "loss": 0.5063, + "step": 7336 + }, + { + "epoch": 0.6081226688769167, + "grad_norm": 0.2468593418598175, + "learning_rate": 1.959801077496892e-06, + "loss": 0.525, + "step": 7337 + }, + { + "epoch": 0.6082055532532118, + "grad_norm": 0.24932661652565002, + "learning_rate": 1.9593866556154167e-06, + "loss": 0.5119, + "step": 7338 + }, + { + "epoch": 0.6082884376295068, + "grad_norm": 0.23979800939559937, + "learning_rate": 1.958972233733941e-06, + "loss": 0.5049, + "step": 7339 + }, + { + "epoch": 0.6083713220058019, + "grad_norm": 0.26612386107444763, + "learning_rate": 1.958557811852466e-06, + "loss": 0.5255, + "step": 7340 + }, + { + "epoch": 0.608454206382097, + "grad_norm": 0.23240657150745392, + "learning_rate": 1.958143389970991e-06, + "loss": 0.4337, + "step": 7341 + }, + { + "epoch": 0.608537090758392, + "grad_norm": 0.26329776644706726, + "learning_rate": 1.9577289680895152e-06, + "loss": 0.52, + "step": 7342 + }, + { + "epoch": 0.6086199751346871, + "grad_norm": 0.24176278710365295, + "learning_rate": 1.95731454620804e-06, + "loss": 0.5181, + "step": 7343 + }, + { + "epoch": 0.6087028595109821, + "grad_norm": 0.24252095818519592, + "learning_rate": 1.9569001243265645e-06, + "loss": 0.5161, + "step": 7344 + }, + { + "epoch": 0.6087857438872772, + "grad_norm": 0.2570688724517822, + "learning_rate": 1.9564857024450894e-06, + "loss": 0.5333, + "step": 7345 + }, + { + "epoch": 0.6088686282635724, + "grad_norm": 0.2429122030735016, + "learning_rate": 1.956071280563614e-06, + "loss": 0.4591, + "step": 7346 + }, + { + "epoch": 0.6089515126398674, + "grad_norm": 0.26712697744369507, + "learning_rate": 1.9556568586821387e-06, + "loss": 0.5521, + "step": 7347 + }, + { + "epoch": 0.6090343970161625, + "grad_norm": 0.2468564510345459, + "learning_rate": 1.955242436800663e-06, + "loss": 0.522, + "step": 7348 + }, + { + "epoch": 0.6091172813924576, + "grad_norm": 0.2540578842163086, + "learning_rate": 1.954828014919188e-06, + "loss": 0.546, + "step": 7349 + }, + { + "epoch": 0.6092001657687526, + "grad_norm": 0.23666951060295105, + "learning_rate": 1.9544135930377124e-06, + "loss": 0.5213, + "step": 7350 + }, + { + "epoch": 0.6092830501450477, + "grad_norm": 0.2538042664527893, + "learning_rate": 1.9539991711562372e-06, + "loss": 0.5085, + "step": 7351 + }, + { + "epoch": 0.6093659345213427, + "grad_norm": 0.24173684418201447, + "learning_rate": 1.953584749274762e-06, + "loss": 0.4904, + "step": 7352 + }, + { + "epoch": 0.6094488188976378, + "grad_norm": 0.23150336742401123, + "learning_rate": 1.9531703273932865e-06, + "loss": 0.4517, + "step": 7353 + }, + { + "epoch": 0.6095317032739329, + "grad_norm": 0.2622227370738983, + "learning_rate": 1.952755905511811e-06, + "loss": 0.5354, + "step": 7354 + }, + { + "epoch": 0.6096145876502279, + "grad_norm": 0.2476886361837387, + "learning_rate": 1.952341483630336e-06, + "loss": 0.506, + "step": 7355 + }, + { + "epoch": 0.609697472026523, + "grad_norm": 0.25034478306770325, + "learning_rate": 1.9519270617488607e-06, + "loss": 0.5059, + "step": 7356 + }, + { + "epoch": 0.609780356402818, + "grad_norm": 0.24372494220733643, + "learning_rate": 1.951512639867385e-06, + "loss": 0.4478, + "step": 7357 + }, + { + "epoch": 0.6098632407791131, + "grad_norm": 0.25975754857063293, + "learning_rate": 1.95109821798591e-06, + "loss": 0.5287, + "step": 7358 + }, + { + "epoch": 0.6099461251554082, + "grad_norm": 0.24329283833503723, + "learning_rate": 1.9506837961044344e-06, + "loss": 0.5092, + "step": 7359 + }, + { + "epoch": 0.6100290095317032, + "grad_norm": 0.25827568769454956, + "learning_rate": 1.9502693742229593e-06, + "loss": 0.5254, + "step": 7360 + }, + { + "epoch": 0.6101118939079984, + "grad_norm": 0.23629237711429596, + "learning_rate": 1.949854952341484e-06, + "loss": 0.4774, + "step": 7361 + }, + { + "epoch": 0.6101947782842934, + "grad_norm": 0.2381962388753891, + "learning_rate": 1.9494405304600085e-06, + "loss": 0.4751, + "step": 7362 + }, + { + "epoch": 0.6102776626605885, + "grad_norm": 0.23836031556129456, + "learning_rate": 1.949026108578533e-06, + "loss": 0.51, + "step": 7363 + }, + { + "epoch": 0.6103605470368836, + "grad_norm": 0.2632398307323456, + "learning_rate": 1.948611686697058e-06, + "loss": 0.5107, + "step": 7364 + }, + { + "epoch": 0.6104434314131786, + "grad_norm": 0.2413366734981537, + "learning_rate": 1.9481972648155827e-06, + "loss": 0.4582, + "step": 7365 + }, + { + "epoch": 0.6105263157894737, + "grad_norm": 0.23222176730632782, + "learning_rate": 1.947782842934107e-06, + "loss": 0.5068, + "step": 7366 + }, + { + "epoch": 0.6106092001657688, + "grad_norm": 0.24522337317466736, + "learning_rate": 1.9473684210526315e-06, + "loss": 0.4859, + "step": 7367 + }, + { + "epoch": 0.6106920845420638, + "grad_norm": 0.2695762515068054, + "learning_rate": 1.9469539991711564e-06, + "loss": 0.5448, + "step": 7368 + }, + { + "epoch": 0.6107749689183589, + "grad_norm": 0.24872125685214996, + "learning_rate": 1.9465395772896813e-06, + "loss": 0.4989, + "step": 7369 + }, + { + "epoch": 0.6108578532946539, + "grad_norm": 0.23196542263031006, + "learning_rate": 1.9461251554082057e-06, + "loss": 0.4991, + "step": 7370 + }, + { + "epoch": 0.610940737670949, + "grad_norm": 0.25505581498146057, + "learning_rate": 1.94571073352673e-06, + "loss": 0.5073, + "step": 7371 + }, + { + "epoch": 0.6110236220472441, + "grad_norm": 0.24392107129096985, + "learning_rate": 1.945296311645255e-06, + "loss": 0.5029, + "step": 7372 + }, + { + "epoch": 0.6111065064235391, + "grad_norm": 0.24102826416492462, + "learning_rate": 1.94488188976378e-06, + "loss": 0.4946, + "step": 7373 + }, + { + "epoch": 0.6111893907998343, + "grad_norm": 0.25109684467315674, + "learning_rate": 1.9444674678823043e-06, + "loss": 0.5038, + "step": 7374 + }, + { + "epoch": 0.6112722751761293, + "grad_norm": 0.255165696144104, + "learning_rate": 1.9440530460008287e-06, + "loss": 0.5027, + "step": 7375 + }, + { + "epoch": 0.6113551595524244, + "grad_norm": 0.2687211334705353, + "learning_rate": 1.9436386241193536e-06, + "loss": 0.5141, + "step": 7376 + }, + { + "epoch": 0.6114380439287195, + "grad_norm": 0.23554135859012604, + "learning_rate": 1.9432242022378784e-06, + "loss": 0.4476, + "step": 7377 + }, + { + "epoch": 0.6115209283050145, + "grad_norm": 0.2520045042037964, + "learning_rate": 1.9428097803564033e-06, + "loss": 0.4938, + "step": 7378 + }, + { + "epoch": 0.6116038126813096, + "grad_norm": 0.23411214351654053, + "learning_rate": 1.9423953584749277e-06, + "loss": 0.4637, + "step": 7379 + }, + { + "epoch": 0.6116866970576046, + "grad_norm": 0.22930866479873657, + "learning_rate": 1.941980936593452e-06, + "loss": 0.4598, + "step": 7380 + }, + { + "epoch": 0.6117695814338997, + "grad_norm": 0.2655690312385559, + "learning_rate": 1.941566514711977e-06, + "loss": 0.5236, + "step": 7381 + }, + { + "epoch": 0.6118524658101948, + "grad_norm": 0.23896177113056183, + "learning_rate": 1.941152092830502e-06, + "loss": 0.4731, + "step": 7382 + }, + { + "epoch": 0.6119353501864898, + "grad_norm": 0.23960784077644348, + "learning_rate": 1.9407376709490263e-06, + "loss": 0.4651, + "step": 7383 + }, + { + "epoch": 0.6120182345627849, + "grad_norm": 0.24571523070335388, + "learning_rate": 1.9403232490675507e-06, + "loss": 0.5068, + "step": 7384 + }, + { + "epoch": 0.61210111893908, + "grad_norm": 0.2591686546802521, + "learning_rate": 1.9399088271860756e-06, + "loss": 0.53, + "step": 7385 + }, + { + "epoch": 0.612184003315375, + "grad_norm": 0.2535070776939392, + "learning_rate": 1.9394944053046004e-06, + "loss": 0.5114, + "step": 7386 + }, + { + "epoch": 0.6122668876916701, + "grad_norm": 0.2577465772628784, + "learning_rate": 1.939079983423125e-06, + "loss": 0.4764, + "step": 7387 + }, + { + "epoch": 0.6123497720679651, + "grad_norm": 0.24453552067279816, + "learning_rate": 1.9386655615416493e-06, + "loss": 0.5118, + "step": 7388 + }, + { + "epoch": 0.6124326564442603, + "grad_norm": 0.23277464509010315, + "learning_rate": 1.938251139660174e-06, + "loss": 0.5022, + "step": 7389 + }, + { + "epoch": 0.6125155408205554, + "grad_norm": 0.24873748421669006, + "learning_rate": 1.937836717778699e-06, + "loss": 0.4918, + "step": 7390 + }, + { + "epoch": 0.6125984251968504, + "grad_norm": 0.23483876883983612, + "learning_rate": 1.9374222958972234e-06, + "loss": 0.4713, + "step": 7391 + }, + { + "epoch": 0.6126813095731455, + "grad_norm": 0.24347984790802002, + "learning_rate": 1.9370078740157483e-06, + "loss": 0.488, + "step": 7392 + }, + { + "epoch": 0.6127641939494405, + "grad_norm": 0.2776225507259369, + "learning_rate": 1.9365934521342727e-06, + "loss": 0.551, + "step": 7393 + }, + { + "epoch": 0.6128470783257356, + "grad_norm": 0.24104759097099304, + "learning_rate": 1.9361790302527976e-06, + "loss": 0.5173, + "step": 7394 + }, + { + "epoch": 0.6129299627020307, + "grad_norm": 0.2390860617160797, + "learning_rate": 1.935764608371322e-06, + "loss": 0.478, + "step": 7395 + }, + { + "epoch": 0.6130128470783257, + "grad_norm": 0.24258475005626678, + "learning_rate": 1.935350186489847e-06, + "loss": 0.5134, + "step": 7396 + }, + { + "epoch": 0.6130957314546208, + "grad_norm": 0.2515009343624115, + "learning_rate": 1.9349357646083713e-06, + "loss": 0.5605, + "step": 7397 + }, + { + "epoch": 0.6131786158309158, + "grad_norm": 0.24492397904396057, + "learning_rate": 1.934521342726896e-06, + "loss": 0.5013, + "step": 7398 + }, + { + "epoch": 0.6132615002072109, + "grad_norm": 0.25059783458709717, + "learning_rate": 1.934106920845421e-06, + "loss": 0.5309, + "step": 7399 + }, + { + "epoch": 0.613344384583506, + "grad_norm": 0.23740354180335999, + "learning_rate": 1.9336924989639454e-06, + "loss": 0.4848, + "step": 7400 + }, + { + "epoch": 0.613427268959801, + "grad_norm": 0.25354132056236267, + "learning_rate": 1.9332780770824703e-06, + "loss": 0.5116, + "step": 7401 + }, + { + "epoch": 0.6135101533360962, + "grad_norm": 0.24052022397518158, + "learning_rate": 1.9328636552009947e-06, + "loss": 0.4886, + "step": 7402 + }, + { + "epoch": 0.6135930377123913, + "grad_norm": 0.25441113114356995, + "learning_rate": 1.9324492333195196e-06, + "loss": 0.5161, + "step": 7403 + }, + { + "epoch": 0.6136759220886863, + "grad_norm": 0.2710849344730377, + "learning_rate": 1.932034811438044e-06, + "loss": 0.5184, + "step": 7404 + }, + { + "epoch": 0.6137588064649814, + "grad_norm": 0.25059476494789124, + "learning_rate": 1.931620389556569e-06, + "loss": 0.4875, + "step": 7405 + }, + { + "epoch": 0.6138416908412764, + "grad_norm": 0.23046742379665375, + "learning_rate": 1.9312059676750933e-06, + "loss": 0.48, + "step": 7406 + }, + { + "epoch": 0.6139245752175715, + "grad_norm": 0.24435769021511078, + "learning_rate": 1.930791545793618e-06, + "loss": 0.4843, + "step": 7407 + }, + { + "epoch": 0.6140074595938666, + "grad_norm": 0.23874661326408386, + "learning_rate": 1.9303771239121426e-06, + "loss": 0.5227, + "step": 7408 + }, + { + "epoch": 0.6140903439701616, + "grad_norm": 0.25187748670578003, + "learning_rate": 1.9299627020306674e-06, + "loss": 0.4858, + "step": 7409 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 0.25944557785987854, + "learning_rate": 1.9295482801491923e-06, + "loss": 0.5563, + "step": 7410 + }, + { + "epoch": 0.6142561127227517, + "grad_norm": 0.2773532271385193, + "learning_rate": 1.9291338582677167e-06, + "loss": 0.5024, + "step": 7411 + }, + { + "epoch": 0.6143389970990468, + "grad_norm": 0.26265084743499756, + "learning_rate": 1.928719436386241e-06, + "loss": 0.4948, + "step": 7412 + }, + { + "epoch": 0.6144218814753419, + "grad_norm": 0.25991320610046387, + "learning_rate": 1.928305014504766e-06, + "loss": 0.5382, + "step": 7413 + }, + { + "epoch": 0.6145047658516369, + "grad_norm": 0.24490919709205627, + "learning_rate": 1.927890592623291e-06, + "loss": 0.509, + "step": 7414 + }, + { + "epoch": 0.614587650227932, + "grad_norm": 0.2434261292219162, + "learning_rate": 1.9274761707418153e-06, + "loss": 0.5069, + "step": 7415 + }, + { + "epoch": 0.614670534604227, + "grad_norm": 0.2596171796321869, + "learning_rate": 1.9270617488603397e-06, + "loss": 0.4816, + "step": 7416 + }, + { + "epoch": 0.6147534189805222, + "grad_norm": 0.24932517111301422, + "learning_rate": 1.9266473269788646e-06, + "loss": 0.5069, + "step": 7417 + }, + { + "epoch": 0.6148363033568173, + "grad_norm": 0.22568076848983765, + "learning_rate": 1.9262329050973894e-06, + "loss": 0.4493, + "step": 7418 + }, + { + "epoch": 0.6149191877331123, + "grad_norm": 0.26414966583251953, + "learning_rate": 1.9258184832159143e-06, + "loss": 0.5403, + "step": 7419 + }, + { + "epoch": 0.6150020721094074, + "grad_norm": 0.25971338152885437, + "learning_rate": 1.9254040613344387e-06, + "loss": 0.5406, + "step": 7420 + }, + { + "epoch": 0.6150849564857025, + "grad_norm": 0.2507694959640503, + "learning_rate": 1.924989639452963e-06, + "loss": 0.492, + "step": 7421 + }, + { + "epoch": 0.6151678408619975, + "grad_norm": 0.27120715379714966, + "learning_rate": 1.924575217571488e-06, + "loss": 0.5321, + "step": 7422 + }, + { + "epoch": 0.6152507252382926, + "grad_norm": 0.25814154744148254, + "learning_rate": 1.924160795690013e-06, + "loss": 0.4794, + "step": 7423 + }, + { + "epoch": 0.6153336096145876, + "grad_norm": 0.23407229781150818, + "learning_rate": 1.9237463738085373e-06, + "loss": 0.4486, + "step": 7424 + }, + { + "epoch": 0.6154164939908827, + "grad_norm": 0.24202901124954224, + "learning_rate": 1.9233319519270617e-06, + "loss": 0.4727, + "step": 7425 + }, + { + "epoch": 0.6154993783671778, + "grad_norm": 0.2779672145843506, + "learning_rate": 1.9229175300455866e-06, + "loss": 0.4996, + "step": 7426 + }, + { + "epoch": 0.6155822627434728, + "grad_norm": 0.2669244706630707, + "learning_rate": 1.9225031081641114e-06, + "loss": 0.4605, + "step": 7427 + }, + { + "epoch": 0.6156651471197679, + "grad_norm": 0.23720207810401917, + "learning_rate": 1.922088686282636e-06, + "loss": 0.4932, + "step": 7428 + }, + { + "epoch": 0.6157480314960629, + "grad_norm": 0.24184705317020416, + "learning_rate": 1.9216742644011603e-06, + "loss": 0.4918, + "step": 7429 + }, + { + "epoch": 0.615830915872358, + "grad_norm": 0.24814115464687347, + "learning_rate": 1.921259842519685e-06, + "loss": 0.55, + "step": 7430 + }, + { + "epoch": 0.6159138002486532, + "grad_norm": 0.24334368109703064, + "learning_rate": 1.92084542063821e-06, + "loss": 0.4946, + "step": 7431 + }, + { + "epoch": 0.6159966846249482, + "grad_norm": 0.25471067428588867, + "learning_rate": 1.9204309987567344e-06, + "loss": 0.4932, + "step": 7432 + }, + { + "epoch": 0.6160795690012433, + "grad_norm": 0.2763904631137848, + "learning_rate": 1.920016576875259e-06, + "loss": 0.5494, + "step": 7433 + }, + { + "epoch": 0.6161624533775384, + "grad_norm": 0.23484402894973755, + "learning_rate": 1.9196021549937837e-06, + "loss": 0.4612, + "step": 7434 + }, + { + "epoch": 0.6162453377538334, + "grad_norm": 0.2398766428232193, + "learning_rate": 1.9191877331123086e-06, + "loss": 0.5129, + "step": 7435 + }, + { + "epoch": 0.6163282221301285, + "grad_norm": 0.2549682855606079, + "learning_rate": 1.918773311230833e-06, + "loss": 0.5091, + "step": 7436 + }, + { + "epoch": 0.6164111065064235, + "grad_norm": 0.25097382068634033, + "learning_rate": 1.918358889349358e-06, + "loss": 0.4888, + "step": 7437 + }, + { + "epoch": 0.6164939908827186, + "grad_norm": 0.2651837170124054, + "learning_rate": 1.9179444674678823e-06, + "loss": 0.5104, + "step": 7438 + }, + { + "epoch": 0.6165768752590137, + "grad_norm": 0.2464555948972702, + "learning_rate": 1.917530045586407e-06, + "loss": 0.4941, + "step": 7439 + }, + { + "epoch": 0.6166597596353087, + "grad_norm": 0.24881207942962646, + "learning_rate": 1.917115623704932e-06, + "loss": 0.5185, + "step": 7440 + }, + { + "epoch": 0.6167426440116038, + "grad_norm": 0.23743866384029388, + "learning_rate": 1.9167012018234564e-06, + "loss": 0.5377, + "step": 7441 + }, + { + "epoch": 0.6168255283878988, + "grad_norm": 0.2445751577615738, + "learning_rate": 1.916286779941981e-06, + "loss": 0.4858, + "step": 7442 + }, + { + "epoch": 0.616908412764194, + "grad_norm": 0.25504881143569946, + "learning_rate": 1.9158723580605057e-06, + "loss": 0.5216, + "step": 7443 + }, + { + "epoch": 0.6169912971404891, + "grad_norm": 0.24288786947727203, + "learning_rate": 1.9154579361790306e-06, + "loss": 0.4791, + "step": 7444 + }, + { + "epoch": 0.6170741815167841, + "grad_norm": 0.26749885082244873, + "learning_rate": 1.915043514297555e-06, + "loss": 0.5308, + "step": 7445 + }, + { + "epoch": 0.6171570658930792, + "grad_norm": 0.285183310508728, + "learning_rate": 1.91462909241608e-06, + "loss": 0.5115, + "step": 7446 + }, + { + "epoch": 0.6172399502693742, + "grad_norm": 0.267252117395401, + "learning_rate": 1.9142146705346043e-06, + "loss": 0.5033, + "step": 7447 + }, + { + "epoch": 0.6173228346456693, + "grad_norm": 0.24484872817993164, + "learning_rate": 1.913800248653129e-06, + "loss": 0.4813, + "step": 7448 + }, + { + "epoch": 0.6174057190219644, + "grad_norm": 0.2396039217710495, + "learning_rate": 1.9133858267716536e-06, + "loss": 0.4885, + "step": 7449 + }, + { + "epoch": 0.6174886033982594, + "grad_norm": 0.2435157150030136, + "learning_rate": 1.9129714048901784e-06, + "loss": 0.5273, + "step": 7450 + }, + { + "epoch": 0.6175714877745545, + "grad_norm": 0.2581246793270111, + "learning_rate": 1.912556983008703e-06, + "loss": 0.5264, + "step": 7451 + }, + { + "epoch": 0.6176543721508496, + "grad_norm": 0.24041827023029327, + "learning_rate": 1.9121425611272277e-06, + "loss": 0.4509, + "step": 7452 + }, + { + "epoch": 0.6177372565271446, + "grad_norm": 0.24676838517189026, + "learning_rate": 1.911728139245752e-06, + "loss": 0.5501, + "step": 7453 + }, + { + "epoch": 0.6178201409034397, + "grad_norm": 0.2682541012763977, + "learning_rate": 1.911313717364277e-06, + "loss": 0.4918, + "step": 7454 + }, + { + "epoch": 0.6179030252797347, + "grad_norm": 0.2507607638835907, + "learning_rate": 1.910899295482802e-06, + "loss": 0.5227, + "step": 7455 + }, + { + "epoch": 0.6179859096560298, + "grad_norm": 0.25044572353363037, + "learning_rate": 1.9104848736013263e-06, + "loss": 0.5222, + "step": 7456 + }, + { + "epoch": 0.618068794032325, + "grad_norm": 0.2461724728345871, + "learning_rate": 1.910070451719851e-06, + "loss": 0.5105, + "step": 7457 + }, + { + "epoch": 0.61815167840862, + "grad_norm": 0.22510294616222382, + "learning_rate": 1.9096560298383756e-06, + "loss": 0.4548, + "step": 7458 + }, + { + "epoch": 0.6182345627849151, + "grad_norm": 0.2560575306415558, + "learning_rate": 1.9092416079569004e-06, + "loss": 0.521, + "step": 7459 + }, + { + "epoch": 0.6183174471612101, + "grad_norm": 0.26849207282066345, + "learning_rate": 1.908827186075425e-06, + "loss": 0.5439, + "step": 7460 + }, + { + "epoch": 0.6184003315375052, + "grad_norm": 0.24310968816280365, + "learning_rate": 1.9084127641939497e-06, + "loss": 0.527, + "step": 7461 + }, + { + "epoch": 0.6184832159138003, + "grad_norm": 0.24602463841438293, + "learning_rate": 1.907998342312474e-06, + "loss": 0.4771, + "step": 7462 + }, + { + "epoch": 0.6185661002900953, + "grad_norm": 0.2462269514799118, + "learning_rate": 1.907583920430999e-06, + "loss": 0.4241, + "step": 7463 + }, + { + "epoch": 0.6186489846663904, + "grad_norm": 0.24001444876194, + "learning_rate": 1.9071694985495235e-06, + "loss": 0.5197, + "step": 7464 + }, + { + "epoch": 0.6187318690426854, + "grad_norm": 0.24025599658489227, + "learning_rate": 1.906755076668048e-06, + "loss": 0.4302, + "step": 7465 + }, + { + "epoch": 0.6188147534189805, + "grad_norm": 0.24170513451099396, + "learning_rate": 1.906340654786573e-06, + "loss": 0.5148, + "step": 7466 + }, + { + "epoch": 0.6188976377952756, + "grad_norm": 0.25742971897125244, + "learning_rate": 1.9059262329050976e-06, + "loss": 0.536, + "step": 7467 + }, + { + "epoch": 0.6189805221715706, + "grad_norm": 0.2287055104970932, + "learning_rate": 1.9055118110236222e-06, + "loss": 0.4667, + "step": 7468 + }, + { + "epoch": 0.6190634065478657, + "grad_norm": 0.2397616058588028, + "learning_rate": 1.9050973891421467e-06, + "loss": 0.4826, + "step": 7469 + }, + { + "epoch": 0.6191462909241608, + "grad_norm": 0.23624151945114136, + "learning_rate": 1.9046829672606715e-06, + "loss": 0.5173, + "step": 7470 + }, + { + "epoch": 0.6192291753004558, + "grad_norm": 0.2584115266799927, + "learning_rate": 1.9042685453791962e-06, + "loss": 0.5302, + "step": 7471 + }, + { + "epoch": 0.619312059676751, + "grad_norm": 0.2533610165119171, + "learning_rate": 1.9038541234977208e-06, + "loss": 0.4747, + "step": 7472 + }, + { + "epoch": 0.619394944053046, + "grad_norm": 0.24479064345359802, + "learning_rate": 1.9034397016162455e-06, + "loss": 0.4798, + "step": 7473 + }, + { + "epoch": 0.6194778284293411, + "grad_norm": 0.23404666781425476, + "learning_rate": 1.90302527973477e-06, + "loss": 0.5209, + "step": 7474 + }, + { + "epoch": 0.6195607128056362, + "grad_norm": 0.25055792927742004, + "learning_rate": 1.9026108578532947e-06, + "loss": 0.535, + "step": 7475 + }, + { + "epoch": 0.6196435971819312, + "grad_norm": 0.2483927309513092, + "learning_rate": 1.9021964359718196e-06, + "loss": 0.4946, + "step": 7476 + }, + { + "epoch": 0.6197264815582263, + "grad_norm": 0.25699397921562195, + "learning_rate": 1.9017820140903442e-06, + "loss": 0.5192, + "step": 7477 + }, + { + "epoch": 0.6198093659345213, + "grad_norm": 0.25947728753089905, + "learning_rate": 1.9013675922088687e-06, + "loss": 0.4995, + "step": 7478 + }, + { + "epoch": 0.6198922503108164, + "grad_norm": 0.25098639726638794, + "learning_rate": 1.9009531703273933e-06, + "loss": 0.5112, + "step": 7479 + }, + { + "epoch": 0.6199751346871115, + "grad_norm": 0.245499849319458, + "learning_rate": 1.9005387484459182e-06, + "loss": 0.4878, + "step": 7480 + }, + { + "epoch": 0.6200580190634065, + "grad_norm": 0.2428981363773346, + "learning_rate": 1.9001243265644428e-06, + "loss": 0.5135, + "step": 7481 + }, + { + "epoch": 0.6201409034397016, + "grad_norm": 0.23936967551708221, + "learning_rate": 1.8997099046829672e-06, + "loss": 0.4644, + "step": 7482 + }, + { + "epoch": 0.6202237878159966, + "grad_norm": 0.2349996417760849, + "learning_rate": 1.899295482801492e-06, + "loss": 0.5072, + "step": 7483 + }, + { + "epoch": 0.6203066721922917, + "grad_norm": 0.2546161115169525, + "learning_rate": 1.8988810609200167e-06, + "loss": 0.4888, + "step": 7484 + }, + { + "epoch": 0.6203895565685869, + "grad_norm": 0.23942138254642487, + "learning_rate": 1.8984666390385414e-06, + "loss": 0.4878, + "step": 7485 + }, + { + "epoch": 0.6204724409448819, + "grad_norm": 0.25063028931617737, + "learning_rate": 1.8980522171570662e-06, + "loss": 0.5326, + "step": 7486 + }, + { + "epoch": 0.620555325321177, + "grad_norm": 0.23901791870594025, + "learning_rate": 1.8976377952755907e-06, + "loss": 0.4706, + "step": 7487 + }, + { + "epoch": 0.6206382096974721, + "grad_norm": 0.259524941444397, + "learning_rate": 1.8972233733941153e-06, + "loss": 0.4956, + "step": 7488 + }, + { + "epoch": 0.6207210940737671, + "grad_norm": 0.2410217970609665, + "learning_rate": 1.89680895151264e-06, + "loss": 0.4887, + "step": 7489 + }, + { + "epoch": 0.6208039784500622, + "grad_norm": 0.2466568797826767, + "learning_rate": 1.8963945296311648e-06, + "loss": 0.4736, + "step": 7490 + }, + { + "epoch": 0.6208868628263572, + "grad_norm": 0.23691633343696594, + "learning_rate": 1.8959801077496893e-06, + "loss": 0.4948, + "step": 7491 + }, + { + "epoch": 0.6209697472026523, + "grad_norm": 0.25328752398490906, + "learning_rate": 1.895565685868214e-06, + "loss": 0.5139, + "step": 7492 + }, + { + "epoch": 0.6210526315789474, + "grad_norm": 0.2604285180568695, + "learning_rate": 1.8951512639867388e-06, + "loss": 0.5238, + "step": 7493 + }, + { + "epoch": 0.6211355159552424, + "grad_norm": 0.2590709924697876, + "learning_rate": 1.8947368421052634e-06, + "loss": 0.5306, + "step": 7494 + }, + { + "epoch": 0.6212184003315375, + "grad_norm": 0.26162683963775635, + "learning_rate": 1.894322420223788e-06, + "loss": 0.5104, + "step": 7495 + }, + { + "epoch": 0.6213012847078325, + "grad_norm": 0.24544689059257507, + "learning_rate": 1.8939079983423125e-06, + "loss": 0.522, + "step": 7496 + }, + { + "epoch": 0.6213841690841276, + "grad_norm": 0.2453528791666031, + "learning_rate": 1.8934935764608373e-06, + "loss": 0.5005, + "step": 7497 + }, + { + "epoch": 0.6214670534604227, + "grad_norm": 0.2592087388038635, + "learning_rate": 1.893079154579362e-06, + "loss": 0.5131, + "step": 7498 + }, + { + "epoch": 0.6215499378367177, + "grad_norm": 0.27567991614341736, + "learning_rate": 1.8926647326978866e-06, + "loss": 0.4758, + "step": 7499 + }, + { + "epoch": 0.6216328222130129, + "grad_norm": 0.2677079439163208, + "learning_rate": 1.892250310816411e-06, + "loss": 0.541, + "step": 7500 + }, + { + "epoch": 0.621715706589308, + "grad_norm": 0.24698272347450256, + "learning_rate": 1.891835888934936e-06, + "loss": 0.541, + "step": 7501 + }, + { + "epoch": 0.621798590965603, + "grad_norm": 0.24163466691970825, + "learning_rate": 1.8914214670534605e-06, + "loss": 0.5156, + "step": 7502 + }, + { + "epoch": 0.6218814753418981, + "grad_norm": 0.2564134895801544, + "learning_rate": 1.8910070451719854e-06, + "loss": 0.5245, + "step": 7503 + }, + { + "epoch": 0.6219643597181931, + "grad_norm": 0.2574622333049774, + "learning_rate": 1.89059262329051e-06, + "loss": 0.5376, + "step": 7504 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 0.24315212666988373, + "learning_rate": 1.8901782014090345e-06, + "loss": 0.4719, + "step": 7505 + }, + { + "epoch": 0.6221301284707833, + "grad_norm": 0.2779858410358429, + "learning_rate": 1.8897637795275591e-06, + "loss": 0.5485, + "step": 7506 + }, + { + "epoch": 0.6222130128470783, + "grad_norm": 0.23328250646591187, + "learning_rate": 1.889349357646084e-06, + "loss": 0.4728, + "step": 7507 + }, + { + "epoch": 0.6222958972233734, + "grad_norm": 0.23502035439014435, + "learning_rate": 1.8889349357646086e-06, + "loss": 0.5059, + "step": 7508 + }, + { + "epoch": 0.6223787815996684, + "grad_norm": 0.26746776700019836, + "learning_rate": 1.888520513883133e-06, + "loss": 0.4991, + "step": 7509 + }, + { + "epoch": 0.6224616659759635, + "grad_norm": 0.25528261065483093, + "learning_rate": 1.8881060920016577e-06, + "loss": 0.5468, + "step": 7510 + }, + { + "epoch": 0.6225445503522586, + "grad_norm": 0.24317477643489838, + "learning_rate": 1.8876916701201825e-06, + "loss": 0.501, + "step": 7511 + }, + { + "epoch": 0.6226274347285536, + "grad_norm": 0.2444218099117279, + "learning_rate": 1.8872772482387072e-06, + "loss": 0.5301, + "step": 7512 + }, + { + "epoch": 0.6227103191048488, + "grad_norm": 0.2712653577327728, + "learning_rate": 1.886862826357232e-06, + "loss": 0.5089, + "step": 7513 + }, + { + "epoch": 0.6227932034811438, + "grad_norm": 0.2723100185394287, + "learning_rate": 1.8864484044757565e-06, + "loss": 0.5087, + "step": 7514 + }, + { + "epoch": 0.6228760878574389, + "grad_norm": 0.24441488087177277, + "learning_rate": 1.8860339825942811e-06, + "loss": 0.4901, + "step": 7515 + }, + { + "epoch": 0.622958972233734, + "grad_norm": 0.24241192638874054, + "learning_rate": 1.8856195607128058e-06, + "loss": 0.478, + "step": 7516 + }, + { + "epoch": 0.623041856610029, + "grad_norm": 0.25302329659461975, + "learning_rate": 1.8852051388313306e-06, + "loss": 0.4839, + "step": 7517 + }, + { + "epoch": 0.6231247409863241, + "grad_norm": 0.2509937584400177, + "learning_rate": 1.884790716949855e-06, + "loss": 0.4924, + "step": 7518 + }, + { + "epoch": 0.6232076253626192, + "grad_norm": 0.25625768303871155, + "learning_rate": 1.8843762950683797e-06, + "loss": 0.5352, + "step": 7519 + }, + { + "epoch": 0.6232905097389142, + "grad_norm": 0.2502114772796631, + "learning_rate": 1.8839618731869043e-06, + "loss": 0.4944, + "step": 7520 + }, + { + "epoch": 0.6233733941152093, + "grad_norm": 0.23853619396686554, + "learning_rate": 1.8835474513054292e-06, + "loss": 0.4893, + "step": 7521 + }, + { + "epoch": 0.6234562784915043, + "grad_norm": 0.24865905940532684, + "learning_rate": 1.8831330294239538e-06, + "loss": 0.5353, + "step": 7522 + }, + { + "epoch": 0.6235391628677994, + "grad_norm": 0.23463228344917297, + "learning_rate": 1.8827186075424783e-06, + "loss": 0.4477, + "step": 7523 + }, + { + "epoch": 0.6236220472440945, + "grad_norm": 0.23698140680789948, + "learning_rate": 1.8823041856610031e-06, + "loss": 0.5083, + "step": 7524 + }, + { + "epoch": 0.6237049316203895, + "grad_norm": 0.2559659481048584, + "learning_rate": 1.8818897637795278e-06, + "loss": 0.5209, + "step": 7525 + }, + { + "epoch": 0.6237878159966846, + "grad_norm": 0.26678740978240967, + "learning_rate": 1.8814753418980524e-06, + "loss": 0.5786, + "step": 7526 + }, + { + "epoch": 0.6238707003729796, + "grad_norm": 0.26126280426979065, + "learning_rate": 1.8810609200165768e-06, + "loss": 0.4957, + "step": 7527 + }, + { + "epoch": 0.6239535847492748, + "grad_norm": 0.25310224294662476, + "learning_rate": 1.8806464981351017e-06, + "loss": 0.5038, + "step": 7528 + }, + { + "epoch": 0.6240364691255699, + "grad_norm": 0.24559944868087769, + "learning_rate": 1.8802320762536263e-06, + "loss": 0.4868, + "step": 7529 + }, + { + "epoch": 0.6241193535018649, + "grad_norm": 0.24255912005901337, + "learning_rate": 1.879817654372151e-06, + "loss": 0.4681, + "step": 7530 + }, + { + "epoch": 0.62420223787816, + "grad_norm": 0.23391379415988922, + "learning_rate": 1.8794032324906758e-06, + "loss": 0.4991, + "step": 7531 + }, + { + "epoch": 0.624285122254455, + "grad_norm": 0.24827013909816742, + "learning_rate": 1.8789888106092003e-06, + "loss": 0.5192, + "step": 7532 + }, + { + "epoch": 0.6243680066307501, + "grad_norm": 0.24476023018360138, + "learning_rate": 1.878574388727725e-06, + "loss": 0.4794, + "step": 7533 + }, + { + "epoch": 0.6244508910070452, + "grad_norm": 0.24288947880268097, + "learning_rate": 1.8781599668462498e-06, + "loss": 0.5202, + "step": 7534 + }, + { + "epoch": 0.6245337753833402, + "grad_norm": 0.2698659300804138, + "learning_rate": 1.8777455449647744e-06, + "loss": 0.5056, + "step": 7535 + }, + { + "epoch": 0.6246166597596353, + "grad_norm": 0.2407282441854477, + "learning_rate": 1.8773311230832988e-06, + "loss": 0.5009, + "step": 7536 + }, + { + "epoch": 0.6246995441359304, + "grad_norm": 0.24486996233463287, + "learning_rate": 1.8769167012018235e-06, + "loss": 0.4671, + "step": 7537 + }, + { + "epoch": 0.6247824285122254, + "grad_norm": 0.2480809986591339, + "learning_rate": 1.8765022793203483e-06, + "loss": 0.4926, + "step": 7538 + }, + { + "epoch": 0.6248653128885205, + "grad_norm": 0.245940163731575, + "learning_rate": 1.876087857438873e-06, + "loss": 0.5203, + "step": 7539 + }, + { + "epoch": 0.6249481972648155, + "grad_norm": 0.27914005517959595, + "learning_rate": 1.8756734355573974e-06, + "loss": 0.4973, + "step": 7540 + }, + { + "epoch": 0.6250310816411107, + "grad_norm": 0.2396421879529953, + "learning_rate": 1.875259013675922e-06, + "loss": 0.5329, + "step": 7541 + }, + { + "epoch": 0.6251139660174058, + "grad_norm": 0.22998638451099396, + "learning_rate": 1.874844591794447e-06, + "loss": 0.4809, + "step": 7542 + }, + { + "epoch": 0.6251968503937008, + "grad_norm": 0.25688380002975464, + "learning_rate": 1.8744301699129716e-06, + "loss": 0.497, + "step": 7543 + }, + { + "epoch": 0.6252797347699959, + "grad_norm": 0.2508845925331116, + "learning_rate": 1.8740157480314964e-06, + "loss": 0.5325, + "step": 7544 + }, + { + "epoch": 0.6253626191462909, + "grad_norm": 0.27134639024734497, + "learning_rate": 1.8736013261500209e-06, + "loss": 0.5305, + "step": 7545 + }, + { + "epoch": 0.625445503522586, + "grad_norm": 0.2723635733127594, + "learning_rate": 1.8731869042685455e-06, + "loss": 0.523, + "step": 7546 + }, + { + "epoch": 0.6255283878988811, + "grad_norm": 0.2502851188182831, + "learning_rate": 1.8727724823870701e-06, + "loss": 0.5407, + "step": 7547 + }, + { + "epoch": 0.6256112722751761, + "grad_norm": 0.24608831107616425, + "learning_rate": 1.872358060505595e-06, + "loss": 0.5085, + "step": 7548 + }, + { + "epoch": 0.6256941566514712, + "grad_norm": 0.31680604815483093, + "learning_rate": 1.8719436386241194e-06, + "loss": 0.534, + "step": 7549 + }, + { + "epoch": 0.6257770410277662, + "grad_norm": 0.24810777604579926, + "learning_rate": 1.871529216742644e-06, + "loss": 0.5009, + "step": 7550 + }, + { + "epoch": 0.6258599254040613, + "grad_norm": 0.2560594379901886, + "learning_rate": 1.8711147948611687e-06, + "loss": 0.5075, + "step": 7551 + }, + { + "epoch": 0.6259428097803564, + "grad_norm": 0.24362914264202118, + "learning_rate": 1.8707003729796936e-06, + "loss": 0.4975, + "step": 7552 + }, + { + "epoch": 0.6260256941566514, + "grad_norm": 0.23871663212776184, + "learning_rate": 1.8702859510982182e-06, + "loss": 0.4676, + "step": 7553 + }, + { + "epoch": 0.6261085785329465, + "grad_norm": 0.2633377015590668, + "learning_rate": 1.8698715292167426e-06, + "loss": 0.529, + "step": 7554 + }, + { + "epoch": 0.6261914629092417, + "grad_norm": 0.24860064685344696, + "learning_rate": 1.8694571073352675e-06, + "loss": 0.512, + "step": 7555 + }, + { + "epoch": 0.6262743472855367, + "grad_norm": 0.25935986638069153, + "learning_rate": 1.8690426854537921e-06, + "loss": 0.4979, + "step": 7556 + }, + { + "epoch": 0.6263572316618318, + "grad_norm": 0.2608540654182434, + "learning_rate": 1.8686282635723168e-06, + "loss": 0.5046, + "step": 7557 + }, + { + "epoch": 0.6264401160381268, + "grad_norm": 0.24915437400341034, + "learning_rate": 1.8682138416908412e-06, + "loss": 0.4817, + "step": 7558 + }, + { + "epoch": 0.6265230004144219, + "grad_norm": 0.24860571324825287, + "learning_rate": 1.867799419809366e-06, + "loss": 0.5, + "step": 7559 + }, + { + "epoch": 0.626605884790717, + "grad_norm": 0.23555965721607208, + "learning_rate": 1.8673849979278907e-06, + "loss": 0.5003, + "step": 7560 + }, + { + "epoch": 0.626688769167012, + "grad_norm": 0.2559857964515686, + "learning_rate": 1.8669705760464154e-06, + "loss": 0.5326, + "step": 7561 + }, + { + "epoch": 0.6267716535433071, + "grad_norm": 0.29279041290283203, + "learning_rate": 1.8665561541649402e-06, + "loss": 0.5215, + "step": 7562 + }, + { + "epoch": 0.6268545379196021, + "grad_norm": 0.22955438494682312, + "learning_rate": 1.8661417322834646e-06, + "loss": 0.491, + "step": 7563 + }, + { + "epoch": 0.6269374222958972, + "grad_norm": 0.2603181302547455, + "learning_rate": 1.8657273104019893e-06, + "loss": 0.5183, + "step": 7564 + }, + { + "epoch": 0.6270203066721923, + "grad_norm": 0.23360446095466614, + "learning_rate": 1.8653128885205141e-06, + "loss": 0.5007, + "step": 7565 + }, + { + "epoch": 0.6271031910484873, + "grad_norm": 0.2583872675895691, + "learning_rate": 1.8648984666390388e-06, + "loss": 0.5051, + "step": 7566 + }, + { + "epoch": 0.6271860754247824, + "grad_norm": 0.25934290885925293, + "learning_rate": 1.8644840447575632e-06, + "loss": 0.5535, + "step": 7567 + }, + { + "epoch": 0.6272689598010774, + "grad_norm": 0.23916813731193542, + "learning_rate": 1.8640696228760879e-06, + "loss": 0.5004, + "step": 7568 + }, + { + "epoch": 0.6273518441773726, + "grad_norm": 0.2352968007326126, + "learning_rate": 1.8636552009946127e-06, + "loss": 0.4762, + "step": 7569 + }, + { + "epoch": 0.6274347285536677, + "grad_norm": 0.23549984395503998, + "learning_rate": 1.8632407791131374e-06, + "loss": 0.4867, + "step": 7570 + }, + { + "epoch": 0.6275176129299627, + "grad_norm": 0.28633496165275574, + "learning_rate": 1.862826357231662e-06, + "loss": 0.4996, + "step": 7571 + }, + { + "epoch": 0.6276004973062578, + "grad_norm": 0.25099077820777893, + "learning_rate": 1.8624119353501867e-06, + "loss": 0.5168, + "step": 7572 + }, + { + "epoch": 0.6276833816825529, + "grad_norm": 0.23867832124233246, + "learning_rate": 1.8619975134687113e-06, + "loss": 0.4726, + "step": 7573 + }, + { + "epoch": 0.6277662660588479, + "grad_norm": 0.251509428024292, + "learning_rate": 1.861583091587236e-06, + "loss": 0.5536, + "step": 7574 + }, + { + "epoch": 0.627849150435143, + "grad_norm": 0.24150985479354858, + "learning_rate": 1.8611686697057608e-06, + "loss": 0.5274, + "step": 7575 + }, + { + "epoch": 0.627932034811438, + "grad_norm": 0.2676033079624176, + "learning_rate": 1.8607542478242852e-06, + "loss": 0.5269, + "step": 7576 + }, + { + "epoch": 0.6280149191877331, + "grad_norm": 0.3129712641239166, + "learning_rate": 1.8603398259428099e-06, + "loss": 0.5082, + "step": 7577 + }, + { + "epoch": 0.6280978035640282, + "grad_norm": 0.28359758853912354, + "learning_rate": 1.8599254040613345e-06, + "loss": 0.5027, + "step": 7578 + }, + { + "epoch": 0.6281806879403232, + "grad_norm": 0.23108188807964325, + "learning_rate": 1.8595109821798594e-06, + "loss": 0.4819, + "step": 7579 + }, + { + "epoch": 0.6282635723166183, + "grad_norm": 0.25028663873672485, + "learning_rate": 1.859096560298384e-06, + "loss": 0.4719, + "step": 7580 + }, + { + "epoch": 0.6283464566929133, + "grad_norm": 0.2575945556163788, + "learning_rate": 1.8586821384169084e-06, + "loss": 0.506, + "step": 7581 + }, + { + "epoch": 0.6284293410692084, + "grad_norm": 0.24821126461029053, + "learning_rate": 1.8582677165354333e-06, + "loss": 0.5453, + "step": 7582 + }, + { + "epoch": 0.6285122254455036, + "grad_norm": 0.25781601667404175, + "learning_rate": 1.857853294653958e-06, + "loss": 0.5345, + "step": 7583 + }, + { + "epoch": 0.6285951098217986, + "grad_norm": 0.2790740728378296, + "learning_rate": 1.8574388727724826e-06, + "loss": 0.5229, + "step": 7584 + }, + { + "epoch": 0.6286779941980937, + "grad_norm": 0.27003249526023865, + "learning_rate": 1.857024450891007e-06, + "loss": 0.5436, + "step": 7585 + }, + { + "epoch": 0.6287608785743888, + "grad_norm": 0.253176748752594, + "learning_rate": 1.8566100290095319e-06, + "loss": 0.514, + "step": 7586 + }, + { + "epoch": 0.6288437629506838, + "grad_norm": 0.23671624064445496, + "learning_rate": 1.8561956071280565e-06, + "loss": 0.4686, + "step": 7587 + }, + { + "epoch": 0.6289266473269789, + "grad_norm": 0.25379809737205505, + "learning_rate": 1.8557811852465812e-06, + "loss": 0.5004, + "step": 7588 + }, + { + "epoch": 0.6290095317032739, + "grad_norm": 0.24802155792713165, + "learning_rate": 1.855366763365106e-06, + "loss": 0.4999, + "step": 7589 + }, + { + "epoch": 0.629092416079569, + "grad_norm": 0.2266547530889511, + "learning_rate": 1.8549523414836304e-06, + "loss": 0.4696, + "step": 7590 + }, + { + "epoch": 0.6291753004558641, + "grad_norm": 0.2541022002696991, + "learning_rate": 1.854537919602155e-06, + "loss": 0.5206, + "step": 7591 + }, + { + "epoch": 0.6292581848321591, + "grad_norm": 0.2271593064069748, + "learning_rate": 1.85412349772068e-06, + "loss": 0.4704, + "step": 7592 + }, + { + "epoch": 0.6293410692084542, + "grad_norm": 0.26065492630004883, + "learning_rate": 1.8537090758392046e-06, + "loss": 0.5093, + "step": 7593 + }, + { + "epoch": 0.6294239535847492, + "grad_norm": 0.25214189291000366, + "learning_rate": 1.853294653957729e-06, + "loss": 0.5032, + "step": 7594 + }, + { + "epoch": 0.6295068379610443, + "grad_norm": 0.2638075351715088, + "learning_rate": 1.8528802320762537e-06, + "loss": 0.5113, + "step": 7595 + }, + { + "epoch": 0.6295897223373395, + "grad_norm": 0.22472910583019257, + "learning_rate": 1.8524658101947785e-06, + "loss": 0.4623, + "step": 7596 + }, + { + "epoch": 0.6296726067136345, + "grad_norm": 0.22820472717285156, + "learning_rate": 1.8520513883133032e-06, + "loss": 0.4755, + "step": 7597 + }, + { + "epoch": 0.6297554910899296, + "grad_norm": 0.25946852564811707, + "learning_rate": 1.8516369664318278e-06, + "loss": 0.4991, + "step": 7598 + }, + { + "epoch": 0.6298383754662246, + "grad_norm": 0.24949218332767487, + "learning_rate": 1.8512225445503522e-06, + "loss": 0.5221, + "step": 7599 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 0.24731364846229553, + "learning_rate": 1.850808122668877e-06, + "loss": 0.5274, + "step": 7600 + }, + { + "epoch": 0.6300041442188148, + "grad_norm": 0.2464110553264618, + "learning_rate": 1.8503937007874017e-06, + "loss": 0.4814, + "step": 7601 + }, + { + "epoch": 0.6300870285951098, + "grad_norm": 0.24505510926246643, + "learning_rate": 1.8499792789059266e-06, + "loss": 0.5442, + "step": 7602 + }, + { + "epoch": 0.6301699129714049, + "grad_norm": 0.2457435578107834, + "learning_rate": 1.849564857024451e-06, + "loss": 0.5156, + "step": 7603 + }, + { + "epoch": 0.6302527973477, + "grad_norm": 0.2426517754793167, + "learning_rate": 1.8491504351429757e-06, + "loss": 0.4465, + "step": 7604 + }, + { + "epoch": 0.630335681723995, + "grad_norm": 0.25834131240844727, + "learning_rate": 1.8487360132615003e-06, + "loss": 0.5227, + "step": 7605 + }, + { + "epoch": 0.6304185661002901, + "grad_norm": 0.25398170948028564, + "learning_rate": 1.8483215913800252e-06, + "loss": 0.4894, + "step": 7606 + }, + { + "epoch": 0.6305014504765851, + "grad_norm": 0.24131745100021362, + "learning_rate": 1.8479071694985498e-06, + "loss": 0.4976, + "step": 7607 + }, + { + "epoch": 0.6305843348528802, + "grad_norm": 0.24505896866321564, + "learning_rate": 1.8474927476170742e-06, + "loss": 0.5251, + "step": 7608 + }, + { + "epoch": 0.6306672192291753, + "grad_norm": 0.24193175137043, + "learning_rate": 1.8470783257355989e-06, + "loss": 0.4782, + "step": 7609 + }, + { + "epoch": 0.6307501036054703, + "grad_norm": 0.4046739339828491, + "learning_rate": 1.8466639038541237e-06, + "loss": 0.5228, + "step": 7610 + }, + { + "epoch": 0.6308329879817655, + "grad_norm": 0.24619907140731812, + "learning_rate": 1.8462494819726484e-06, + "loss": 0.5063, + "step": 7611 + }, + { + "epoch": 0.6309158723580605, + "grad_norm": 0.2558591961860657, + "learning_rate": 1.8458350600911728e-06, + "loss": 0.5116, + "step": 7612 + }, + { + "epoch": 0.6309987567343556, + "grad_norm": 0.2400951385498047, + "learning_rate": 1.8454206382096977e-06, + "loss": 0.4827, + "step": 7613 + }, + { + "epoch": 0.6310816411106507, + "grad_norm": 0.280203640460968, + "learning_rate": 1.8450062163282223e-06, + "loss": 0.5097, + "step": 7614 + }, + { + "epoch": 0.6311645254869457, + "grad_norm": 0.23356036841869354, + "learning_rate": 1.844591794446747e-06, + "loss": 0.493, + "step": 7615 + }, + { + "epoch": 0.6312474098632408, + "grad_norm": 0.2751774489879608, + "learning_rate": 1.8441773725652714e-06, + "loss": 0.5631, + "step": 7616 + }, + { + "epoch": 0.6313302942395358, + "grad_norm": 0.2486829310655594, + "learning_rate": 1.8437629506837962e-06, + "loss": 0.499, + "step": 7617 + }, + { + "epoch": 0.6314131786158309, + "grad_norm": 0.2391642928123474, + "learning_rate": 1.8433485288023209e-06, + "loss": 0.4798, + "step": 7618 + }, + { + "epoch": 0.631496062992126, + "grad_norm": 0.27398592233657837, + "learning_rate": 1.8429341069208455e-06, + "loss": 0.5583, + "step": 7619 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 0.26633211970329285, + "learning_rate": 1.8425196850393704e-06, + "loss": 0.5176, + "step": 7620 + }, + { + "epoch": 0.6316618317447161, + "grad_norm": 0.25045305490493774, + "learning_rate": 1.8421052631578948e-06, + "loss": 0.4339, + "step": 7621 + }, + { + "epoch": 0.6317447161210112, + "grad_norm": 0.23077239096164703, + "learning_rate": 1.8416908412764195e-06, + "loss": 0.4171, + "step": 7622 + }, + { + "epoch": 0.6318276004973062, + "grad_norm": 0.26001641154289246, + "learning_rate": 1.8412764193949443e-06, + "loss": 0.4833, + "step": 7623 + }, + { + "epoch": 0.6319104848736014, + "grad_norm": 0.23063069581985474, + "learning_rate": 1.840861997513469e-06, + "loss": 0.4744, + "step": 7624 + }, + { + "epoch": 0.6319933692498964, + "grad_norm": 0.23150113224983215, + "learning_rate": 1.8404475756319934e-06, + "loss": 0.4684, + "step": 7625 + }, + { + "epoch": 0.6320762536261915, + "grad_norm": 0.2521877884864807, + "learning_rate": 1.840033153750518e-06, + "loss": 0.5012, + "step": 7626 + }, + { + "epoch": 0.6321591380024866, + "grad_norm": 0.24209314584732056, + "learning_rate": 1.8396187318690429e-06, + "loss": 0.4782, + "step": 7627 + }, + { + "epoch": 0.6322420223787816, + "grad_norm": 0.27349603176116943, + "learning_rate": 1.8392043099875675e-06, + "loss": 0.4851, + "step": 7628 + }, + { + "epoch": 0.6323249067550767, + "grad_norm": 0.23484483361244202, + "learning_rate": 1.8387898881060922e-06, + "loss": 0.4668, + "step": 7629 + }, + { + "epoch": 0.6324077911313717, + "grad_norm": 0.2267313301563263, + "learning_rate": 1.8383754662246166e-06, + "loss": 0.444, + "step": 7630 + }, + { + "epoch": 0.6324906755076668, + "grad_norm": 0.2619278132915497, + "learning_rate": 1.8379610443431415e-06, + "loss": 0.5305, + "step": 7631 + }, + { + "epoch": 0.6325735598839619, + "grad_norm": 0.26336535811424255, + "learning_rate": 1.8375466224616661e-06, + "loss": 0.5489, + "step": 7632 + }, + { + "epoch": 0.6326564442602569, + "grad_norm": 0.25769907236099243, + "learning_rate": 1.837132200580191e-06, + "loss": 0.5109, + "step": 7633 + }, + { + "epoch": 0.632739328636552, + "grad_norm": 0.24540941417217255, + "learning_rate": 1.8367177786987154e-06, + "loss": 0.501, + "step": 7634 + }, + { + "epoch": 0.632822213012847, + "grad_norm": 0.2327808290719986, + "learning_rate": 1.83630335681724e-06, + "loss": 0.5047, + "step": 7635 + }, + { + "epoch": 0.6329050973891421, + "grad_norm": 0.2594437301158905, + "learning_rate": 1.8358889349357647e-06, + "loss": 0.5347, + "step": 7636 + }, + { + "epoch": 0.6329879817654372, + "grad_norm": 0.2593572437763214, + "learning_rate": 1.8354745130542895e-06, + "loss": 0.5189, + "step": 7637 + }, + { + "epoch": 0.6330708661417322, + "grad_norm": 0.2463100552558899, + "learning_rate": 1.8350600911728142e-06, + "loss": 0.4984, + "step": 7638 + }, + { + "epoch": 0.6331537505180274, + "grad_norm": 0.2791823446750641, + "learning_rate": 1.8346456692913386e-06, + "loss": 0.5549, + "step": 7639 + }, + { + "epoch": 0.6332366348943225, + "grad_norm": 0.27160415053367615, + "learning_rate": 1.8342312474098633e-06, + "loss": 0.5133, + "step": 7640 + }, + { + "epoch": 0.6333195192706175, + "grad_norm": 0.27682963013648987, + "learning_rate": 1.8338168255283881e-06, + "loss": 0.5316, + "step": 7641 + }, + { + "epoch": 0.6334024036469126, + "grad_norm": 0.24093572795391083, + "learning_rate": 1.8334024036469128e-06, + "loss": 0.4518, + "step": 7642 + }, + { + "epoch": 0.6334852880232076, + "grad_norm": 0.2519181966781616, + "learning_rate": 1.8329879817654372e-06, + "loss": 0.5183, + "step": 7643 + }, + { + "epoch": 0.6335681723995027, + "grad_norm": 0.26346278190612793, + "learning_rate": 1.832573559883962e-06, + "loss": 0.5356, + "step": 7644 + }, + { + "epoch": 0.6336510567757978, + "grad_norm": 0.2510942220687866, + "learning_rate": 1.8321591380024867e-06, + "loss": 0.4991, + "step": 7645 + }, + { + "epoch": 0.6337339411520928, + "grad_norm": 0.2538674771785736, + "learning_rate": 1.8317447161210113e-06, + "loss": 0.5264, + "step": 7646 + }, + { + "epoch": 0.6338168255283879, + "grad_norm": 0.24933061003684998, + "learning_rate": 1.8313302942395362e-06, + "loss": 0.4934, + "step": 7647 + }, + { + "epoch": 0.6338997099046829, + "grad_norm": 0.24819441139698029, + "learning_rate": 1.8309158723580606e-06, + "loss": 0.4804, + "step": 7648 + }, + { + "epoch": 0.633982594280978, + "grad_norm": 0.24275648593902588, + "learning_rate": 1.8305014504765853e-06, + "loss": 0.4943, + "step": 7649 + }, + { + "epoch": 0.6340654786572731, + "grad_norm": 0.23486724495887756, + "learning_rate": 1.83008702859511e-06, + "loss": 0.4955, + "step": 7650 + }, + { + "epoch": 0.6341483630335681, + "grad_norm": 0.2353447824716568, + "learning_rate": 1.8296726067136348e-06, + "loss": 0.4951, + "step": 7651 + }, + { + "epoch": 0.6342312474098633, + "grad_norm": 0.23336441814899445, + "learning_rate": 1.8292581848321592e-06, + "loss": 0.4732, + "step": 7652 + }, + { + "epoch": 0.6343141317861584, + "grad_norm": 0.23743723332881927, + "learning_rate": 1.8288437629506838e-06, + "loss": 0.4579, + "step": 7653 + }, + { + "epoch": 0.6343970161624534, + "grad_norm": 0.2450876384973526, + "learning_rate": 1.8284293410692087e-06, + "loss": 0.4985, + "step": 7654 + }, + { + "epoch": 0.6344799005387485, + "grad_norm": 0.2691735327243805, + "learning_rate": 1.8280149191877333e-06, + "loss": 0.5231, + "step": 7655 + }, + { + "epoch": 0.6345627849150435, + "grad_norm": 0.2692229747772217, + "learning_rate": 1.827600497306258e-06, + "loss": 0.497, + "step": 7656 + }, + { + "epoch": 0.6346456692913386, + "grad_norm": 0.23677654564380646, + "learning_rate": 1.8271860754247824e-06, + "loss": 0.465, + "step": 7657 + }, + { + "epoch": 0.6347285536676337, + "grad_norm": 0.23710711300373077, + "learning_rate": 1.8267716535433073e-06, + "loss": 0.4901, + "step": 7658 + }, + { + "epoch": 0.6348114380439287, + "grad_norm": 0.2610302269458771, + "learning_rate": 1.826357231661832e-06, + "loss": 0.5093, + "step": 7659 + }, + { + "epoch": 0.6348943224202238, + "grad_norm": 0.24664422869682312, + "learning_rate": 1.8259428097803566e-06, + "loss": 0.5214, + "step": 7660 + }, + { + "epoch": 0.6349772067965188, + "grad_norm": 0.2481367588043213, + "learning_rate": 1.8255283878988812e-06, + "loss": 0.5135, + "step": 7661 + }, + { + "epoch": 0.6350600911728139, + "grad_norm": 0.25573912262916565, + "learning_rate": 1.8251139660174058e-06, + "loss": 0.4907, + "step": 7662 + }, + { + "epoch": 0.635142975549109, + "grad_norm": 0.23787499964237213, + "learning_rate": 1.8246995441359305e-06, + "loss": 0.4904, + "step": 7663 + }, + { + "epoch": 0.635225859925404, + "grad_norm": 0.26319923996925354, + "learning_rate": 1.8242851222544553e-06, + "loss": 0.5156, + "step": 7664 + }, + { + "epoch": 0.6353087443016991, + "grad_norm": 0.34920135140419006, + "learning_rate": 1.82387070037298e-06, + "loss": 0.5081, + "step": 7665 + }, + { + "epoch": 0.6353916286779941, + "grad_norm": 0.2723716199398041, + "learning_rate": 1.8234562784915044e-06, + "loss": 0.4463, + "step": 7666 + }, + { + "epoch": 0.6354745130542893, + "grad_norm": 0.24365027248859406, + "learning_rate": 1.823041856610029e-06, + "loss": 0.4945, + "step": 7667 + }, + { + "epoch": 0.6355573974305844, + "grad_norm": 0.272487610578537, + "learning_rate": 1.822627434728554e-06, + "loss": 0.5356, + "step": 7668 + }, + { + "epoch": 0.6356402818068794, + "grad_norm": 0.27329081296920776, + "learning_rate": 1.8222130128470786e-06, + "loss": 0.531, + "step": 7669 + }, + { + "epoch": 0.6357231661831745, + "grad_norm": 0.24880139529705048, + "learning_rate": 1.821798590965603e-06, + "loss": 0.4987, + "step": 7670 + }, + { + "epoch": 0.6358060505594696, + "grad_norm": 0.2537411153316498, + "learning_rate": 1.8213841690841278e-06, + "loss": 0.4678, + "step": 7671 + }, + { + "epoch": 0.6358889349357646, + "grad_norm": 0.24836230278015137, + "learning_rate": 1.8209697472026525e-06, + "loss": 0.4688, + "step": 7672 + }, + { + "epoch": 0.6359718193120597, + "grad_norm": 0.24117743968963623, + "learning_rate": 1.8205553253211771e-06, + "loss": 0.5161, + "step": 7673 + }, + { + "epoch": 0.6360547036883547, + "grad_norm": 0.2381865680217743, + "learning_rate": 1.820140903439702e-06, + "loss": 0.4905, + "step": 7674 + }, + { + "epoch": 0.6361375880646498, + "grad_norm": 0.25106844305992126, + "learning_rate": 1.8197264815582264e-06, + "loss": 0.4893, + "step": 7675 + }, + { + "epoch": 0.6362204724409449, + "grad_norm": 0.2660585641860962, + "learning_rate": 1.819312059676751e-06, + "loss": 0.5268, + "step": 7676 + }, + { + "epoch": 0.6363033568172399, + "grad_norm": 0.2575131356716156, + "learning_rate": 1.8188976377952757e-06, + "loss": 0.4789, + "step": 7677 + }, + { + "epoch": 0.636386241193535, + "grad_norm": 0.22086475789546967, + "learning_rate": 1.8184832159138006e-06, + "loss": 0.4404, + "step": 7678 + }, + { + "epoch": 0.63646912556983, + "grad_norm": 0.23993496596813202, + "learning_rate": 1.818068794032325e-06, + "loss": 0.4974, + "step": 7679 + }, + { + "epoch": 0.6365520099461252, + "grad_norm": 0.22702574729919434, + "learning_rate": 1.8176543721508496e-06, + "loss": 0.457, + "step": 7680 + }, + { + "epoch": 0.6366348943224203, + "grad_norm": 0.2454414814710617, + "learning_rate": 1.8172399502693743e-06, + "loss": 0.4984, + "step": 7681 + }, + { + "epoch": 0.6367177786987153, + "grad_norm": 0.2642146646976471, + "learning_rate": 1.8168255283878991e-06, + "loss": 0.5227, + "step": 7682 + }, + { + "epoch": 0.6368006630750104, + "grad_norm": 0.24749986827373505, + "learning_rate": 1.8164111065064236e-06, + "loss": 0.4877, + "step": 7683 + }, + { + "epoch": 0.6368835474513054, + "grad_norm": 0.23293127119541168, + "learning_rate": 1.8159966846249482e-06, + "loss": 0.4448, + "step": 7684 + }, + { + "epoch": 0.6369664318276005, + "grad_norm": 0.2548583745956421, + "learning_rate": 1.815582262743473e-06, + "loss": 0.5093, + "step": 7685 + }, + { + "epoch": 0.6370493162038956, + "grad_norm": 0.25644415616989136, + "learning_rate": 1.8151678408619977e-06, + "loss": 0.4968, + "step": 7686 + }, + { + "epoch": 0.6371322005801906, + "grad_norm": 0.24852629005908966, + "learning_rate": 1.8147534189805224e-06, + "loss": 0.5261, + "step": 7687 + }, + { + "epoch": 0.6372150849564857, + "grad_norm": 0.2407263219356537, + "learning_rate": 1.8143389970990468e-06, + "loss": 0.5035, + "step": 7688 + }, + { + "epoch": 0.6372979693327808, + "grad_norm": 0.23153944313526154, + "learning_rate": 1.8139245752175716e-06, + "loss": 0.4857, + "step": 7689 + }, + { + "epoch": 0.6373808537090758, + "grad_norm": 0.2621701955795288, + "learning_rate": 1.8135101533360963e-06, + "loss": 0.5585, + "step": 7690 + }, + { + "epoch": 0.6374637380853709, + "grad_norm": 0.264218807220459, + "learning_rate": 1.813095731454621e-06, + "loss": 0.5302, + "step": 7691 + }, + { + "epoch": 0.6375466224616659, + "grad_norm": 0.23753978312015533, + "learning_rate": 1.8126813095731456e-06, + "loss": 0.5237, + "step": 7692 + }, + { + "epoch": 0.637629506837961, + "grad_norm": 0.2560185194015503, + "learning_rate": 1.8122668876916702e-06, + "loss": 0.49, + "step": 7693 + }, + { + "epoch": 0.6377123912142562, + "grad_norm": 0.24593426287174225, + "learning_rate": 1.8118524658101949e-06, + "loss": 0.4819, + "step": 7694 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 0.23267513513565063, + "learning_rate": 1.8114380439287197e-06, + "loss": 0.5011, + "step": 7695 + }, + { + "epoch": 0.6378781599668463, + "grad_norm": 0.2314089834690094, + "learning_rate": 1.8110236220472444e-06, + "loss": 0.4715, + "step": 7696 + }, + { + "epoch": 0.6379610443431413, + "grad_norm": 0.25834324955940247, + "learning_rate": 1.8106092001657688e-06, + "loss": 0.509, + "step": 7697 + }, + { + "epoch": 0.6380439287194364, + "grad_norm": 0.256652295589447, + "learning_rate": 1.8101947782842934e-06, + "loss": 0.5232, + "step": 7698 + }, + { + "epoch": 0.6381268130957315, + "grad_norm": 0.24005848169326782, + "learning_rate": 1.8097803564028183e-06, + "loss": 0.4944, + "step": 7699 + }, + { + "epoch": 0.6382096974720265, + "grad_norm": 0.24952974915504456, + "learning_rate": 1.809365934521343e-06, + "loss": 0.4977, + "step": 7700 + }, + { + "epoch": 0.6382925818483216, + "grad_norm": 0.2528371810913086, + "learning_rate": 1.8089515126398674e-06, + "loss": 0.5595, + "step": 7701 + }, + { + "epoch": 0.6383754662246166, + "grad_norm": 0.2548118233680725, + "learning_rate": 1.8085370907583922e-06, + "loss": 0.5035, + "step": 7702 + }, + { + "epoch": 0.6384583506009117, + "grad_norm": 0.25400683283805847, + "learning_rate": 1.8081226688769169e-06, + "loss": 0.5242, + "step": 7703 + }, + { + "epoch": 0.6385412349772068, + "grad_norm": 0.2495017945766449, + "learning_rate": 1.8077082469954415e-06, + "loss": 0.5066, + "step": 7704 + }, + { + "epoch": 0.6386241193535018, + "grad_norm": 0.2640254497528076, + "learning_rate": 1.8072938251139664e-06, + "loss": 0.4766, + "step": 7705 + }, + { + "epoch": 0.6387070037297969, + "grad_norm": 0.2366587072610855, + "learning_rate": 1.8068794032324908e-06, + "loss": 0.5124, + "step": 7706 + }, + { + "epoch": 0.638789888106092, + "grad_norm": 0.23772473633289337, + "learning_rate": 1.8064649813510154e-06, + "loss": 0.4717, + "step": 7707 + }, + { + "epoch": 0.638872772482387, + "grad_norm": 0.21747583150863647, + "learning_rate": 1.80605055946954e-06, + "loss": 0.4796, + "step": 7708 + }, + { + "epoch": 0.6389556568586822, + "grad_norm": 0.24602487683296204, + "learning_rate": 1.805636137588065e-06, + "loss": 0.4807, + "step": 7709 + }, + { + "epoch": 0.6390385412349772, + "grad_norm": 0.26050180196762085, + "learning_rate": 1.8052217157065894e-06, + "loss": 0.5746, + "step": 7710 + }, + { + "epoch": 0.6391214256112723, + "grad_norm": 0.22876940667629242, + "learning_rate": 1.804807293825114e-06, + "loss": 0.4719, + "step": 7711 + }, + { + "epoch": 0.6392043099875674, + "grad_norm": 0.31667661666870117, + "learning_rate": 1.8043928719436389e-06, + "loss": 0.5095, + "step": 7712 + }, + { + "epoch": 0.6392871943638624, + "grad_norm": 0.2455551028251648, + "learning_rate": 1.8039784500621635e-06, + "loss": 0.5352, + "step": 7713 + }, + { + "epoch": 0.6393700787401575, + "grad_norm": 0.23896028101444244, + "learning_rate": 1.8035640281806882e-06, + "loss": 0.4814, + "step": 7714 + }, + { + "epoch": 0.6394529631164525, + "grad_norm": 0.2485388219356537, + "learning_rate": 1.8031496062992126e-06, + "loss": 0.493, + "step": 7715 + }, + { + "epoch": 0.6395358474927476, + "grad_norm": 0.24952785670757294, + "learning_rate": 1.8027351844177374e-06, + "loss": 0.5012, + "step": 7716 + }, + { + "epoch": 0.6396187318690427, + "grad_norm": 0.2547861337661743, + "learning_rate": 1.802320762536262e-06, + "loss": 0.5319, + "step": 7717 + }, + { + "epoch": 0.6397016162453377, + "grad_norm": 0.2714291214942932, + "learning_rate": 1.8019063406547867e-06, + "loss": 0.5125, + "step": 7718 + }, + { + "epoch": 0.6397845006216328, + "grad_norm": 0.22455944120883942, + "learning_rate": 1.8014919187733112e-06, + "loss": 0.4531, + "step": 7719 + }, + { + "epoch": 0.639867384997928, + "grad_norm": 0.2411024123430252, + "learning_rate": 1.801077496891836e-06, + "loss": 0.4695, + "step": 7720 + }, + { + "epoch": 0.639950269374223, + "grad_norm": 0.2511427402496338, + "learning_rate": 1.8006630750103607e-06, + "loss": 0.4908, + "step": 7721 + }, + { + "epoch": 0.6400331537505181, + "grad_norm": 0.26337742805480957, + "learning_rate": 1.8002486531288855e-06, + "loss": 0.562, + "step": 7722 + }, + { + "epoch": 0.6401160381268131, + "grad_norm": 0.25398021936416626, + "learning_rate": 1.7998342312474102e-06, + "loss": 0.4732, + "step": 7723 + }, + { + "epoch": 0.6401989225031082, + "grad_norm": 0.2392445057630539, + "learning_rate": 1.7994198093659346e-06, + "loss": 0.4784, + "step": 7724 + }, + { + "epoch": 0.6402818068794033, + "grad_norm": 0.25457483530044556, + "learning_rate": 1.7990053874844592e-06, + "loss": 0.5129, + "step": 7725 + }, + { + "epoch": 0.6403646912556983, + "grad_norm": 0.2501407265663147, + "learning_rate": 1.798590965602984e-06, + "loss": 0.5173, + "step": 7726 + }, + { + "epoch": 0.6404475756319934, + "grad_norm": 0.24539750814437866, + "learning_rate": 1.7981765437215087e-06, + "loss": 0.4815, + "step": 7727 + }, + { + "epoch": 0.6405304600082884, + "grad_norm": 0.2515053153038025, + "learning_rate": 1.7977621218400332e-06, + "loss": 0.4916, + "step": 7728 + }, + { + "epoch": 0.6406133443845835, + "grad_norm": 0.2444605529308319, + "learning_rate": 1.7973476999585578e-06, + "loss": 0.4774, + "step": 7729 + }, + { + "epoch": 0.6406962287608786, + "grad_norm": 0.2565617561340332, + "learning_rate": 1.7969332780770827e-06, + "loss": 0.5184, + "step": 7730 + }, + { + "epoch": 0.6407791131371736, + "grad_norm": 0.2675284147262573, + "learning_rate": 1.7965188561956073e-06, + "loss": 0.5383, + "step": 7731 + }, + { + "epoch": 0.6408619975134687, + "grad_norm": 0.23677098751068115, + "learning_rate": 1.7961044343141322e-06, + "loss": 0.4875, + "step": 7732 + }, + { + "epoch": 0.6409448818897637, + "grad_norm": 0.24390548467636108, + "learning_rate": 1.7956900124326566e-06, + "loss": 0.4763, + "step": 7733 + }, + { + "epoch": 0.6410277662660588, + "grad_norm": 0.25217244029045105, + "learning_rate": 1.7952755905511812e-06, + "loss": 0.4875, + "step": 7734 + }, + { + "epoch": 0.641110650642354, + "grad_norm": 0.2443474680185318, + "learning_rate": 1.7948611686697059e-06, + "loss": 0.5176, + "step": 7735 + }, + { + "epoch": 0.641193535018649, + "grad_norm": 0.2529766261577606, + "learning_rate": 1.7944467467882307e-06, + "loss": 0.5381, + "step": 7736 + }, + { + "epoch": 0.6412764193949441, + "grad_norm": 0.24812041223049164, + "learning_rate": 1.7940323249067552e-06, + "loss": 0.4614, + "step": 7737 + }, + { + "epoch": 0.6413593037712392, + "grad_norm": 0.25929105281829834, + "learning_rate": 1.7936179030252798e-06, + "loss": 0.5141, + "step": 7738 + }, + { + "epoch": 0.6414421881475342, + "grad_norm": 0.25131797790527344, + "learning_rate": 1.7932034811438045e-06, + "loss": 0.5303, + "step": 7739 + }, + { + "epoch": 0.6415250725238293, + "grad_norm": 0.24478642642498016, + "learning_rate": 1.7927890592623293e-06, + "loss": 0.4772, + "step": 7740 + }, + { + "epoch": 0.6416079569001243, + "grad_norm": 0.26448675990104675, + "learning_rate": 1.792374637380854e-06, + "loss": 0.5394, + "step": 7741 + }, + { + "epoch": 0.6416908412764194, + "grad_norm": 0.24396726489067078, + "learning_rate": 1.7919602154993784e-06, + "loss": 0.5054, + "step": 7742 + }, + { + "epoch": 0.6417737256527145, + "grad_norm": 0.23974889516830444, + "learning_rate": 1.7915457936179032e-06, + "loss": 0.5055, + "step": 7743 + }, + { + "epoch": 0.6418566100290095, + "grad_norm": 0.25443148612976074, + "learning_rate": 1.7911313717364279e-06, + "loss": 0.4585, + "step": 7744 + }, + { + "epoch": 0.6419394944053046, + "grad_norm": 0.24035565555095673, + "learning_rate": 1.7907169498549525e-06, + "loss": 0.5089, + "step": 7745 + }, + { + "epoch": 0.6420223787815996, + "grad_norm": 0.2581258714199066, + "learning_rate": 1.790302527973477e-06, + "loss": 0.5056, + "step": 7746 + }, + { + "epoch": 0.6421052631578947, + "grad_norm": 0.24253590404987335, + "learning_rate": 1.7898881060920018e-06, + "loss": 0.491, + "step": 7747 + }, + { + "epoch": 0.6421881475341898, + "grad_norm": 0.2412966936826706, + "learning_rate": 1.7894736842105265e-06, + "loss": 0.52, + "step": 7748 + }, + { + "epoch": 0.6422710319104848, + "grad_norm": 0.2326711267232895, + "learning_rate": 1.789059262329051e-06, + "loss": 0.4846, + "step": 7749 + }, + { + "epoch": 0.64235391628678, + "grad_norm": 0.24665701389312744, + "learning_rate": 1.788644840447576e-06, + "loss": 0.5099, + "step": 7750 + }, + { + "epoch": 0.642436800663075, + "grad_norm": 0.2616342604160309, + "learning_rate": 1.7882304185661004e-06, + "loss": 0.4799, + "step": 7751 + }, + { + "epoch": 0.6425196850393701, + "grad_norm": 0.24989037215709686, + "learning_rate": 1.787815996684625e-06, + "loss": 0.4922, + "step": 7752 + }, + { + "epoch": 0.6426025694156652, + "grad_norm": 0.2601117789745331, + "learning_rate": 1.7874015748031499e-06, + "loss": 0.5106, + "step": 7753 + }, + { + "epoch": 0.6426854537919602, + "grad_norm": 0.2518794536590576, + "learning_rate": 1.7869871529216745e-06, + "loss": 0.5409, + "step": 7754 + }, + { + "epoch": 0.6427683381682553, + "grad_norm": 0.24436146020889282, + "learning_rate": 1.786572731040199e-06, + "loss": 0.5012, + "step": 7755 + }, + { + "epoch": 0.6428512225445504, + "grad_norm": 0.2995038330554962, + "learning_rate": 1.7861583091587236e-06, + "loss": 0.5234, + "step": 7756 + }, + { + "epoch": 0.6429341069208454, + "grad_norm": 0.2434602975845337, + "learning_rate": 1.7857438872772485e-06, + "loss": 0.4881, + "step": 7757 + }, + { + "epoch": 0.6430169912971405, + "grad_norm": 0.23761247098445892, + "learning_rate": 1.785329465395773e-06, + "loss": 0.464, + "step": 7758 + }, + { + "epoch": 0.6430998756734355, + "grad_norm": 0.2583456039428711, + "learning_rate": 1.7849150435142975e-06, + "loss": 0.4965, + "step": 7759 + }, + { + "epoch": 0.6431827600497306, + "grad_norm": 0.23872028291225433, + "learning_rate": 1.7845006216328222e-06, + "loss": 0.5234, + "step": 7760 + }, + { + "epoch": 0.6432656444260257, + "grad_norm": 0.24730566143989563, + "learning_rate": 1.784086199751347e-06, + "loss": 0.494, + "step": 7761 + }, + { + "epoch": 0.6433485288023207, + "grad_norm": 0.24383623898029327, + "learning_rate": 1.7836717778698717e-06, + "loss": 0.5718, + "step": 7762 + }, + { + "epoch": 0.6434314131786159, + "grad_norm": 0.23658230900764465, + "learning_rate": 1.7832573559883965e-06, + "loss": 0.494, + "step": 7763 + }, + { + "epoch": 0.6435142975549109, + "grad_norm": 0.2595076858997345, + "learning_rate": 1.782842934106921e-06, + "loss": 0.5057, + "step": 7764 + }, + { + "epoch": 0.643597181931206, + "grad_norm": 0.24787494540214539, + "learning_rate": 1.7824285122254456e-06, + "loss": 0.4805, + "step": 7765 + }, + { + "epoch": 0.6436800663075011, + "grad_norm": 0.25746461749076843, + "learning_rate": 1.7820140903439703e-06, + "loss": 0.5252, + "step": 7766 + }, + { + "epoch": 0.6437629506837961, + "grad_norm": 0.2522696554660797, + "learning_rate": 1.7815996684624951e-06, + "loss": 0.531, + "step": 7767 + }, + { + "epoch": 0.6438458350600912, + "grad_norm": 0.2595117390155792, + "learning_rate": 1.7811852465810195e-06, + "loss": 0.5338, + "step": 7768 + }, + { + "epoch": 0.6439287194363862, + "grad_norm": 0.24621225893497467, + "learning_rate": 1.7807708246995442e-06, + "loss": 0.5121, + "step": 7769 + }, + { + "epoch": 0.6440116038126813, + "grad_norm": 0.251640647649765, + "learning_rate": 1.7803564028180688e-06, + "loss": 0.5155, + "step": 7770 + }, + { + "epoch": 0.6440944881889764, + "grad_norm": 0.23084083199501038, + "learning_rate": 1.7799419809365937e-06, + "loss": 0.4937, + "step": 7771 + }, + { + "epoch": 0.6441773725652714, + "grad_norm": 0.2669827342033386, + "learning_rate": 1.7795275590551183e-06, + "loss": 0.5347, + "step": 7772 + }, + { + "epoch": 0.6442602569415665, + "grad_norm": 0.26042604446411133, + "learning_rate": 1.7791131371736428e-06, + "loss": 0.5251, + "step": 7773 + }, + { + "epoch": 0.6443431413178616, + "grad_norm": 0.2505899667739868, + "learning_rate": 1.7786987152921676e-06, + "loss": 0.5029, + "step": 7774 + }, + { + "epoch": 0.6444260256941566, + "grad_norm": 0.2521170973777771, + "learning_rate": 1.7782842934106923e-06, + "loss": 0.5086, + "step": 7775 + }, + { + "epoch": 0.6445089100704517, + "grad_norm": 0.25953057408332825, + "learning_rate": 1.777869871529217e-06, + "loss": 0.5367, + "step": 7776 + }, + { + "epoch": 0.6445917944467467, + "grad_norm": 0.22475972771644592, + "learning_rate": 1.7774554496477413e-06, + "loss": 0.474, + "step": 7777 + }, + { + "epoch": 0.6446746788230419, + "grad_norm": 0.22308114171028137, + "learning_rate": 1.7770410277662662e-06, + "loss": 0.45, + "step": 7778 + }, + { + "epoch": 0.644757563199337, + "grad_norm": 0.2563580870628357, + "learning_rate": 1.7766266058847908e-06, + "loss": 0.5294, + "step": 7779 + }, + { + "epoch": 0.644840447575632, + "grad_norm": 0.24953563511371613, + "learning_rate": 1.7762121840033155e-06, + "loss": 0.5573, + "step": 7780 + }, + { + "epoch": 0.6449233319519271, + "grad_norm": 0.2521913945674896, + "learning_rate": 1.7757977621218403e-06, + "loss": 0.4873, + "step": 7781 + }, + { + "epoch": 0.6450062163282221, + "grad_norm": 0.24280312657356262, + "learning_rate": 1.7753833402403648e-06, + "loss": 0.512, + "step": 7782 + }, + { + "epoch": 0.6450891007045172, + "grad_norm": 0.25078389048576355, + "learning_rate": 1.7749689183588894e-06, + "loss": 0.4969, + "step": 7783 + }, + { + "epoch": 0.6451719850808123, + "grad_norm": 0.24926304817199707, + "learning_rate": 1.7745544964774143e-06, + "loss": 0.5544, + "step": 7784 + }, + { + "epoch": 0.6452548694571073, + "grad_norm": 0.2626540958881378, + "learning_rate": 1.774140074595939e-06, + "loss": 0.5311, + "step": 7785 + }, + { + "epoch": 0.6453377538334024, + "grad_norm": 0.24160194396972656, + "learning_rate": 1.7737256527144633e-06, + "loss": 0.4857, + "step": 7786 + }, + { + "epoch": 0.6454206382096974, + "grad_norm": 0.2350286990404129, + "learning_rate": 1.773311230832988e-06, + "loss": 0.4586, + "step": 7787 + }, + { + "epoch": 0.6455035225859925, + "grad_norm": 0.24653245508670807, + "learning_rate": 1.7728968089515128e-06, + "loss": 0.5033, + "step": 7788 + }, + { + "epoch": 0.6455864069622876, + "grad_norm": 0.22938792407512665, + "learning_rate": 1.7724823870700375e-06, + "loss": 0.4604, + "step": 7789 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 0.23871803283691406, + "learning_rate": 1.7720679651885621e-06, + "loss": 0.5023, + "step": 7790 + }, + { + "epoch": 0.6457521757148778, + "grad_norm": 0.24862299859523773, + "learning_rate": 1.7716535433070868e-06, + "loss": 0.4795, + "step": 7791 + }, + { + "epoch": 0.6458350600911729, + "grad_norm": 0.25013405084609985, + "learning_rate": 1.7712391214256114e-06, + "loss": 0.4897, + "step": 7792 + }, + { + "epoch": 0.6459179444674679, + "grad_norm": 0.2495444118976593, + "learning_rate": 1.770824699544136e-06, + "loss": 0.49, + "step": 7793 + }, + { + "epoch": 0.646000828843763, + "grad_norm": 0.23844560980796814, + "learning_rate": 1.770410277662661e-06, + "loss": 0.4626, + "step": 7794 + }, + { + "epoch": 0.646083713220058, + "grad_norm": 0.24260585010051727, + "learning_rate": 1.7699958557811853e-06, + "loss": 0.5106, + "step": 7795 + }, + { + "epoch": 0.6461665975963531, + "grad_norm": 0.2570129334926605, + "learning_rate": 1.76958143389971e-06, + "loss": 0.4904, + "step": 7796 + }, + { + "epoch": 0.6462494819726482, + "grad_norm": 0.24316009879112244, + "learning_rate": 1.7691670120182346e-06, + "loss": 0.5223, + "step": 7797 + }, + { + "epoch": 0.6463323663489432, + "grad_norm": 0.24332979321479797, + "learning_rate": 1.7687525901367595e-06, + "loss": 0.4889, + "step": 7798 + }, + { + "epoch": 0.6464152507252383, + "grad_norm": 0.23439987003803253, + "learning_rate": 1.7683381682552841e-06, + "loss": 0.4896, + "step": 7799 + }, + { + "epoch": 0.6464981351015333, + "grad_norm": 0.24991700053215027, + "learning_rate": 1.7679237463738086e-06, + "loss": 0.5406, + "step": 7800 + }, + { + "epoch": 0.6465810194778284, + "grad_norm": 0.23675890266895294, + "learning_rate": 1.7675093244923334e-06, + "loss": 0.4801, + "step": 7801 + }, + { + "epoch": 0.6466639038541235, + "grad_norm": 0.24231281876564026, + "learning_rate": 1.767094902610858e-06, + "loss": 0.4966, + "step": 7802 + }, + { + "epoch": 0.6467467882304185, + "grad_norm": 0.24822701513767242, + "learning_rate": 1.7666804807293827e-06, + "loss": 0.5089, + "step": 7803 + }, + { + "epoch": 0.6468296726067136, + "grad_norm": 0.2680915296077728, + "learning_rate": 1.7662660588479071e-06, + "loss": 0.5479, + "step": 7804 + }, + { + "epoch": 0.6469125569830088, + "grad_norm": 0.2379804104566574, + "learning_rate": 1.765851636966432e-06, + "loss": 0.4782, + "step": 7805 + }, + { + "epoch": 0.6469954413593038, + "grad_norm": 0.25839662551879883, + "learning_rate": 1.7654372150849566e-06, + "loss": 0.4876, + "step": 7806 + }, + { + "epoch": 0.6470783257355989, + "grad_norm": 0.24441909790039062, + "learning_rate": 1.7650227932034813e-06, + "loss": 0.5314, + "step": 7807 + }, + { + "epoch": 0.6471612101118939, + "grad_norm": 0.250338613986969, + "learning_rate": 1.7646083713220061e-06, + "loss": 0.4827, + "step": 7808 + }, + { + "epoch": 0.647244094488189, + "grad_norm": 0.23741836845874786, + "learning_rate": 1.7641939494405306e-06, + "loss": 0.4904, + "step": 7809 + }, + { + "epoch": 0.6473269788644841, + "grad_norm": 0.2232343852519989, + "learning_rate": 1.7637795275590552e-06, + "loss": 0.4722, + "step": 7810 + }, + { + "epoch": 0.6474098632407791, + "grad_norm": 0.22239306569099426, + "learning_rate": 1.76336510567758e-06, + "loss": 0.4768, + "step": 7811 + }, + { + "epoch": 0.6474927476170742, + "grad_norm": 0.2456687092781067, + "learning_rate": 1.7629506837961047e-06, + "loss": 0.5373, + "step": 7812 + }, + { + "epoch": 0.6475756319933692, + "grad_norm": 0.2553707957267761, + "learning_rate": 1.7625362619146291e-06, + "loss": 0.5306, + "step": 7813 + }, + { + "epoch": 0.6476585163696643, + "grad_norm": 0.6216596364974976, + "learning_rate": 1.7621218400331538e-06, + "loss": 0.4933, + "step": 7814 + }, + { + "epoch": 0.6477414007459594, + "grad_norm": 0.28101983666419983, + "learning_rate": 1.7617074181516786e-06, + "loss": 0.5253, + "step": 7815 + }, + { + "epoch": 0.6478242851222544, + "grad_norm": 0.26170027256011963, + "learning_rate": 1.7612929962702033e-06, + "loss": 0.5588, + "step": 7816 + }, + { + "epoch": 0.6479071694985495, + "grad_norm": 0.26070553064346313, + "learning_rate": 1.760878574388728e-06, + "loss": 0.5355, + "step": 7817 + }, + { + "epoch": 0.6479900538748445, + "grad_norm": 0.25692567229270935, + "learning_rate": 1.7604641525072524e-06, + "loss": 0.5338, + "step": 7818 + }, + { + "epoch": 0.6480729382511397, + "grad_norm": 0.24368374049663544, + "learning_rate": 1.7600497306257772e-06, + "loss": 0.4768, + "step": 7819 + }, + { + "epoch": 0.6481558226274348, + "grad_norm": 0.25318941473960876, + "learning_rate": 1.7596353087443019e-06, + "loss": 0.5407, + "step": 7820 + }, + { + "epoch": 0.6482387070037298, + "grad_norm": 0.23098093271255493, + "learning_rate": 1.7592208868628267e-06, + "loss": 0.479, + "step": 7821 + }, + { + "epoch": 0.6483215913800249, + "grad_norm": 0.2420310080051422, + "learning_rate": 1.7588064649813511e-06, + "loss": 0.5115, + "step": 7822 + }, + { + "epoch": 0.64840447575632, + "grad_norm": 0.25238460302352905, + "learning_rate": 1.7583920430998758e-06, + "loss": 0.5194, + "step": 7823 + }, + { + "epoch": 0.648487360132615, + "grad_norm": 0.26453226804733276, + "learning_rate": 1.7579776212184004e-06, + "loss": 0.4916, + "step": 7824 + }, + { + "epoch": 0.6485702445089101, + "grad_norm": 0.2616351246833801, + "learning_rate": 1.7575631993369253e-06, + "loss": 0.4801, + "step": 7825 + }, + { + "epoch": 0.6486531288852051, + "grad_norm": 0.2859729826450348, + "learning_rate": 1.75714877745545e-06, + "loss": 0.5255, + "step": 7826 + }, + { + "epoch": 0.6487360132615002, + "grad_norm": 0.24229682981967926, + "learning_rate": 1.7567343555739744e-06, + "loss": 0.5099, + "step": 7827 + }, + { + "epoch": 0.6488188976377953, + "grad_norm": 0.24668774008750916, + "learning_rate": 1.756319933692499e-06, + "loss": 0.506, + "step": 7828 + }, + { + "epoch": 0.6489017820140903, + "grad_norm": 0.2594258189201355, + "learning_rate": 1.7559055118110239e-06, + "loss": 0.5241, + "step": 7829 + }, + { + "epoch": 0.6489846663903854, + "grad_norm": 0.24059627950191498, + "learning_rate": 1.7554910899295485e-06, + "loss": 0.5661, + "step": 7830 + }, + { + "epoch": 0.6490675507666804, + "grad_norm": 0.2391817420721054, + "learning_rate": 1.755076668048073e-06, + "loss": 0.4873, + "step": 7831 + }, + { + "epoch": 0.6491504351429755, + "grad_norm": 0.23767654597759247, + "learning_rate": 1.7546622461665978e-06, + "loss": 0.4953, + "step": 7832 + }, + { + "epoch": 0.6492333195192707, + "grad_norm": 0.23950253427028656, + "learning_rate": 1.7542478242851224e-06, + "loss": 0.4784, + "step": 7833 + }, + { + "epoch": 0.6493162038955657, + "grad_norm": 0.3027753233909607, + "learning_rate": 1.753833402403647e-06, + "loss": 0.5522, + "step": 7834 + }, + { + "epoch": 0.6493990882718608, + "grad_norm": 0.24823294579982758, + "learning_rate": 1.7534189805221715e-06, + "loss": 0.5132, + "step": 7835 + }, + { + "epoch": 0.6494819726481558, + "grad_norm": 0.2529648244380951, + "learning_rate": 1.7530045586406964e-06, + "loss": 0.506, + "step": 7836 + }, + { + "epoch": 0.6495648570244509, + "grad_norm": 0.23683106899261475, + "learning_rate": 1.752590136759221e-06, + "loss": 0.4785, + "step": 7837 + }, + { + "epoch": 0.649647741400746, + "grad_norm": 0.24487370252609253, + "learning_rate": 1.7521757148777456e-06, + "loss": 0.4955, + "step": 7838 + }, + { + "epoch": 0.649730625777041, + "grad_norm": 0.2801283597946167, + "learning_rate": 1.7517612929962705e-06, + "loss": 0.5278, + "step": 7839 + }, + { + "epoch": 0.6498135101533361, + "grad_norm": 0.26792845129966736, + "learning_rate": 1.751346871114795e-06, + "loss": 0.4996, + "step": 7840 + }, + { + "epoch": 0.6498963945296312, + "grad_norm": 0.3456627428531647, + "learning_rate": 1.7509324492333196e-06, + "loss": 0.4805, + "step": 7841 + }, + { + "epoch": 0.6499792789059262, + "grad_norm": 0.24254442751407623, + "learning_rate": 1.7505180273518444e-06, + "loss": 0.5064, + "step": 7842 + }, + { + "epoch": 0.6500621632822213, + "grad_norm": 0.2645808756351471, + "learning_rate": 1.750103605470369e-06, + "loss": 0.535, + "step": 7843 + }, + { + "epoch": 0.6501450476585163, + "grad_norm": 0.25302913784980774, + "learning_rate": 1.7496891835888935e-06, + "loss": 0.5196, + "step": 7844 + }, + { + "epoch": 0.6502279320348114, + "grad_norm": 0.22713203728199005, + "learning_rate": 1.7492747617074182e-06, + "loss": 0.471, + "step": 7845 + }, + { + "epoch": 0.6503108164111066, + "grad_norm": 0.25905662775039673, + "learning_rate": 1.748860339825943e-06, + "loss": 0.4789, + "step": 7846 + }, + { + "epoch": 0.6503937007874016, + "grad_norm": 0.2459513545036316, + "learning_rate": 1.7484459179444677e-06, + "loss": 0.4834, + "step": 7847 + }, + { + "epoch": 0.6504765851636967, + "grad_norm": 0.2719389498233795, + "learning_rate": 1.7480314960629923e-06, + "loss": 0.5601, + "step": 7848 + }, + { + "epoch": 0.6505594695399917, + "grad_norm": 0.22424718737602234, + "learning_rate": 1.7476170741815167e-06, + "loss": 0.5081, + "step": 7849 + }, + { + "epoch": 0.6506423539162868, + "grad_norm": 0.26968473196029663, + "learning_rate": 1.7472026523000416e-06, + "loss": 0.5267, + "step": 7850 + }, + { + "epoch": 0.6507252382925819, + "grad_norm": 0.2383064180612564, + "learning_rate": 1.7467882304185662e-06, + "loss": 0.5128, + "step": 7851 + }, + { + "epoch": 0.6508081226688769, + "grad_norm": 0.24002143740653992, + "learning_rate": 1.746373808537091e-06, + "loss": 0.5081, + "step": 7852 + }, + { + "epoch": 0.650891007045172, + "grad_norm": 0.24560049176216125, + "learning_rate": 1.7459593866556155e-06, + "loss": 0.4778, + "step": 7853 + }, + { + "epoch": 0.650973891421467, + "grad_norm": 0.2327665090560913, + "learning_rate": 1.7455449647741402e-06, + "loss": 0.4612, + "step": 7854 + }, + { + "epoch": 0.6510567757977621, + "grad_norm": 0.2598660886287689, + "learning_rate": 1.7451305428926648e-06, + "loss": 0.5049, + "step": 7855 + }, + { + "epoch": 0.6511396601740572, + "grad_norm": 0.23962032794952393, + "learning_rate": 1.7447161210111897e-06, + "loss": 0.4793, + "step": 7856 + }, + { + "epoch": 0.6512225445503522, + "grad_norm": 0.2684350609779358, + "learning_rate": 1.7443016991297143e-06, + "loss": 0.5048, + "step": 7857 + }, + { + "epoch": 0.6513054289266473, + "grad_norm": 0.26617351174354553, + "learning_rate": 1.7438872772482387e-06, + "loss": 0.5449, + "step": 7858 + }, + { + "epoch": 0.6513883133029424, + "grad_norm": 0.25482049584388733, + "learning_rate": 1.7434728553667634e-06, + "loss": 0.5395, + "step": 7859 + }, + { + "epoch": 0.6514711976792374, + "grad_norm": 0.23305180668830872, + "learning_rate": 1.7430584334852882e-06, + "loss": 0.4813, + "step": 7860 + }, + { + "epoch": 0.6515540820555326, + "grad_norm": 0.24485649168491364, + "learning_rate": 1.7426440116038129e-06, + "loss": 0.5125, + "step": 7861 + }, + { + "epoch": 0.6516369664318276, + "grad_norm": 0.24762769043445587, + "learning_rate": 1.7422295897223373e-06, + "loss": 0.5078, + "step": 7862 + }, + { + "epoch": 0.6517198508081227, + "grad_norm": 0.2531333267688751, + "learning_rate": 1.7418151678408622e-06, + "loss": 0.511, + "step": 7863 + }, + { + "epoch": 0.6518027351844178, + "grad_norm": 0.2525922656059265, + "learning_rate": 1.7414007459593868e-06, + "loss": 0.4805, + "step": 7864 + }, + { + "epoch": 0.6518856195607128, + "grad_norm": 0.2449767291545868, + "learning_rate": 1.7409863240779114e-06, + "loss": 0.5067, + "step": 7865 + }, + { + "epoch": 0.6519685039370079, + "grad_norm": 0.2611372172832489, + "learning_rate": 1.7405719021964363e-06, + "loss": 0.5225, + "step": 7866 + }, + { + "epoch": 0.6520513883133029, + "grad_norm": 0.24857650697231293, + "learning_rate": 1.7401574803149607e-06, + "loss": 0.4953, + "step": 7867 + }, + { + "epoch": 0.652134272689598, + "grad_norm": 0.2629993259906769, + "learning_rate": 1.7397430584334854e-06, + "loss": 0.4826, + "step": 7868 + }, + { + "epoch": 0.6522171570658931, + "grad_norm": 0.23750701546669006, + "learning_rate": 1.73932863655201e-06, + "loss": 0.5153, + "step": 7869 + }, + { + "epoch": 0.6523000414421881, + "grad_norm": 0.2311258763074875, + "learning_rate": 1.7389142146705349e-06, + "loss": 0.4344, + "step": 7870 + }, + { + "epoch": 0.6523829258184832, + "grad_norm": 0.24050229787826538, + "learning_rate": 1.7384997927890593e-06, + "loss": 0.4536, + "step": 7871 + }, + { + "epoch": 0.6524658101947783, + "grad_norm": 0.24542291462421417, + "learning_rate": 1.738085370907584e-06, + "loss": 0.4998, + "step": 7872 + }, + { + "epoch": 0.6525486945710733, + "grad_norm": 0.24044160544872284, + "learning_rate": 1.7376709490261088e-06, + "loss": 0.4861, + "step": 7873 + }, + { + "epoch": 0.6526315789473685, + "grad_norm": 0.23353908956050873, + "learning_rate": 1.7372565271446335e-06, + "loss": 0.493, + "step": 7874 + }, + { + "epoch": 0.6527144633236635, + "grad_norm": 0.24285952746868134, + "learning_rate": 1.736842105263158e-06, + "loss": 0.474, + "step": 7875 + }, + { + "epoch": 0.6527973476999586, + "grad_norm": 0.23059210181236267, + "learning_rate": 1.7364276833816825e-06, + "loss": 0.4608, + "step": 7876 + }, + { + "epoch": 0.6528802320762537, + "grad_norm": 0.24564673006534576, + "learning_rate": 1.7360132615002074e-06, + "loss": 0.4839, + "step": 7877 + }, + { + "epoch": 0.6529631164525487, + "grad_norm": 0.24888835847377777, + "learning_rate": 1.735598839618732e-06, + "loss": 0.4951, + "step": 7878 + }, + { + "epoch": 0.6530460008288438, + "grad_norm": 0.24478267133235931, + "learning_rate": 1.7351844177372567e-06, + "loss": 0.465, + "step": 7879 + }, + { + "epoch": 0.6531288852051388, + "grad_norm": 0.26540690660476685, + "learning_rate": 1.7347699958557813e-06, + "loss": 0.5178, + "step": 7880 + }, + { + "epoch": 0.6532117695814339, + "grad_norm": 0.2616828978061676, + "learning_rate": 1.734355573974306e-06, + "loss": 0.5536, + "step": 7881 + }, + { + "epoch": 0.653294653957729, + "grad_norm": 0.23513266444206238, + "learning_rate": 1.7339411520928306e-06, + "loss": 0.4752, + "step": 7882 + }, + { + "epoch": 0.653377538334024, + "grad_norm": 0.25621432065963745, + "learning_rate": 1.7335267302113555e-06, + "loss": 0.509, + "step": 7883 + }, + { + "epoch": 0.6534604227103191, + "grad_norm": 0.24743317067623138, + "learning_rate": 1.73311230832988e-06, + "loss": 0.4766, + "step": 7884 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 0.24016372859477997, + "learning_rate": 1.7326978864484045e-06, + "loss": 0.4545, + "step": 7885 + }, + { + "epoch": 0.6536261914629092, + "grad_norm": 0.23132674396038055, + "learning_rate": 1.7322834645669292e-06, + "loss": 0.4684, + "step": 7886 + }, + { + "epoch": 0.6537090758392043, + "grad_norm": 0.2535393238067627, + "learning_rate": 1.731869042685454e-06, + "loss": 0.5059, + "step": 7887 + }, + { + "epoch": 0.6537919602154993, + "grad_norm": 0.26022979617118835, + "learning_rate": 1.7314546208039787e-06, + "loss": 0.5744, + "step": 7888 + }, + { + "epoch": 0.6538748445917945, + "grad_norm": 0.24657602608203888, + "learning_rate": 1.731040198922503e-06, + "loss": 0.5054, + "step": 7889 + }, + { + "epoch": 0.6539577289680896, + "grad_norm": 0.2438710480928421, + "learning_rate": 1.730625777041028e-06, + "loss": 0.5001, + "step": 7890 + }, + { + "epoch": 0.6540406133443846, + "grad_norm": 0.24051623046398163, + "learning_rate": 1.7302113551595526e-06, + "loss": 0.5009, + "step": 7891 + }, + { + "epoch": 0.6541234977206797, + "grad_norm": 0.242283433675766, + "learning_rate": 1.7297969332780772e-06, + "loss": 0.4403, + "step": 7892 + }, + { + "epoch": 0.6542063820969747, + "grad_norm": 0.25167417526245117, + "learning_rate": 1.729382511396602e-06, + "loss": 0.5022, + "step": 7893 + }, + { + "epoch": 0.6542892664732698, + "grad_norm": 0.26350274682044983, + "learning_rate": 1.7289680895151265e-06, + "loss": 0.5409, + "step": 7894 + }, + { + "epoch": 0.6543721508495649, + "grad_norm": 0.2544015049934387, + "learning_rate": 1.7285536676336512e-06, + "loss": 0.5432, + "step": 7895 + }, + { + "epoch": 0.6544550352258599, + "grad_norm": 0.2505149245262146, + "learning_rate": 1.7281392457521758e-06, + "loss": 0.5019, + "step": 7896 + }, + { + "epoch": 0.654537919602155, + "grad_norm": 0.23337474465370178, + "learning_rate": 1.7277248238707007e-06, + "loss": 0.478, + "step": 7897 + }, + { + "epoch": 0.65462080397845, + "grad_norm": 0.2426503449678421, + "learning_rate": 1.7273104019892251e-06, + "loss": 0.4685, + "step": 7898 + }, + { + "epoch": 0.6547036883547451, + "grad_norm": 0.23869533836841583, + "learning_rate": 1.7268959801077498e-06, + "loss": 0.4524, + "step": 7899 + }, + { + "epoch": 0.6547865727310402, + "grad_norm": 0.2509955167770386, + "learning_rate": 1.7264815582262744e-06, + "loss": 0.5029, + "step": 7900 + }, + { + "epoch": 0.6548694571073352, + "grad_norm": 0.3825349509716034, + "learning_rate": 1.7260671363447993e-06, + "loss": 0.6004, + "step": 7901 + }, + { + "epoch": 0.6549523414836304, + "grad_norm": 0.2163592427968979, + "learning_rate": 1.725652714463324e-06, + "loss": 0.4493, + "step": 7902 + }, + { + "epoch": 0.6550352258599254, + "grad_norm": 0.2525366246700287, + "learning_rate": 1.7252382925818483e-06, + "loss": 0.5161, + "step": 7903 + }, + { + "epoch": 0.6551181102362205, + "grad_norm": 0.25721845030784607, + "learning_rate": 1.7248238707003732e-06, + "loss": 0.5027, + "step": 7904 + }, + { + "epoch": 0.6552009946125156, + "grad_norm": 0.23486793041229248, + "learning_rate": 1.7244094488188978e-06, + "loss": 0.4652, + "step": 7905 + }, + { + "epoch": 0.6552838789888106, + "grad_norm": 0.24502864480018616, + "learning_rate": 1.7239950269374225e-06, + "loss": 0.4516, + "step": 7906 + }, + { + "epoch": 0.6553667633651057, + "grad_norm": 0.23420147597789764, + "learning_rate": 1.723580605055947e-06, + "loss": 0.4795, + "step": 7907 + }, + { + "epoch": 0.6554496477414008, + "grad_norm": 0.2618664503097534, + "learning_rate": 1.7231661831744718e-06, + "loss": 0.4827, + "step": 7908 + }, + { + "epoch": 0.6555325321176958, + "grad_norm": 0.269438773393631, + "learning_rate": 1.7227517612929964e-06, + "loss": 0.5024, + "step": 7909 + }, + { + "epoch": 0.6556154164939909, + "grad_norm": 0.26682141423225403, + "learning_rate": 1.722337339411521e-06, + "loss": 0.5397, + "step": 7910 + }, + { + "epoch": 0.6556983008702859, + "grad_norm": 0.2433667927980423, + "learning_rate": 1.7219229175300457e-06, + "loss": 0.4892, + "step": 7911 + }, + { + "epoch": 0.655781185246581, + "grad_norm": 0.24291662871837616, + "learning_rate": 1.7215084956485703e-06, + "loss": 0.4993, + "step": 7912 + }, + { + "epoch": 0.6558640696228761, + "grad_norm": 0.2820020318031311, + "learning_rate": 1.721094073767095e-06, + "loss": 0.5486, + "step": 7913 + }, + { + "epoch": 0.6559469539991711, + "grad_norm": 0.24336910247802734, + "learning_rate": 1.7206796518856198e-06, + "loss": 0.4815, + "step": 7914 + }, + { + "epoch": 0.6560298383754662, + "grad_norm": 0.23381608724594116, + "learning_rate": 1.7202652300041445e-06, + "loss": 0.5298, + "step": 7915 + }, + { + "epoch": 0.6561127227517612, + "grad_norm": 0.2620185911655426, + "learning_rate": 1.719850808122669e-06, + "loss": 0.5174, + "step": 7916 + }, + { + "epoch": 0.6561956071280564, + "grad_norm": 0.25843536853790283, + "learning_rate": 1.7194363862411935e-06, + "loss": 0.5151, + "step": 7917 + }, + { + "epoch": 0.6562784915043515, + "grad_norm": 0.24251143634319305, + "learning_rate": 1.7190219643597184e-06, + "loss": 0.4837, + "step": 7918 + }, + { + "epoch": 0.6563613758806465, + "grad_norm": 0.2526211142539978, + "learning_rate": 1.718607542478243e-06, + "loss": 0.5051, + "step": 7919 + }, + { + "epoch": 0.6564442602569416, + "grad_norm": 0.2496795505285263, + "learning_rate": 1.7181931205967675e-06, + "loss": 0.5059, + "step": 7920 + }, + { + "epoch": 0.6565271446332366, + "grad_norm": 0.25373122096061707, + "learning_rate": 1.7177786987152923e-06, + "loss": 0.5517, + "step": 7921 + }, + { + "epoch": 0.6566100290095317, + "grad_norm": 0.24372567236423492, + "learning_rate": 1.717364276833817e-06, + "loss": 0.5049, + "step": 7922 + }, + { + "epoch": 0.6566929133858268, + "grad_norm": 0.2514585256576538, + "learning_rate": 1.7169498549523416e-06, + "loss": 0.5424, + "step": 7923 + }, + { + "epoch": 0.6567757977621218, + "grad_norm": 0.251819372177124, + "learning_rate": 1.7165354330708665e-06, + "loss": 0.4965, + "step": 7924 + }, + { + "epoch": 0.6568586821384169, + "grad_norm": 0.23979401588439941, + "learning_rate": 1.716121011189391e-06, + "loss": 0.4899, + "step": 7925 + }, + { + "epoch": 0.656941566514712, + "grad_norm": 0.24322205781936646, + "learning_rate": 1.7157065893079156e-06, + "loss": 0.5145, + "step": 7926 + }, + { + "epoch": 0.657024450891007, + "grad_norm": 0.25255465507507324, + "learning_rate": 1.7152921674264402e-06, + "loss": 0.4832, + "step": 7927 + }, + { + "epoch": 0.6571073352673021, + "grad_norm": 0.27114248275756836, + "learning_rate": 1.714877745544965e-06, + "loss": 0.5212, + "step": 7928 + }, + { + "epoch": 0.6571902196435971, + "grad_norm": 0.24916893243789673, + "learning_rate": 1.7144633236634895e-06, + "loss": 0.53, + "step": 7929 + }, + { + "epoch": 0.6572731040198923, + "grad_norm": 0.25640642642974854, + "learning_rate": 1.7140489017820141e-06, + "loss": 0.5096, + "step": 7930 + }, + { + "epoch": 0.6573559883961874, + "grad_norm": 0.22368453443050385, + "learning_rate": 1.713634479900539e-06, + "loss": 0.5076, + "step": 7931 + }, + { + "epoch": 0.6574388727724824, + "grad_norm": 0.2398207187652588, + "learning_rate": 1.7132200580190636e-06, + "loss": 0.512, + "step": 7932 + }, + { + "epoch": 0.6575217571487775, + "grad_norm": 0.2642784118652344, + "learning_rate": 1.7128056361375883e-06, + "loss": 0.5041, + "step": 7933 + }, + { + "epoch": 0.6576046415250725, + "grad_norm": 0.25452640652656555, + "learning_rate": 1.7123912142561127e-06, + "loss": 0.4653, + "step": 7934 + }, + { + "epoch": 0.6576875259013676, + "grad_norm": 0.2526521682739258, + "learning_rate": 1.7119767923746376e-06, + "loss": 0.4937, + "step": 7935 + }, + { + "epoch": 0.6577704102776627, + "grad_norm": 0.26008909940719604, + "learning_rate": 1.7115623704931622e-06, + "loss": 0.496, + "step": 7936 + }, + { + "epoch": 0.6578532946539577, + "grad_norm": 0.24040435254573822, + "learning_rate": 1.7111479486116868e-06, + "loss": 0.5056, + "step": 7937 + }, + { + "epoch": 0.6579361790302528, + "grad_norm": 0.2831359803676605, + "learning_rate": 1.7107335267302113e-06, + "loss": 0.5094, + "step": 7938 + }, + { + "epoch": 0.6580190634065478, + "grad_norm": 0.2512769401073456, + "learning_rate": 1.7103191048487361e-06, + "loss": 0.4582, + "step": 7939 + }, + { + "epoch": 0.6581019477828429, + "grad_norm": 0.237775519490242, + "learning_rate": 1.7099046829672608e-06, + "loss": 0.4729, + "step": 7940 + }, + { + "epoch": 0.658184832159138, + "grad_norm": 0.23108866810798645, + "learning_rate": 1.7094902610857856e-06, + "loss": 0.4507, + "step": 7941 + }, + { + "epoch": 0.658267716535433, + "grad_norm": 0.2417508363723755, + "learning_rate": 1.7090758392043103e-06, + "loss": 0.4973, + "step": 7942 + }, + { + "epoch": 0.6583506009117281, + "grad_norm": 0.25631484389305115, + "learning_rate": 1.7086614173228347e-06, + "loss": 0.5251, + "step": 7943 + }, + { + "epoch": 0.6584334852880233, + "grad_norm": 0.2626124322414398, + "learning_rate": 1.7082469954413593e-06, + "loss": 0.4984, + "step": 7944 + }, + { + "epoch": 0.6585163696643183, + "grad_norm": 0.27709490060806274, + "learning_rate": 1.7078325735598842e-06, + "loss": 0.5253, + "step": 7945 + }, + { + "epoch": 0.6585992540406134, + "grad_norm": 0.25556910037994385, + "learning_rate": 1.7074181516784088e-06, + "loss": 0.5101, + "step": 7946 + }, + { + "epoch": 0.6586821384169084, + "grad_norm": 0.26009708642959595, + "learning_rate": 1.7070037297969333e-06, + "loss": 0.5251, + "step": 7947 + }, + { + "epoch": 0.6587650227932035, + "grad_norm": 0.24332624673843384, + "learning_rate": 1.706589307915458e-06, + "loss": 0.5103, + "step": 7948 + }, + { + "epoch": 0.6588479071694986, + "grad_norm": 0.2520004212856293, + "learning_rate": 1.7061748860339828e-06, + "loss": 0.478, + "step": 7949 + }, + { + "epoch": 0.6589307915457936, + "grad_norm": 0.24779875576496124, + "learning_rate": 1.7057604641525074e-06, + "loss": 0.5119, + "step": 7950 + }, + { + "epoch": 0.6590136759220887, + "grad_norm": 0.25731921195983887, + "learning_rate": 1.7053460422710323e-06, + "loss": 0.4837, + "step": 7951 + }, + { + "epoch": 0.6590965602983837, + "grad_norm": 0.24882179498672485, + "learning_rate": 1.7049316203895567e-06, + "loss": 0.5234, + "step": 7952 + }, + { + "epoch": 0.6591794446746788, + "grad_norm": 0.2448381632566452, + "learning_rate": 1.7045171985080814e-06, + "loss": 0.5157, + "step": 7953 + }, + { + "epoch": 0.6592623290509739, + "grad_norm": 0.24876268208026886, + "learning_rate": 1.704102776626606e-06, + "loss": 0.5208, + "step": 7954 + }, + { + "epoch": 0.6593452134272689, + "grad_norm": 0.24749276041984558, + "learning_rate": 1.7036883547451308e-06, + "loss": 0.5302, + "step": 7955 + }, + { + "epoch": 0.659428097803564, + "grad_norm": 0.2580161392688751, + "learning_rate": 1.7032739328636553e-06, + "loss": 0.4746, + "step": 7956 + }, + { + "epoch": 0.6595109821798592, + "grad_norm": 0.2472294569015503, + "learning_rate": 1.70285951098218e-06, + "loss": 0.493, + "step": 7957 + }, + { + "epoch": 0.6595938665561542, + "grad_norm": 0.25723570585250854, + "learning_rate": 1.7024450891007046e-06, + "loss": 0.5292, + "step": 7958 + }, + { + "epoch": 0.6596767509324493, + "grad_norm": 0.2781605124473572, + "learning_rate": 1.7020306672192294e-06, + "loss": 0.563, + "step": 7959 + }, + { + "epoch": 0.6597596353087443, + "grad_norm": 0.2584344744682312, + "learning_rate": 1.701616245337754e-06, + "loss": 0.5462, + "step": 7960 + }, + { + "epoch": 0.6598425196850394, + "grad_norm": 0.2521789073944092, + "learning_rate": 1.7012018234562785e-06, + "loss": 0.5122, + "step": 7961 + }, + { + "epoch": 0.6599254040613345, + "grad_norm": 0.2634831368923187, + "learning_rate": 1.7007874015748034e-06, + "loss": 0.5421, + "step": 7962 + }, + { + "epoch": 0.6600082884376295, + "grad_norm": 0.25734013319015503, + "learning_rate": 1.700372979693328e-06, + "loss": 0.5161, + "step": 7963 + }, + { + "epoch": 0.6600911728139246, + "grad_norm": 0.251810222864151, + "learning_rate": 1.6999585578118526e-06, + "loss": 0.5187, + "step": 7964 + }, + { + "epoch": 0.6601740571902196, + "grad_norm": 0.236708402633667, + "learning_rate": 1.699544135930377e-06, + "loss": 0.5042, + "step": 7965 + }, + { + "epoch": 0.6602569415665147, + "grad_norm": 0.24985726177692413, + "learning_rate": 1.699129714048902e-06, + "loss": 0.499, + "step": 7966 + }, + { + "epoch": 0.6603398259428098, + "grad_norm": 0.24493515491485596, + "learning_rate": 1.6987152921674266e-06, + "loss": 0.465, + "step": 7967 + }, + { + "epoch": 0.6604227103191048, + "grad_norm": 0.23378655314445496, + "learning_rate": 1.6983008702859512e-06, + "loss": 0.4994, + "step": 7968 + }, + { + "epoch": 0.6605055946953999, + "grad_norm": 0.2524559795856476, + "learning_rate": 1.697886448404476e-06, + "loss": 0.5263, + "step": 7969 + }, + { + "epoch": 0.6605884790716949, + "grad_norm": 0.22494979202747345, + "learning_rate": 1.6974720265230005e-06, + "loss": 0.4688, + "step": 7970 + }, + { + "epoch": 0.66067136344799, + "grad_norm": 0.2590310871601105, + "learning_rate": 1.6970576046415251e-06, + "loss": 0.511, + "step": 7971 + }, + { + "epoch": 0.6607542478242852, + "grad_norm": 0.28386229276657104, + "learning_rate": 1.69664318276005e-06, + "loss": 0.4952, + "step": 7972 + }, + { + "epoch": 0.6608371322005802, + "grad_norm": 0.25496533513069153, + "learning_rate": 1.6962287608785746e-06, + "loss": 0.5027, + "step": 7973 + }, + { + "epoch": 0.6609200165768753, + "grad_norm": 0.24557948112487793, + "learning_rate": 1.695814338997099e-06, + "loss": 0.5194, + "step": 7974 + }, + { + "epoch": 0.6610029009531704, + "grad_norm": 0.2314382940530777, + "learning_rate": 1.6953999171156237e-06, + "loss": 0.4564, + "step": 7975 + }, + { + "epoch": 0.6610857853294654, + "grad_norm": 0.22526215016841888, + "learning_rate": 1.6949854952341486e-06, + "loss": 0.4631, + "step": 7976 + }, + { + "epoch": 0.6611686697057605, + "grad_norm": 0.25050050020217896, + "learning_rate": 1.6945710733526732e-06, + "loss": 0.5216, + "step": 7977 + }, + { + "epoch": 0.6612515540820555, + "grad_norm": 0.2451496422290802, + "learning_rate": 1.6941566514711979e-06, + "loss": 0.537, + "step": 7978 + }, + { + "epoch": 0.6613344384583506, + "grad_norm": 0.23536956310272217, + "learning_rate": 1.6937422295897223e-06, + "loss": 0.4801, + "step": 7979 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 0.25338464975357056, + "learning_rate": 1.6933278077082472e-06, + "loss": 0.5237, + "step": 7980 + }, + { + "epoch": 0.6615002072109407, + "grad_norm": 0.2427174597978592, + "learning_rate": 1.6929133858267718e-06, + "loss": 0.467, + "step": 7981 + }, + { + "epoch": 0.6615830915872358, + "grad_norm": 0.23126228153705597, + "learning_rate": 1.6924989639452966e-06, + "loss": 0.5133, + "step": 7982 + }, + { + "epoch": 0.6616659759635308, + "grad_norm": 0.2650034725666046, + "learning_rate": 1.692084542063821e-06, + "loss": 0.5246, + "step": 7983 + }, + { + "epoch": 0.6617488603398259, + "grad_norm": 0.2261463701725006, + "learning_rate": 1.6916701201823457e-06, + "loss": 0.4885, + "step": 7984 + }, + { + "epoch": 0.661831744716121, + "grad_norm": 0.25075799226760864, + "learning_rate": 1.6912556983008704e-06, + "loss": 0.5401, + "step": 7985 + }, + { + "epoch": 0.661914629092416, + "grad_norm": 0.23204606771469116, + "learning_rate": 1.6908412764193952e-06, + "loss": 0.4715, + "step": 7986 + }, + { + "epoch": 0.6619975134687112, + "grad_norm": 0.23981519043445587, + "learning_rate": 1.6904268545379197e-06, + "loss": 0.4961, + "step": 7987 + }, + { + "epoch": 0.6620803978450062, + "grad_norm": 0.23738932609558105, + "learning_rate": 1.6900124326564443e-06, + "loss": 0.5197, + "step": 7988 + }, + { + "epoch": 0.6621632822213013, + "grad_norm": 0.24658600986003876, + "learning_rate": 1.689598010774969e-06, + "loss": 0.5009, + "step": 7989 + }, + { + "epoch": 0.6622461665975964, + "grad_norm": 0.24416044354438782, + "learning_rate": 1.6891835888934938e-06, + "loss": 0.5119, + "step": 7990 + }, + { + "epoch": 0.6623290509738914, + "grad_norm": 0.2556711435317993, + "learning_rate": 1.6887691670120184e-06, + "loss": 0.5394, + "step": 7991 + }, + { + "epoch": 0.6624119353501865, + "grad_norm": 0.2515352666378021, + "learning_rate": 1.6883547451305429e-06, + "loss": 0.5164, + "step": 7992 + }, + { + "epoch": 0.6624948197264816, + "grad_norm": 0.24788054823875427, + "learning_rate": 1.6879403232490677e-06, + "loss": 0.5044, + "step": 7993 + }, + { + "epoch": 0.6625777041027766, + "grad_norm": 0.23214559257030487, + "learning_rate": 1.6875259013675924e-06, + "loss": 0.4684, + "step": 7994 + }, + { + "epoch": 0.6626605884790717, + "grad_norm": 0.24513059854507446, + "learning_rate": 1.687111479486117e-06, + "loss": 0.5768, + "step": 7995 + }, + { + "epoch": 0.6627434728553667, + "grad_norm": 0.25097548961639404, + "learning_rate": 1.6866970576046414e-06, + "loss": 0.5296, + "step": 7996 + }, + { + "epoch": 0.6628263572316618, + "grad_norm": 0.24778324365615845, + "learning_rate": 1.6862826357231663e-06, + "loss": 0.5048, + "step": 7997 + }, + { + "epoch": 0.662909241607957, + "grad_norm": 0.25305071473121643, + "learning_rate": 1.685868213841691e-06, + "loss": 0.5182, + "step": 7998 + }, + { + "epoch": 0.662992125984252, + "grad_norm": 0.2393908053636551, + "learning_rate": 1.6854537919602156e-06, + "loss": 0.4857, + "step": 7999 + }, + { + "epoch": 0.6630750103605471, + "grad_norm": 0.2677978277206421, + "learning_rate": 1.6850393700787404e-06, + "loss": 0.5566, + "step": 8000 + }, + { + "epoch": 0.6631578947368421, + "grad_norm": 0.2581248879432678, + "learning_rate": 1.6846249481972649e-06, + "loss": 0.5144, + "step": 8001 + }, + { + "epoch": 0.6632407791131372, + "grad_norm": 0.2574666738510132, + "learning_rate": 1.6842105263157895e-06, + "loss": 0.5357, + "step": 8002 + }, + { + "epoch": 0.6633236634894323, + "grad_norm": 0.2420210838317871, + "learning_rate": 1.6837961044343144e-06, + "loss": 0.5063, + "step": 8003 + }, + { + "epoch": 0.6634065478657273, + "grad_norm": 0.2438889890909195, + "learning_rate": 1.683381682552839e-06, + "loss": 0.4607, + "step": 8004 + }, + { + "epoch": 0.6634894322420224, + "grad_norm": 0.23636355996131897, + "learning_rate": 1.6829672606713635e-06, + "loss": 0.4636, + "step": 8005 + }, + { + "epoch": 0.6635723166183174, + "grad_norm": 0.2307988554239273, + "learning_rate": 1.682552838789888e-06, + "loss": 0.4483, + "step": 8006 + }, + { + "epoch": 0.6636552009946125, + "grad_norm": 0.24622422456741333, + "learning_rate": 1.682138416908413e-06, + "loss": 0.4814, + "step": 8007 + }, + { + "epoch": 0.6637380853709076, + "grad_norm": 0.23634539544582367, + "learning_rate": 1.6817239950269376e-06, + "loss": 0.4776, + "step": 8008 + }, + { + "epoch": 0.6638209697472026, + "grad_norm": 0.24329033493995667, + "learning_rate": 1.6813095731454622e-06, + "loss": 0.4755, + "step": 8009 + }, + { + "epoch": 0.6639038541234977, + "grad_norm": 0.23743118345737457, + "learning_rate": 1.6808951512639869e-06, + "loss": 0.4603, + "step": 8010 + }, + { + "epoch": 0.6639867384997928, + "grad_norm": 0.26393792033195496, + "learning_rate": 1.6804807293825115e-06, + "loss": 0.4956, + "step": 8011 + }, + { + "epoch": 0.6640696228760878, + "grad_norm": 0.23465433716773987, + "learning_rate": 1.6800663075010362e-06, + "loss": 0.5079, + "step": 8012 + }, + { + "epoch": 0.664152507252383, + "grad_norm": 0.24019309878349304, + "learning_rate": 1.679651885619561e-06, + "loss": 0.4882, + "step": 8013 + }, + { + "epoch": 0.664235391628678, + "grad_norm": 0.2545405924320221, + "learning_rate": 1.6792374637380855e-06, + "loss": 0.5023, + "step": 8014 + }, + { + "epoch": 0.6643182760049731, + "grad_norm": 0.24277298152446747, + "learning_rate": 1.67882304185661e-06, + "loss": 0.5034, + "step": 8015 + }, + { + "epoch": 0.6644011603812682, + "grad_norm": 0.2744760811328888, + "learning_rate": 1.6784086199751347e-06, + "loss": 0.5402, + "step": 8016 + }, + { + "epoch": 0.6644840447575632, + "grad_norm": 0.23624424636363983, + "learning_rate": 1.6779941980936596e-06, + "loss": 0.481, + "step": 8017 + }, + { + "epoch": 0.6645669291338583, + "grad_norm": 0.2410036325454712, + "learning_rate": 1.6775797762121842e-06, + "loss": 0.4459, + "step": 8018 + }, + { + "epoch": 0.6646498135101533, + "grad_norm": 0.2528389096260071, + "learning_rate": 1.6771653543307087e-06, + "loss": 0.5119, + "step": 8019 + }, + { + "epoch": 0.6647326978864484, + "grad_norm": 0.24147643148899078, + "learning_rate": 1.6767509324492335e-06, + "loss": 0.4877, + "step": 8020 + }, + { + "epoch": 0.6648155822627435, + "grad_norm": 0.2550840675830841, + "learning_rate": 1.6763365105677582e-06, + "loss": 0.5645, + "step": 8021 + }, + { + "epoch": 0.6648984666390385, + "grad_norm": 0.25338536500930786, + "learning_rate": 1.6759220886862828e-06, + "loss": 0.49, + "step": 8022 + }, + { + "epoch": 0.6649813510153336, + "grad_norm": 0.2606537938117981, + "learning_rate": 1.6755076668048072e-06, + "loss": 0.5673, + "step": 8023 + }, + { + "epoch": 0.6650642353916287, + "grad_norm": 0.23945938050746918, + "learning_rate": 1.675093244923332e-06, + "loss": 0.4851, + "step": 8024 + }, + { + "epoch": 0.6651471197679237, + "grad_norm": 0.22834376990795135, + "learning_rate": 1.6746788230418567e-06, + "loss": 0.5092, + "step": 8025 + }, + { + "epoch": 0.6652300041442188, + "grad_norm": 0.24260655045509338, + "learning_rate": 1.6742644011603814e-06, + "loss": 0.4823, + "step": 8026 + }, + { + "epoch": 0.6653128885205138, + "grad_norm": 0.2592114806175232, + "learning_rate": 1.6738499792789062e-06, + "loss": 0.5364, + "step": 8027 + }, + { + "epoch": 0.665395772896809, + "grad_norm": 0.24672308564186096, + "learning_rate": 1.6734355573974307e-06, + "loss": 0.5031, + "step": 8028 + }, + { + "epoch": 0.6654786572731041, + "grad_norm": 0.2574688792228699, + "learning_rate": 1.6730211355159553e-06, + "loss": 0.5037, + "step": 8029 + }, + { + "epoch": 0.6655615416493991, + "grad_norm": 0.2368810772895813, + "learning_rate": 1.6726067136344802e-06, + "loss": 0.4716, + "step": 8030 + }, + { + "epoch": 0.6656444260256942, + "grad_norm": 0.24259810149669647, + "learning_rate": 1.6721922917530048e-06, + "loss": 0.5002, + "step": 8031 + }, + { + "epoch": 0.6657273104019892, + "grad_norm": 0.23703067004680634, + "learning_rate": 1.6717778698715293e-06, + "loss": 0.4849, + "step": 8032 + }, + { + "epoch": 0.6658101947782843, + "grad_norm": 0.23937910795211792, + "learning_rate": 1.6713634479900539e-06, + "loss": 0.4907, + "step": 8033 + }, + { + "epoch": 0.6658930791545794, + "grad_norm": 0.23541878163814545, + "learning_rate": 1.6709490261085787e-06, + "loss": 0.4402, + "step": 8034 + }, + { + "epoch": 0.6659759635308744, + "grad_norm": 0.25597843527793884, + "learning_rate": 1.6705346042271034e-06, + "loss": 0.5507, + "step": 8035 + }, + { + "epoch": 0.6660588479071695, + "grad_norm": 0.23979955911636353, + "learning_rate": 1.670120182345628e-06, + "loss": 0.4932, + "step": 8036 + }, + { + "epoch": 0.6661417322834645, + "grad_norm": 0.2749473452568054, + "learning_rate": 1.6697057604641525e-06, + "loss": 0.54, + "step": 8037 + }, + { + "epoch": 0.6662246166597596, + "grad_norm": 0.23410288989543915, + "learning_rate": 1.6692913385826773e-06, + "loss": 0.5137, + "step": 8038 + }, + { + "epoch": 0.6663075010360547, + "grad_norm": 0.2462262660264969, + "learning_rate": 1.668876916701202e-06, + "loss": 0.491, + "step": 8039 + }, + { + "epoch": 0.6663903854123497, + "grad_norm": 0.25500181317329407, + "learning_rate": 1.6684624948197268e-06, + "loss": 0.5001, + "step": 8040 + }, + { + "epoch": 0.6664732697886449, + "grad_norm": 0.2539308965206146, + "learning_rate": 1.6680480729382513e-06, + "loss": 0.5169, + "step": 8041 + }, + { + "epoch": 0.66655615416494, + "grad_norm": 0.25135841965675354, + "learning_rate": 1.667633651056776e-06, + "loss": 0.4969, + "step": 8042 + }, + { + "epoch": 0.666639038541235, + "grad_norm": 0.2491493672132492, + "learning_rate": 1.6672192291753005e-06, + "loss": 0.5276, + "step": 8043 + }, + { + "epoch": 0.6667219229175301, + "grad_norm": 0.2624061703681946, + "learning_rate": 1.6668048072938254e-06, + "loss": 0.4969, + "step": 8044 + }, + { + "checkpoint_runtime": 498.3332 + }, + { + "epoch": 0.6668048072938251, + "grad_norm": 0.23898115754127502, + "learning_rate": 1.66639038541235e-06, + "loss": 0.4938, + "step": 8045 + }, + { + "epoch": 0.6668876916701202, + "grad_norm": 0.24300970137119293, + "learning_rate": 1.6659759635308745e-06, + "loss": 0.4991, + "step": 8046 + }, + { + "epoch": 0.6669705760464153, + "grad_norm": 0.24400970339775085, + "learning_rate": 1.6655615416493991e-06, + "loss": 0.4876, + "step": 8047 + }, + { + "epoch": 0.6670534604227103, + "grad_norm": 0.2729347348213196, + "learning_rate": 1.665147119767924e-06, + "loss": 0.5302, + "step": 8048 + }, + { + "epoch": 0.6671363447990054, + "grad_norm": 0.2753441035747528, + "learning_rate": 1.6647326978864486e-06, + "loss": 0.5182, + "step": 8049 + }, + { + "epoch": 0.6672192291753004, + "grad_norm": 0.2521180212497711, + "learning_rate": 1.664318276004973e-06, + "loss": 0.5223, + "step": 8050 + }, + { + "epoch": 0.6673021135515955, + "grad_norm": 0.24737338721752167, + "learning_rate": 1.663903854123498e-06, + "loss": 0.538, + "step": 8051 + }, + { + "epoch": 0.6673849979278906, + "grad_norm": 0.2450684756040573, + "learning_rate": 1.6634894322420225e-06, + "loss": 0.5085, + "step": 8052 + }, + { + "epoch": 0.6674678823041856, + "grad_norm": 0.2316514253616333, + "learning_rate": 1.6630750103605472e-06, + "loss": 0.4945, + "step": 8053 + }, + { + "epoch": 0.6675507666804807, + "grad_norm": 0.24414819478988647, + "learning_rate": 1.662660588479072e-06, + "loss": 0.5145, + "step": 8054 + }, + { + "epoch": 0.6676336510567757, + "grad_norm": 0.23529665172100067, + "learning_rate": 1.6622461665975965e-06, + "loss": 0.4903, + "step": 8055 + }, + { + "epoch": 0.6677165354330709, + "grad_norm": 0.25986891984939575, + "learning_rate": 1.6618317447161211e-06, + "loss": 0.4777, + "step": 8056 + }, + { + "epoch": 0.667799419809366, + "grad_norm": 0.25294724106788635, + "learning_rate": 1.6614173228346458e-06, + "loss": 0.5248, + "step": 8057 + }, + { + "epoch": 0.667882304185661, + "grad_norm": 0.2444480061531067, + "learning_rate": 1.6610029009531706e-06, + "loss": 0.5071, + "step": 8058 + }, + { + "epoch": 0.6679651885619561, + "grad_norm": 0.2372560203075409, + "learning_rate": 1.660588479071695e-06, + "loss": 0.4823, + "step": 8059 + }, + { + "epoch": 0.6680480729382512, + "grad_norm": 0.23158448934555054, + "learning_rate": 1.6601740571902197e-06, + "loss": 0.4718, + "step": 8060 + }, + { + "epoch": 0.6681309573145462, + "grad_norm": 0.26710858941078186, + "learning_rate": 1.6597596353087445e-06, + "loss": 0.5237, + "step": 8061 + }, + { + "epoch": 0.6682138416908413, + "grad_norm": 0.2431776076555252, + "learning_rate": 1.6593452134272692e-06, + "loss": 0.4646, + "step": 8062 + }, + { + "epoch": 0.6682967260671363, + "grad_norm": 0.25357648730278015, + "learning_rate": 1.6589307915457936e-06, + "loss": 0.5344, + "step": 8063 + }, + { + "epoch": 0.6683796104434314, + "grad_norm": 0.24220089614391327, + "learning_rate": 1.6585163696643183e-06, + "loss": 0.4938, + "step": 8064 + }, + { + "epoch": 0.6684624948197265, + "grad_norm": 0.2464463710784912, + "learning_rate": 1.6581019477828431e-06, + "loss": 0.5486, + "step": 8065 + }, + { + "epoch": 0.6685453791960215, + "grad_norm": 0.231320321559906, + "learning_rate": 1.6576875259013678e-06, + "loss": 0.4599, + "step": 8066 + }, + { + "epoch": 0.6686282635723166, + "grad_norm": 0.2502411901950836, + "learning_rate": 1.6572731040198924e-06, + "loss": 0.5256, + "step": 8067 + }, + { + "epoch": 0.6687111479486116, + "grad_norm": 0.2654097378253937, + "learning_rate": 1.6568586821384168e-06, + "loss": 0.5062, + "step": 8068 + }, + { + "epoch": 0.6687940323249068, + "grad_norm": 0.2431391328573227, + "learning_rate": 1.6564442602569417e-06, + "loss": 0.4784, + "step": 8069 + }, + { + "epoch": 0.6688769167012019, + "grad_norm": 0.24967752397060394, + "learning_rate": 1.6560298383754663e-06, + "loss": 0.4733, + "step": 8070 + }, + { + "epoch": 0.6689598010774969, + "grad_norm": 0.24905797839164734, + "learning_rate": 1.6556154164939912e-06, + "loss": 0.4789, + "step": 8071 + }, + { + "epoch": 0.669042685453792, + "grad_norm": 0.24527227878570557, + "learning_rate": 1.6552009946125156e-06, + "loss": 0.4876, + "step": 8072 + }, + { + "epoch": 0.669125569830087, + "grad_norm": 0.26216837763786316, + "learning_rate": 1.6547865727310403e-06, + "loss": 0.5498, + "step": 8073 + }, + { + "epoch": 0.6692084542063821, + "grad_norm": 0.229136660695076, + "learning_rate": 1.654372150849565e-06, + "loss": 0.4869, + "step": 8074 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 0.24014896154403687, + "learning_rate": 1.6539577289680898e-06, + "loss": 0.4926, + "step": 8075 + }, + { + "epoch": 0.6693742229589722, + "grad_norm": 0.23893707990646362, + "learning_rate": 1.6535433070866144e-06, + "loss": 0.5002, + "step": 8076 + }, + { + "epoch": 0.6694571073352673, + "grad_norm": 0.23787716031074524, + "learning_rate": 1.6531288852051388e-06, + "loss": 0.478, + "step": 8077 + }, + { + "epoch": 0.6695399917115624, + "grad_norm": 0.23816396296024323, + "learning_rate": 1.6527144633236635e-06, + "loss": 0.491, + "step": 8078 + }, + { + "epoch": 0.6696228760878574, + "grad_norm": 0.24269425868988037, + "learning_rate": 1.6523000414421883e-06, + "loss": 0.4536, + "step": 8079 + }, + { + "epoch": 0.6697057604641525, + "grad_norm": 0.26355084776878357, + "learning_rate": 1.651885619560713e-06, + "loss": 0.5087, + "step": 8080 + }, + { + "epoch": 0.6697886448404475, + "grad_norm": 0.25338390469551086, + "learning_rate": 1.6514711976792374e-06, + "loss": 0.5024, + "step": 8081 + }, + { + "epoch": 0.6698715292167426, + "grad_norm": 0.24070964753627777, + "learning_rate": 1.6510567757977623e-06, + "loss": 0.4949, + "step": 8082 + }, + { + "epoch": 0.6699544135930378, + "grad_norm": 0.23326262831687927, + "learning_rate": 1.650642353916287e-06, + "loss": 0.4951, + "step": 8083 + }, + { + "epoch": 0.6700372979693328, + "grad_norm": 0.24067804217338562, + "learning_rate": 1.6502279320348116e-06, + "loss": 0.4962, + "step": 8084 + }, + { + "epoch": 0.6701201823456279, + "grad_norm": 0.26357409358024597, + "learning_rate": 1.6498135101533364e-06, + "loss": 0.563, + "step": 8085 + }, + { + "epoch": 0.6702030667219229, + "grad_norm": 0.23746053874492645, + "learning_rate": 1.6493990882718608e-06, + "loss": 0.4847, + "step": 8086 + }, + { + "epoch": 0.670285951098218, + "grad_norm": 0.29126161336898804, + "learning_rate": 1.6489846663903855e-06, + "loss": 0.5112, + "step": 8087 + }, + { + "epoch": 0.6703688354745131, + "grad_norm": 0.2650367021560669, + "learning_rate": 1.6485702445089101e-06, + "loss": 0.4841, + "step": 8088 + }, + { + "epoch": 0.6704517198508081, + "grad_norm": 0.2501189410686493, + "learning_rate": 1.648155822627435e-06, + "loss": 0.5192, + "step": 8089 + }, + { + "epoch": 0.6705346042271032, + "grad_norm": 0.2813457250595093, + "learning_rate": 1.6477414007459594e-06, + "loss": 0.5729, + "step": 8090 + }, + { + "epoch": 0.6706174886033982, + "grad_norm": 0.2555299699306488, + "learning_rate": 1.647326978864484e-06, + "loss": 0.512, + "step": 8091 + }, + { + "epoch": 0.6707003729796933, + "grad_norm": 0.24021750688552856, + "learning_rate": 1.646912556983009e-06, + "loss": 0.5349, + "step": 8092 + }, + { + "epoch": 0.6707832573559884, + "grad_norm": 0.25288254022598267, + "learning_rate": 1.6464981351015336e-06, + "loss": 0.5081, + "step": 8093 + }, + { + "epoch": 0.6708661417322834, + "grad_norm": 0.2408309429883957, + "learning_rate": 1.6460837132200582e-06, + "loss": 0.5116, + "step": 8094 + }, + { + "epoch": 0.6709490261085785, + "grad_norm": 0.24823586642742157, + "learning_rate": 1.6456692913385826e-06, + "loss": 0.5295, + "step": 8095 + }, + { + "epoch": 0.6710319104848737, + "grad_norm": 0.2545433044433594, + "learning_rate": 1.6452548694571075e-06, + "loss": 0.4856, + "step": 8096 + }, + { + "epoch": 0.6711147948611687, + "grad_norm": 0.25930115580558777, + "learning_rate": 1.6448404475756321e-06, + "loss": 0.4868, + "step": 8097 + }, + { + "epoch": 0.6711976792374638, + "grad_norm": 0.26367974281311035, + "learning_rate": 1.6444260256941568e-06, + "loss": 0.5077, + "step": 8098 + }, + { + "epoch": 0.6712805636137588, + "grad_norm": 0.26366400718688965, + "learning_rate": 1.6440116038126814e-06, + "loss": 0.5181, + "step": 8099 + }, + { + "epoch": 0.6713634479900539, + "grad_norm": 0.24165499210357666, + "learning_rate": 1.643597181931206e-06, + "loss": 0.5285, + "step": 8100 + }, + { + "epoch": 0.671446332366349, + "grad_norm": 0.23637782037258148, + "learning_rate": 1.6431827600497307e-06, + "loss": 0.4764, + "step": 8101 + }, + { + "epoch": 0.671529216742644, + "grad_norm": 0.22694194316864014, + "learning_rate": 1.6427683381682556e-06, + "loss": 0.4774, + "step": 8102 + }, + { + "epoch": 0.6716121011189391, + "grad_norm": 0.25829172134399414, + "learning_rate": 1.6423539162867802e-06, + "loss": 0.4942, + "step": 8103 + }, + { + "epoch": 0.6716949854952341, + "grad_norm": 0.2749623954296112, + "learning_rate": 1.6419394944053046e-06, + "loss": 0.4888, + "step": 8104 + }, + { + "epoch": 0.6717778698715292, + "grad_norm": 0.25396278500556946, + "learning_rate": 1.6415250725238293e-06, + "loss": 0.5279, + "step": 8105 + }, + { + "epoch": 0.6718607542478243, + "grad_norm": 0.24592967331409454, + "learning_rate": 1.6411106506423541e-06, + "loss": 0.4726, + "step": 8106 + }, + { + "epoch": 0.6719436386241193, + "grad_norm": 0.2534211277961731, + "learning_rate": 1.6406962287608788e-06, + "loss": 0.4968, + "step": 8107 + }, + { + "epoch": 0.6720265230004144, + "grad_norm": 0.26766830682754517, + "learning_rate": 1.6402818068794032e-06, + "loss": 0.5424, + "step": 8108 + }, + { + "epoch": 0.6721094073767095, + "grad_norm": 0.23090530931949615, + "learning_rate": 1.639867384997928e-06, + "loss": 0.4627, + "step": 8109 + }, + { + "epoch": 0.6721922917530045, + "grad_norm": 0.24694810807704926, + "learning_rate": 1.6394529631164527e-06, + "loss": 0.5173, + "step": 8110 + }, + { + "epoch": 0.6722751761292997, + "grad_norm": 0.27466094493865967, + "learning_rate": 1.6390385412349774e-06, + "loss": 0.4925, + "step": 8111 + }, + { + "epoch": 0.6723580605055947, + "grad_norm": 0.2493082731962204, + "learning_rate": 1.6386241193535022e-06, + "loss": 0.5179, + "step": 8112 + }, + { + "epoch": 0.6724409448818898, + "grad_norm": 0.241798534989357, + "learning_rate": 1.6382096974720266e-06, + "loss": 0.5412, + "step": 8113 + }, + { + "epoch": 0.6725238292581849, + "grad_norm": 0.26131489872932434, + "learning_rate": 1.6377952755905513e-06, + "loss": 0.5275, + "step": 8114 + }, + { + "epoch": 0.6726067136344799, + "grad_norm": 0.2607245147228241, + "learning_rate": 1.637380853709076e-06, + "loss": 0.5598, + "step": 8115 + }, + { + "epoch": 0.672689598010775, + "grad_norm": 0.2482404112815857, + "learning_rate": 1.6369664318276008e-06, + "loss": 0.5457, + "step": 8116 + }, + { + "epoch": 0.67277248238707, + "grad_norm": 0.24730733036994934, + "learning_rate": 1.6365520099461252e-06, + "loss": 0.5396, + "step": 8117 + }, + { + "epoch": 0.6728553667633651, + "grad_norm": 0.2372027188539505, + "learning_rate": 1.6361375880646499e-06, + "loss": 0.5122, + "step": 8118 + }, + { + "epoch": 0.6729382511396602, + "grad_norm": 0.2466852217912674, + "learning_rate": 1.6357231661831745e-06, + "loss": 0.4892, + "step": 8119 + }, + { + "epoch": 0.6730211355159552, + "grad_norm": 0.26800787448883057, + "learning_rate": 1.6353087443016994e-06, + "loss": 0.5096, + "step": 8120 + }, + { + "epoch": 0.6731040198922503, + "grad_norm": 0.25659558176994324, + "learning_rate": 1.634894322420224e-06, + "loss": 0.5327, + "step": 8121 + }, + { + "epoch": 0.6731869042685453, + "grad_norm": 0.25051403045654297, + "learning_rate": 1.6344799005387484e-06, + "loss": 0.5061, + "step": 8122 + }, + { + "epoch": 0.6732697886448404, + "grad_norm": 0.25170114636421204, + "learning_rate": 1.6340654786572733e-06, + "loss": 0.4905, + "step": 8123 + }, + { + "epoch": 0.6733526730211356, + "grad_norm": 0.27340638637542725, + "learning_rate": 1.633651056775798e-06, + "loss": 0.5226, + "step": 8124 + }, + { + "epoch": 0.6734355573974306, + "grad_norm": 0.24480745196342468, + "learning_rate": 1.6332366348943226e-06, + "loss": 0.483, + "step": 8125 + }, + { + "epoch": 0.6735184417737257, + "grad_norm": 0.2474764883518219, + "learning_rate": 1.632822213012847e-06, + "loss": 0.5126, + "step": 8126 + }, + { + "epoch": 0.6736013261500208, + "grad_norm": 0.24004848301410675, + "learning_rate": 1.6324077911313719e-06, + "loss": 0.5086, + "step": 8127 + }, + { + "epoch": 0.6736842105263158, + "grad_norm": 0.23006008565425873, + "learning_rate": 1.6319933692498965e-06, + "loss": 0.4841, + "step": 8128 + }, + { + "epoch": 0.6737670949026109, + "grad_norm": 0.24970050156116486, + "learning_rate": 1.6315789473684212e-06, + "loss": 0.4954, + "step": 8129 + }, + { + "epoch": 0.6738499792789059, + "grad_norm": 0.23457267880439758, + "learning_rate": 1.631164525486946e-06, + "loss": 0.4785, + "step": 8130 + }, + { + "epoch": 0.673932863655201, + "grad_norm": 0.25580883026123047, + "learning_rate": 1.6307501036054704e-06, + "loss": 0.5171, + "step": 8131 + }, + { + "epoch": 0.6740157480314961, + "grad_norm": 0.23761561512947083, + "learning_rate": 1.630335681723995e-06, + "loss": 0.4741, + "step": 8132 + }, + { + "epoch": 0.6740986324077911, + "grad_norm": 0.22715169191360474, + "learning_rate": 1.62992125984252e-06, + "loss": 0.4816, + "step": 8133 + }, + { + "epoch": 0.6741815167840862, + "grad_norm": 0.23192757368087769, + "learning_rate": 1.6295068379610446e-06, + "loss": 0.4823, + "step": 8134 + }, + { + "epoch": 0.6742644011603812, + "grad_norm": 0.29332292079925537, + "learning_rate": 1.629092416079569e-06, + "loss": 0.5099, + "step": 8135 + }, + { + "epoch": 0.6743472855366763, + "grad_norm": 0.2329431176185608, + "learning_rate": 1.6286779941980937e-06, + "loss": 0.4559, + "step": 8136 + }, + { + "epoch": 0.6744301699129714, + "grad_norm": 0.24443630874156952, + "learning_rate": 1.6282635723166185e-06, + "loss": 0.519, + "step": 8137 + }, + { + "epoch": 0.6745130542892664, + "grad_norm": 0.25947752594947815, + "learning_rate": 1.6278491504351432e-06, + "loss": 0.4767, + "step": 8138 + }, + { + "epoch": 0.6745959386655616, + "grad_norm": 0.2628006637096405, + "learning_rate": 1.6274347285536676e-06, + "loss": 0.5071, + "step": 8139 + }, + { + "epoch": 0.6746788230418566, + "grad_norm": 0.24653872847557068, + "learning_rate": 1.6270203066721924e-06, + "loss": 0.5442, + "step": 8140 + }, + { + "epoch": 0.6747617074181517, + "grad_norm": 0.2483346313238144, + "learning_rate": 1.626605884790717e-06, + "loss": 0.4893, + "step": 8141 + }, + { + "epoch": 0.6748445917944468, + "grad_norm": 0.23399725556373596, + "learning_rate": 1.6261914629092417e-06, + "loss": 0.4709, + "step": 8142 + }, + { + "epoch": 0.6749274761707418, + "grad_norm": 0.23497071862220764, + "learning_rate": 1.6257770410277666e-06, + "loss": 0.4883, + "step": 8143 + }, + { + "epoch": 0.6750103605470369, + "grad_norm": 0.2433125376701355, + "learning_rate": 1.625362619146291e-06, + "loss": 0.4726, + "step": 8144 + }, + { + "epoch": 0.675093244923332, + "grad_norm": 0.2664434313774109, + "learning_rate": 1.6249481972648157e-06, + "loss": 0.4974, + "step": 8145 + }, + { + "epoch": 0.675176129299627, + "grad_norm": 0.24235035479068756, + "learning_rate": 1.6245337753833403e-06, + "loss": 0.5236, + "step": 8146 + }, + { + "epoch": 0.6752590136759221, + "grad_norm": 0.25295913219451904, + "learning_rate": 1.6241193535018652e-06, + "loss": 0.5217, + "step": 8147 + }, + { + "epoch": 0.6753418980522171, + "grad_norm": 0.2550230920314789, + "learning_rate": 1.6237049316203896e-06, + "loss": 0.523, + "step": 8148 + }, + { + "epoch": 0.6754247824285122, + "grad_norm": 0.25955432653427124, + "learning_rate": 1.6232905097389142e-06, + "loss": 0.4853, + "step": 8149 + }, + { + "epoch": 0.6755076668048073, + "grad_norm": 0.23169420659542084, + "learning_rate": 1.622876087857439e-06, + "loss": 0.479, + "step": 8150 + }, + { + "epoch": 0.6755905511811023, + "grad_norm": 0.23271675407886505, + "learning_rate": 1.6224616659759637e-06, + "loss": 0.4916, + "step": 8151 + }, + { + "epoch": 0.6756734355573975, + "grad_norm": 0.2468981295824051, + "learning_rate": 1.6220472440944884e-06, + "loss": 0.4828, + "step": 8152 + }, + { + "epoch": 0.6757563199336925, + "grad_norm": 0.24258668720722198, + "learning_rate": 1.6216328222130128e-06, + "loss": 0.4924, + "step": 8153 + }, + { + "epoch": 0.6758392043099876, + "grad_norm": 0.23191259801387787, + "learning_rate": 1.6212184003315377e-06, + "loss": 0.4808, + "step": 8154 + }, + { + "epoch": 0.6759220886862827, + "grad_norm": 0.24354790151119232, + "learning_rate": 1.6208039784500623e-06, + "loss": 0.515, + "step": 8155 + }, + { + "epoch": 0.6760049730625777, + "grad_norm": 0.24169549345970154, + "learning_rate": 1.620389556568587e-06, + "loss": 0.5299, + "step": 8156 + }, + { + "epoch": 0.6760878574388728, + "grad_norm": 0.2411361038684845, + "learning_rate": 1.6199751346871114e-06, + "loss": 0.5036, + "step": 8157 + }, + { + "epoch": 0.6761707418151678, + "grad_norm": 0.2432798594236374, + "learning_rate": 1.6195607128056362e-06, + "loss": 0.5321, + "step": 8158 + }, + { + "epoch": 0.6762536261914629, + "grad_norm": 0.23382939398288727, + "learning_rate": 1.6191462909241609e-06, + "loss": 0.4694, + "step": 8159 + }, + { + "epoch": 0.676336510567758, + "grad_norm": 0.2606668770313263, + "learning_rate": 1.6187318690426857e-06, + "loss": 0.5288, + "step": 8160 + }, + { + "epoch": 0.676419394944053, + "grad_norm": 0.25492745637893677, + "learning_rate": 1.6183174471612104e-06, + "loss": 0.495, + "step": 8161 + }, + { + "epoch": 0.6765022793203481, + "grad_norm": 0.27056658267974854, + "learning_rate": 1.6179030252797348e-06, + "loss": 0.4804, + "step": 8162 + }, + { + "epoch": 0.6765851636966432, + "grad_norm": 0.2444959580898285, + "learning_rate": 1.6174886033982595e-06, + "loss": 0.5173, + "step": 8163 + }, + { + "epoch": 0.6766680480729382, + "grad_norm": 0.2745412588119507, + "learning_rate": 1.6170741815167843e-06, + "loss": 0.5158, + "step": 8164 + }, + { + "epoch": 0.6767509324492333, + "grad_norm": 0.24490121006965637, + "learning_rate": 1.616659759635309e-06, + "loss": 0.4961, + "step": 8165 + }, + { + "epoch": 0.6768338168255283, + "grad_norm": 0.26058530807495117, + "learning_rate": 1.6162453377538334e-06, + "loss": 0.5314, + "step": 8166 + }, + { + "epoch": 0.6769167012018235, + "grad_norm": 0.23753610253334045, + "learning_rate": 1.615830915872358e-06, + "loss": 0.4848, + "step": 8167 + }, + { + "epoch": 0.6769995855781186, + "grad_norm": 0.257586270570755, + "learning_rate": 1.6154164939908829e-06, + "loss": 0.4904, + "step": 8168 + }, + { + "epoch": 0.6770824699544136, + "grad_norm": 0.24642257392406464, + "learning_rate": 1.6150020721094075e-06, + "loss": 0.5182, + "step": 8169 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 0.24668122828006744, + "learning_rate": 1.6145876502279324e-06, + "loss": 0.4701, + "step": 8170 + }, + { + "epoch": 0.6772482387070037, + "grad_norm": 0.23751994967460632, + "learning_rate": 1.6141732283464568e-06, + "loss": 0.5119, + "step": 8171 + }, + { + "epoch": 0.6773311230832988, + "grad_norm": 0.23951254785060883, + "learning_rate": 1.6137588064649815e-06, + "loss": 0.4734, + "step": 8172 + }, + { + "epoch": 0.6774140074595939, + "grad_norm": 0.24581864476203918, + "learning_rate": 1.6133443845835061e-06, + "loss": 0.5231, + "step": 8173 + }, + { + "epoch": 0.6774968918358889, + "grad_norm": 0.2344207465648651, + "learning_rate": 1.612929962702031e-06, + "loss": 0.4648, + "step": 8174 + }, + { + "epoch": 0.677579776212184, + "grad_norm": 0.24873098731040955, + "learning_rate": 1.6125155408205554e-06, + "loss": 0.5359, + "step": 8175 + }, + { + "epoch": 0.6776626605884791, + "grad_norm": 0.2565556764602661, + "learning_rate": 1.61210111893908e-06, + "loss": 0.4905, + "step": 8176 + }, + { + "epoch": 0.6777455449647741, + "grad_norm": 0.25140464305877686, + "learning_rate": 1.6116866970576047e-06, + "loss": 0.5062, + "step": 8177 + }, + { + "epoch": 0.6778284293410692, + "grad_norm": 0.24127836525440216, + "learning_rate": 1.6112722751761295e-06, + "loss": 0.5328, + "step": 8178 + }, + { + "epoch": 0.6779113137173642, + "grad_norm": 0.23488683998584747, + "learning_rate": 1.6108578532946542e-06, + "loss": 0.4803, + "step": 8179 + }, + { + "epoch": 0.6779941980936594, + "grad_norm": 0.22608764469623566, + "learning_rate": 1.6104434314131786e-06, + "loss": 0.4689, + "step": 8180 + }, + { + "epoch": 0.6780770824699545, + "grad_norm": 0.2348918318748474, + "learning_rate": 1.6100290095317035e-06, + "loss": 0.4797, + "step": 8181 + }, + { + "epoch": 0.6781599668462495, + "grad_norm": 0.24670246243476868, + "learning_rate": 1.6096145876502281e-06, + "loss": 0.545, + "step": 8182 + }, + { + "epoch": 0.6782428512225446, + "grad_norm": 0.2377563714981079, + "learning_rate": 1.6092001657687528e-06, + "loss": 0.4949, + "step": 8183 + }, + { + "epoch": 0.6783257355988396, + "grad_norm": 0.24946482479572296, + "learning_rate": 1.6087857438872772e-06, + "loss": 0.5065, + "step": 8184 + }, + { + "epoch": 0.6784086199751347, + "grad_norm": 0.24333594739437103, + "learning_rate": 1.608371322005802e-06, + "loss": 0.4759, + "step": 8185 + }, + { + "epoch": 0.6784915043514298, + "grad_norm": 0.2662675678730011, + "learning_rate": 1.6079569001243267e-06, + "loss": 0.5041, + "step": 8186 + }, + { + "epoch": 0.6785743887277248, + "grad_norm": 0.23173965513706207, + "learning_rate": 1.6075424782428513e-06, + "loss": 0.4739, + "step": 8187 + }, + { + "epoch": 0.6786572731040199, + "grad_norm": 0.23867300152778625, + "learning_rate": 1.6071280563613762e-06, + "loss": 0.4896, + "step": 8188 + }, + { + "epoch": 0.6787401574803149, + "grad_norm": 0.2374475598335266, + "learning_rate": 1.6067136344799006e-06, + "loss": 0.5184, + "step": 8189 + }, + { + "epoch": 0.67882304185661, + "grad_norm": 0.23655076324939728, + "learning_rate": 1.6062992125984253e-06, + "loss": 0.463, + "step": 8190 + }, + { + "epoch": 0.6789059262329051, + "grad_norm": 0.2514820694923401, + "learning_rate": 1.6058847907169501e-06, + "loss": 0.5106, + "step": 8191 + }, + { + "epoch": 0.6789888106092001, + "grad_norm": 0.27555331587791443, + "learning_rate": 1.6054703688354748e-06, + "loss": 0.5064, + "step": 8192 + }, + { + "epoch": 0.6790716949854952, + "grad_norm": 0.23592247068881989, + "learning_rate": 1.6050559469539992e-06, + "loss": 0.4893, + "step": 8193 + }, + { + "epoch": 0.6791545793617904, + "grad_norm": 0.2611584961414337, + "learning_rate": 1.6046415250725238e-06, + "loss": 0.4951, + "step": 8194 + }, + { + "epoch": 0.6792374637380854, + "grad_norm": 0.26804620027542114, + "learning_rate": 1.6042271031910487e-06, + "loss": 0.5714, + "step": 8195 + }, + { + "epoch": 0.6793203481143805, + "grad_norm": 0.23535782098770142, + "learning_rate": 1.6038126813095733e-06, + "loss": 0.489, + "step": 8196 + }, + { + "epoch": 0.6794032324906755, + "grad_norm": 0.23114807903766632, + "learning_rate": 1.603398259428098e-06, + "loss": 0.4626, + "step": 8197 + }, + { + "epoch": 0.6794861168669706, + "grad_norm": 0.2363682985305786, + "learning_rate": 1.6029838375466224e-06, + "loss": 0.4958, + "step": 8198 + }, + { + "epoch": 0.6795690012432657, + "grad_norm": 0.28053683042526245, + "learning_rate": 1.6025694156651473e-06, + "loss": 0.4559, + "step": 8199 + }, + { + "epoch": 0.6796518856195607, + "grad_norm": 0.2508561611175537, + "learning_rate": 1.602154993783672e-06, + "loss": 0.5364, + "step": 8200 + }, + { + "epoch": 0.6797347699958558, + "grad_norm": 0.26386022567749023, + "learning_rate": 1.6017405719021968e-06, + "loss": 0.5225, + "step": 8201 + }, + { + "epoch": 0.6798176543721508, + "grad_norm": 0.24508553743362427, + "learning_rate": 1.6013261500207212e-06, + "loss": 0.5062, + "step": 8202 + }, + { + "epoch": 0.6799005387484459, + "grad_norm": 0.257393479347229, + "learning_rate": 1.6009117281392458e-06, + "loss": 0.5272, + "step": 8203 + }, + { + "epoch": 0.679983423124741, + "grad_norm": 0.24898111820220947, + "learning_rate": 1.6004973062577705e-06, + "loss": 0.5118, + "step": 8204 + }, + { + "epoch": 0.680066307501036, + "grad_norm": 0.23558610677719116, + "learning_rate": 1.6000828843762953e-06, + "loss": 0.4894, + "step": 8205 + }, + { + "epoch": 0.6801491918773311, + "grad_norm": 0.23347395658493042, + "learning_rate": 1.59966846249482e-06, + "loss": 0.478, + "step": 8206 + }, + { + "epoch": 0.6802320762536261, + "grad_norm": 0.25028321146965027, + "learning_rate": 1.5992540406133444e-06, + "loss": 0.4926, + "step": 8207 + }, + { + "epoch": 0.6803149606299213, + "grad_norm": 0.2605452537536621, + "learning_rate": 1.598839618731869e-06, + "loss": 0.5204, + "step": 8208 + }, + { + "epoch": 0.6803978450062164, + "grad_norm": 0.2414514571428299, + "learning_rate": 1.598425196850394e-06, + "loss": 0.4898, + "step": 8209 + }, + { + "epoch": 0.6804807293825114, + "grad_norm": 0.23547787964344025, + "learning_rate": 1.5980107749689186e-06, + "loss": 0.4676, + "step": 8210 + }, + { + "epoch": 0.6805636137588065, + "grad_norm": 0.23882907629013062, + "learning_rate": 1.597596353087443e-06, + "loss": 0.4876, + "step": 8211 + }, + { + "epoch": 0.6806464981351016, + "grad_norm": 0.24600446224212646, + "learning_rate": 1.5971819312059678e-06, + "loss": 0.4838, + "step": 8212 + }, + { + "epoch": 0.6807293825113966, + "grad_norm": 0.24378730356693268, + "learning_rate": 1.5967675093244925e-06, + "loss": 0.5158, + "step": 8213 + }, + { + "epoch": 0.6808122668876917, + "grad_norm": 0.24381886422634125, + "learning_rate": 1.5963530874430171e-06, + "loss": 0.5212, + "step": 8214 + }, + { + "epoch": 0.6808951512639867, + "grad_norm": 0.26165011525154114, + "learning_rate": 1.5959386655615416e-06, + "loss": 0.5282, + "step": 8215 + }, + { + "epoch": 0.6809780356402818, + "grad_norm": 0.23891794681549072, + "learning_rate": 1.5955242436800664e-06, + "loss": 0.4983, + "step": 8216 + }, + { + "epoch": 0.6810609200165769, + "grad_norm": 0.2565023899078369, + "learning_rate": 1.595109821798591e-06, + "loss": 0.522, + "step": 8217 + }, + { + "epoch": 0.6811438043928719, + "grad_norm": 0.2561299204826355, + "learning_rate": 1.5946953999171157e-06, + "loss": 0.5522, + "step": 8218 + }, + { + "epoch": 0.681226688769167, + "grad_norm": 0.24498051404953003, + "learning_rate": 1.5942809780356406e-06, + "loss": 0.5261, + "step": 8219 + }, + { + "epoch": 0.681309573145462, + "grad_norm": 0.2762969434261322, + "learning_rate": 1.593866556154165e-06, + "loss": 0.5388, + "step": 8220 + }, + { + "epoch": 0.6813924575217571, + "grad_norm": 0.24863865971565247, + "learning_rate": 1.5934521342726896e-06, + "loss": 0.531, + "step": 8221 + }, + { + "epoch": 0.6814753418980523, + "grad_norm": 0.23971965909004211, + "learning_rate": 1.5930377123912145e-06, + "loss": 0.5116, + "step": 8222 + }, + { + "epoch": 0.6815582262743473, + "grad_norm": 0.23554173111915588, + "learning_rate": 1.5926232905097391e-06, + "loss": 0.5004, + "step": 8223 + }, + { + "epoch": 0.6816411106506424, + "grad_norm": 0.23859864473342896, + "learning_rate": 1.5922088686282636e-06, + "loss": 0.5036, + "step": 8224 + }, + { + "epoch": 0.6817239950269374, + "grad_norm": 0.23519380390644073, + "learning_rate": 1.5917944467467882e-06, + "loss": 0.5021, + "step": 8225 + }, + { + "epoch": 0.6818068794032325, + "grad_norm": 0.24629110097885132, + "learning_rate": 1.591380024865313e-06, + "loss": 0.4982, + "step": 8226 + }, + { + "epoch": 0.6818897637795276, + "grad_norm": 0.23304888606071472, + "learning_rate": 1.5909656029838377e-06, + "loss": 0.5162, + "step": 8227 + }, + { + "epoch": 0.6819726481558226, + "grad_norm": 0.22150743007659912, + "learning_rate": 1.5905511811023624e-06, + "loss": 0.4593, + "step": 8228 + }, + { + "epoch": 0.6820555325321177, + "grad_norm": 0.24611206352710724, + "learning_rate": 1.590136759220887e-06, + "loss": 0.5203, + "step": 8229 + }, + { + "epoch": 0.6821384169084128, + "grad_norm": 0.25631916522979736, + "learning_rate": 1.5897223373394116e-06, + "loss": 0.5156, + "step": 8230 + }, + { + "epoch": 0.6822213012847078, + "grad_norm": 0.22713398933410645, + "learning_rate": 1.5893079154579363e-06, + "loss": 0.4851, + "step": 8231 + }, + { + "epoch": 0.6823041856610029, + "grad_norm": 0.24740014970302582, + "learning_rate": 1.5888934935764611e-06, + "loss": 0.5659, + "step": 8232 + }, + { + "epoch": 0.6823870700372979, + "grad_norm": 0.2452453374862671, + "learning_rate": 1.5884790716949856e-06, + "loss": 0.5157, + "step": 8233 + }, + { + "epoch": 0.682469954413593, + "grad_norm": 0.24798263609409332, + "learning_rate": 1.5880646498135102e-06, + "loss": 0.4929, + "step": 8234 + }, + { + "epoch": 0.6825528387898881, + "grad_norm": 0.24031569063663483, + "learning_rate": 1.5876502279320349e-06, + "loss": 0.5265, + "step": 8235 + }, + { + "epoch": 0.6826357231661832, + "grad_norm": 0.23768775165081024, + "learning_rate": 1.5872358060505597e-06, + "loss": 0.4642, + "step": 8236 + }, + { + "epoch": 0.6827186075424783, + "grad_norm": 0.25767606496810913, + "learning_rate": 1.5868213841690844e-06, + "loss": 0.4775, + "step": 8237 + }, + { + "epoch": 0.6828014919187733, + "grad_norm": 0.2463589459657669, + "learning_rate": 1.5864069622876088e-06, + "loss": 0.485, + "step": 8238 + }, + { + "epoch": 0.6828843762950684, + "grad_norm": 0.24850201606750488, + "learning_rate": 1.5859925404061336e-06, + "loss": 0.5368, + "step": 8239 + }, + { + "epoch": 0.6829672606713635, + "grad_norm": 0.23130208253860474, + "learning_rate": 1.5855781185246583e-06, + "loss": 0.5046, + "step": 8240 + }, + { + "epoch": 0.6830501450476585, + "grad_norm": 0.2690594792366028, + "learning_rate": 1.585163696643183e-06, + "loss": 0.5583, + "step": 8241 + }, + { + "epoch": 0.6831330294239536, + "grad_norm": 0.22900034487247467, + "learning_rate": 1.5847492747617074e-06, + "loss": 0.4883, + "step": 8242 + }, + { + "epoch": 0.6832159138002486, + "grad_norm": 0.24432432651519775, + "learning_rate": 1.5843348528802322e-06, + "loss": 0.5062, + "step": 8243 + }, + { + "epoch": 0.6832987981765437, + "grad_norm": 0.24029165506362915, + "learning_rate": 1.5839204309987569e-06, + "loss": 0.4633, + "step": 8244 + }, + { + "epoch": 0.6833816825528388, + "grad_norm": 0.24637408554553986, + "learning_rate": 1.5835060091172815e-06, + "loss": 0.506, + "step": 8245 + }, + { + "epoch": 0.6834645669291338, + "grad_norm": 0.2641081213951111, + "learning_rate": 1.5830915872358064e-06, + "loss": 0.5054, + "step": 8246 + }, + { + "epoch": 0.6835474513054289, + "grad_norm": 0.25474268198013306, + "learning_rate": 1.5826771653543308e-06, + "loss": 0.5015, + "step": 8247 + }, + { + "epoch": 0.683630335681724, + "grad_norm": 0.2395784556865692, + "learning_rate": 1.5822627434728554e-06, + "loss": 0.4951, + "step": 8248 + }, + { + "epoch": 0.683713220058019, + "grad_norm": 0.24772636592388153, + "learning_rate": 1.5818483215913803e-06, + "loss": 0.5106, + "step": 8249 + }, + { + "epoch": 0.6837961044343142, + "grad_norm": 0.255646675825119, + "learning_rate": 1.581433899709905e-06, + "loss": 0.5276, + "step": 8250 + }, + { + "epoch": 0.6838789888106092, + "grad_norm": 0.23127180337905884, + "learning_rate": 1.5810194778284294e-06, + "loss": 0.4321, + "step": 8251 + }, + { + "epoch": 0.6839618731869043, + "grad_norm": 0.23446670174598694, + "learning_rate": 1.580605055946954e-06, + "loss": 0.4702, + "step": 8252 + }, + { + "epoch": 0.6840447575631994, + "grad_norm": 0.2317572385072708, + "learning_rate": 1.5801906340654789e-06, + "loss": 0.4561, + "step": 8253 + }, + { + "epoch": 0.6841276419394944, + "grad_norm": 0.2641199231147766, + "learning_rate": 1.5797762121840035e-06, + "loss": 0.5713, + "step": 8254 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 0.23611246049404144, + "learning_rate": 1.5793617903025282e-06, + "loss": 0.4891, + "step": 8255 + }, + { + "epoch": 0.6842934106920845, + "grad_norm": 0.2534436583518982, + "learning_rate": 1.5789473684210526e-06, + "loss": 0.5303, + "step": 8256 + }, + { + "epoch": 0.6843762950683796, + "grad_norm": 0.22692489624023438, + "learning_rate": 1.5785329465395774e-06, + "loss": 0.48, + "step": 8257 + }, + { + "epoch": 0.6844591794446747, + "grad_norm": 0.2242012768983841, + "learning_rate": 1.578118524658102e-06, + "loss": 0.4717, + "step": 8258 + }, + { + "epoch": 0.6845420638209697, + "grad_norm": 0.2291249781847, + "learning_rate": 1.577704102776627e-06, + "loss": 0.5118, + "step": 8259 + }, + { + "epoch": 0.6846249481972648, + "grad_norm": 0.2510213851928711, + "learning_rate": 1.5772896808951514e-06, + "loss": 0.541, + "step": 8260 + }, + { + "epoch": 0.6847078325735599, + "grad_norm": 0.24976646900177002, + "learning_rate": 1.576875259013676e-06, + "loss": 0.508, + "step": 8261 + }, + { + "epoch": 0.6847907169498549, + "grad_norm": 0.235773503780365, + "learning_rate": 1.5764608371322007e-06, + "loss": 0.4849, + "step": 8262 + }, + { + "epoch": 0.68487360132615, + "grad_norm": 0.24863041937351227, + "learning_rate": 1.5760464152507255e-06, + "loss": 0.4879, + "step": 8263 + }, + { + "epoch": 0.684956485702445, + "grad_norm": 0.2358422428369522, + "learning_rate": 1.5756319933692502e-06, + "loss": 0.5019, + "step": 8264 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 0.2552856504917145, + "learning_rate": 1.5752175714877746e-06, + "loss": 0.5219, + "step": 8265 + }, + { + "epoch": 0.6851222544550353, + "grad_norm": 0.24859578907489777, + "learning_rate": 1.5748031496062992e-06, + "loss": 0.5257, + "step": 8266 + }, + { + "epoch": 0.6852051388313303, + "grad_norm": 0.23313263058662415, + "learning_rate": 1.574388727724824e-06, + "loss": 0.4614, + "step": 8267 + }, + { + "epoch": 0.6852880232076254, + "grad_norm": 0.24220742285251617, + "learning_rate": 1.5739743058433487e-06, + "loss": 0.4729, + "step": 8268 + }, + { + "epoch": 0.6853709075839204, + "grad_norm": 0.2620247006416321, + "learning_rate": 1.5735598839618732e-06, + "loss": 0.5172, + "step": 8269 + }, + { + "epoch": 0.6854537919602155, + "grad_norm": 0.24700692296028137, + "learning_rate": 1.573145462080398e-06, + "loss": 0.473, + "step": 8270 + }, + { + "epoch": 0.6855366763365106, + "grad_norm": 0.26505357027053833, + "learning_rate": 1.5727310401989227e-06, + "loss": 0.5027, + "step": 8271 + }, + { + "epoch": 0.6856195607128056, + "grad_norm": 0.2595752775669098, + "learning_rate": 1.5723166183174473e-06, + "loss": 0.5431, + "step": 8272 + }, + { + "epoch": 0.6857024450891007, + "grad_norm": 0.2665795683860779, + "learning_rate": 1.5719021964359722e-06, + "loss": 0.5378, + "step": 8273 + }, + { + "epoch": 0.6857853294653957, + "grad_norm": 0.2418244481086731, + "learning_rate": 1.5714877745544966e-06, + "loss": 0.4849, + "step": 8274 + }, + { + "epoch": 0.6858682138416908, + "grad_norm": 0.23648960888385773, + "learning_rate": 1.5710733526730212e-06, + "loss": 0.4839, + "step": 8275 + }, + { + "epoch": 0.6859510982179859, + "grad_norm": 0.2595992386341095, + "learning_rate": 1.5706589307915459e-06, + "loss": 0.5266, + "step": 8276 + }, + { + "epoch": 0.6860339825942809, + "grad_norm": 0.23139338195323944, + "learning_rate": 1.5702445089100707e-06, + "loss": 0.4579, + "step": 8277 + }, + { + "epoch": 0.6861168669705761, + "grad_norm": 0.2652495503425598, + "learning_rate": 1.5698300870285952e-06, + "loss": 0.5466, + "step": 8278 + }, + { + "epoch": 0.6861997513468712, + "grad_norm": 0.2683718800544739, + "learning_rate": 1.5694156651471198e-06, + "loss": 0.5281, + "step": 8279 + }, + { + "epoch": 0.6862826357231662, + "grad_norm": 0.27592700719833374, + "learning_rate": 1.5690012432656447e-06, + "loss": 0.5552, + "step": 8280 + }, + { + "epoch": 0.6863655200994613, + "grad_norm": 0.24752801656723022, + "learning_rate": 1.5685868213841693e-06, + "loss": 0.525, + "step": 8281 + }, + { + "epoch": 0.6864484044757563, + "grad_norm": 0.23892733454704285, + "learning_rate": 1.5681723995026937e-06, + "loss": 0.4595, + "step": 8282 + }, + { + "epoch": 0.6865312888520514, + "grad_norm": 0.22760513424873352, + "learning_rate": 1.5677579776212184e-06, + "loss": 0.4729, + "step": 8283 + }, + { + "epoch": 0.6866141732283465, + "grad_norm": 0.24936127662658691, + "learning_rate": 1.5673435557397432e-06, + "loss": 0.4967, + "step": 8284 + }, + { + "epoch": 0.6866970576046415, + "grad_norm": 0.23002296686172485, + "learning_rate": 1.5669291338582679e-06, + "loss": 0.4608, + "step": 8285 + }, + { + "epoch": 0.6867799419809366, + "grad_norm": 0.2406623661518097, + "learning_rate": 1.5665147119767925e-06, + "loss": 0.4755, + "step": 8286 + }, + { + "epoch": 0.6868628263572316, + "grad_norm": 0.23925575613975525, + "learning_rate": 1.566100290095317e-06, + "loss": 0.484, + "step": 8287 + }, + { + "epoch": 0.6869457107335267, + "grad_norm": 0.2474452257156372, + "learning_rate": 1.5656858682138418e-06, + "loss": 0.495, + "step": 8288 + }, + { + "epoch": 0.6870285951098218, + "grad_norm": 0.2546044588088989, + "learning_rate": 1.5652714463323665e-06, + "loss": 0.4936, + "step": 8289 + }, + { + "epoch": 0.6871114794861168, + "grad_norm": 0.2527429461479187, + "learning_rate": 1.5648570244508913e-06, + "loss": 0.488, + "step": 8290 + }, + { + "epoch": 0.687194363862412, + "grad_norm": 0.26367855072021484, + "learning_rate": 1.5644426025694157e-06, + "loss": 0.5028, + "step": 8291 + }, + { + "epoch": 0.687277248238707, + "grad_norm": 0.35563740134239197, + "learning_rate": 1.5640281806879404e-06, + "loss": 0.4842, + "step": 8292 + }, + { + "epoch": 0.6873601326150021, + "grad_norm": 0.2742883861064911, + "learning_rate": 1.563613758806465e-06, + "loss": 0.5256, + "step": 8293 + }, + { + "epoch": 0.6874430169912972, + "grad_norm": 0.24858330190181732, + "learning_rate": 1.5631993369249899e-06, + "loss": 0.5214, + "step": 8294 + }, + { + "epoch": 0.6875259013675922, + "grad_norm": 0.2404656857252121, + "learning_rate": 1.5627849150435145e-06, + "loss": 0.4858, + "step": 8295 + }, + { + "epoch": 0.6876087857438873, + "grad_norm": 0.2575848400592804, + "learning_rate": 1.562370493162039e-06, + "loss": 0.5313, + "step": 8296 + }, + { + "epoch": 0.6876916701201824, + "grad_norm": 0.2347864806652069, + "learning_rate": 1.5619560712805636e-06, + "loss": 0.5147, + "step": 8297 + }, + { + "epoch": 0.6877745544964774, + "grad_norm": 0.2600526213645935, + "learning_rate": 1.5615416493990885e-06, + "loss": 0.4442, + "step": 8298 + }, + { + "epoch": 0.6878574388727725, + "grad_norm": 0.24436889588832855, + "learning_rate": 1.561127227517613e-06, + "loss": 0.4992, + "step": 8299 + }, + { + "epoch": 0.6879403232490675, + "grad_norm": 0.24623075127601624, + "learning_rate": 1.5607128056361375e-06, + "loss": 0.5011, + "step": 8300 + }, + { + "epoch": 0.6880232076253626, + "grad_norm": 0.24979569017887115, + "learning_rate": 1.5602983837546624e-06, + "loss": 0.537, + "step": 8301 + }, + { + "epoch": 0.6881060920016577, + "grad_norm": 0.2347949743270874, + "learning_rate": 1.559883961873187e-06, + "loss": 0.4654, + "step": 8302 + }, + { + "epoch": 0.6881889763779527, + "grad_norm": 0.26734161376953125, + "learning_rate": 1.5594695399917117e-06, + "loss": 0.5457, + "step": 8303 + }, + { + "epoch": 0.6882718607542478, + "grad_norm": 0.23590627312660217, + "learning_rate": 1.5590551181102365e-06, + "loss": 0.4659, + "step": 8304 + }, + { + "epoch": 0.6883547451305428, + "grad_norm": 0.22420179843902588, + "learning_rate": 1.558640696228761e-06, + "loss": 0.4909, + "step": 8305 + }, + { + "epoch": 0.688437629506838, + "grad_norm": 0.25914478302001953, + "learning_rate": 1.5582262743472856e-06, + "loss": 0.4817, + "step": 8306 + }, + { + "epoch": 0.6885205138831331, + "grad_norm": 0.2692803144454956, + "learning_rate": 1.5578118524658103e-06, + "loss": 0.5738, + "step": 8307 + }, + { + "epoch": 0.6886033982594281, + "grad_norm": 0.23552151024341583, + "learning_rate": 1.5573974305843351e-06, + "loss": 0.4818, + "step": 8308 + }, + { + "epoch": 0.6886862826357232, + "grad_norm": 0.25755083560943604, + "learning_rate": 1.5569830087028595e-06, + "loss": 0.5068, + "step": 8309 + }, + { + "epoch": 0.6887691670120182, + "grad_norm": 0.2517620325088501, + "learning_rate": 1.5565685868213842e-06, + "loss": 0.5183, + "step": 8310 + }, + { + "epoch": 0.6888520513883133, + "grad_norm": 0.2316666692495346, + "learning_rate": 1.556154164939909e-06, + "loss": 0.4526, + "step": 8311 + }, + { + "epoch": 0.6889349357646084, + "grad_norm": 0.25332018733024597, + "learning_rate": 1.5557397430584337e-06, + "loss": 0.4882, + "step": 8312 + }, + { + "epoch": 0.6890178201409034, + "grad_norm": 0.27144747972488403, + "learning_rate": 1.5553253211769583e-06, + "loss": 0.52, + "step": 8313 + }, + { + "epoch": 0.6891007045171985, + "grad_norm": 0.3167646527290344, + "learning_rate": 1.5549108992954828e-06, + "loss": 0.5071, + "step": 8314 + }, + { + "epoch": 0.6891835888934936, + "grad_norm": 0.23122791945934296, + "learning_rate": 1.5544964774140076e-06, + "loss": 0.5048, + "step": 8315 + }, + { + "epoch": 0.6892664732697886, + "grad_norm": 0.23155005276203156, + "learning_rate": 1.5540820555325323e-06, + "loss": 0.4643, + "step": 8316 + }, + { + "epoch": 0.6893493576460837, + "grad_norm": 0.24817681312561035, + "learning_rate": 1.553667633651057e-06, + "loss": 0.4842, + "step": 8317 + }, + { + "epoch": 0.6894322420223787, + "grad_norm": 0.22873055934906006, + "learning_rate": 1.5532532117695815e-06, + "loss": 0.4586, + "step": 8318 + }, + { + "epoch": 0.6895151263986738, + "grad_norm": 0.26574206352233887, + "learning_rate": 1.5528387898881062e-06, + "loss": 0.5016, + "step": 8319 + }, + { + "epoch": 0.689598010774969, + "grad_norm": 0.242578387260437, + "learning_rate": 1.5524243680066308e-06, + "loss": 0.4901, + "step": 8320 + }, + { + "epoch": 0.689680895151264, + "grad_norm": 0.27214154601097107, + "learning_rate": 1.5520099461251557e-06, + "loss": 0.5273, + "step": 8321 + }, + { + "epoch": 0.6897637795275591, + "grad_norm": 0.25154992938041687, + "learning_rate": 1.5515955242436803e-06, + "loss": 0.5057, + "step": 8322 + }, + { + "epoch": 0.6898466639038541, + "grad_norm": 0.24354486167430878, + "learning_rate": 1.5511811023622048e-06, + "loss": 0.466, + "step": 8323 + }, + { + "epoch": 0.6899295482801492, + "grad_norm": 0.23920214176177979, + "learning_rate": 1.5507666804807294e-06, + "loss": 0.5185, + "step": 8324 + }, + { + "epoch": 0.6900124326564443, + "grad_norm": 0.24655498564243317, + "learning_rate": 1.5503522585992543e-06, + "loss": 0.5283, + "step": 8325 + }, + { + "epoch": 0.6900953170327393, + "grad_norm": 0.2417418360710144, + "learning_rate": 1.549937836717779e-06, + "loss": 0.5055, + "step": 8326 + }, + { + "epoch": 0.6901782014090344, + "grad_norm": 0.2511945962905884, + "learning_rate": 1.5495234148363033e-06, + "loss": 0.5372, + "step": 8327 + }, + { + "epoch": 0.6902610857853295, + "grad_norm": 0.2372310310602188, + "learning_rate": 1.5491089929548282e-06, + "loss": 0.4777, + "step": 8328 + }, + { + "epoch": 0.6903439701616245, + "grad_norm": 0.2709999680519104, + "learning_rate": 1.5486945710733528e-06, + "loss": 0.5188, + "step": 8329 + }, + { + "epoch": 0.6904268545379196, + "grad_norm": 0.24308538436889648, + "learning_rate": 1.5482801491918775e-06, + "loss": 0.4411, + "step": 8330 + }, + { + "epoch": 0.6905097389142146, + "grad_norm": 0.24693584442138672, + "learning_rate": 1.5478657273104023e-06, + "loss": 0.4794, + "step": 8331 + }, + { + "epoch": 0.6905926232905097, + "grad_norm": 0.24414710700511932, + "learning_rate": 1.5474513054289268e-06, + "loss": 0.4891, + "step": 8332 + }, + { + "epoch": 0.6906755076668049, + "grad_norm": 0.2394716739654541, + "learning_rate": 1.5470368835474514e-06, + "loss": 0.4658, + "step": 8333 + }, + { + "epoch": 0.6907583920430999, + "grad_norm": 0.24152973294258118, + "learning_rate": 1.546622461665976e-06, + "loss": 0.4582, + "step": 8334 + }, + { + "epoch": 0.690841276419395, + "grad_norm": 0.2746962010860443, + "learning_rate": 1.546208039784501e-06, + "loss": 0.5663, + "step": 8335 + }, + { + "epoch": 0.69092416079569, + "grad_norm": 0.2514088451862335, + "learning_rate": 1.5457936179030253e-06, + "loss": 0.5131, + "step": 8336 + }, + { + "epoch": 0.6910070451719851, + "grad_norm": 0.23919901251792908, + "learning_rate": 1.54537919602155e-06, + "loss": 0.4639, + "step": 8337 + }, + { + "epoch": 0.6910899295482802, + "grad_norm": 0.2713187038898468, + "learning_rate": 1.5449647741400746e-06, + "loss": 0.5517, + "step": 8338 + }, + { + "epoch": 0.6911728139245752, + "grad_norm": 0.2399529665708542, + "learning_rate": 1.5445503522585995e-06, + "loss": 0.4913, + "step": 8339 + }, + { + "epoch": 0.6912556983008703, + "grad_norm": 0.2333318591117859, + "learning_rate": 1.5441359303771241e-06, + "loss": 0.4927, + "step": 8340 + }, + { + "epoch": 0.6913385826771653, + "grad_norm": 0.24193806946277618, + "learning_rate": 1.5437215084956486e-06, + "loss": 0.5309, + "step": 8341 + }, + { + "epoch": 0.6914214670534604, + "grad_norm": 0.24117913842201233, + "learning_rate": 1.5433070866141734e-06, + "loss": 0.4842, + "step": 8342 + }, + { + "epoch": 0.6915043514297555, + "grad_norm": 0.24553988873958588, + "learning_rate": 1.542892664732698e-06, + "loss": 0.4836, + "step": 8343 + }, + { + "epoch": 0.6915872358060505, + "grad_norm": 0.24563266336917877, + "learning_rate": 1.5424782428512227e-06, + "loss": 0.5267, + "step": 8344 + }, + { + "epoch": 0.6916701201823456, + "grad_norm": 0.24982744455337524, + "learning_rate": 1.5420638209697471e-06, + "loss": 0.5062, + "step": 8345 + }, + { + "epoch": 0.6917530045586407, + "grad_norm": 0.24736806750297546, + "learning_rate": 1.541649399088272e-06, + "loss": 0.4948, + "step": 8346 + }, + { + "epoch": 0.6918358889349357, + "grad_norm": 0.25244516134262085, + "learning_rate": 1.5412349772067966e-06, + "loss": 0.5452, + "step": 8347 + }, + { + "epoch": 0.6919187733112309, + "grad_norm": 0.2440262734889984, + "learning_rate": 1.5408205553253213e-06, + "loss": 0.4961, + "step": 8348 + }, + { + "epoch": 0.6920016576875259, + "grad_norm": 0.2412625104188919, + "learning_rate": 1.5404061334438461e-06, + "loss": 0.503, + "step": 8349 + }, + { + "epoch": 0.692084542063821, + "grad_norm": 0.22938407957553864, + "learning_rate": 1.5399917115623706e-06, + "loss": 0.4661, + "step": 8350 + }, + { + "epoch": 0.6921674264401161, + "grad_norm": 0.2512044906616211, + "learning_rate": 1.5395772896808952e-06, + "loss": 0.5317, + "step": 8351 + }, + { + "epoch": 0.6922503108164111, + "grad_norm": 0.24228604137897491, + "learning_rate": 1.53916286779942e-06, + "loss": 0.4774, + "step": 8352 + }, + { + "epoch": 0.6923331951927062, + "grad_norm": 0.25557583570480347, + "learning_rate": 1.5387484459179447e-06, + "loss": 0.499, + "step": 8353 + }, + { + "epoch": 0.6924160795690012, + "grad_norm": 0.2796831429004669, + "learning_rate": 1.5383340240364691e-06, + "loss": 0.5469, + "step": 8354 + }, + { + "epoch": 0.6924989639452963, + "grad_norm": 0.25290942192077637, + "learning_rate": 1.5379196021549938e-06, + "loss": 0.5182, + "step": 8355 + }, + { + "epoch": 0.6925818483215914, + "grad_norm": 0.25738459825515747, + "learning_rate": 1.5375051802735186e-06, + "loss": 0.5223, + "step": 8356 + }, + { + "epoch": 0.6926647326978864, + "grad_norm": 0.24911700189113617, + "learning_rate": 1.5370907583920433e-06, + "loss": 0.5096, + "step": 8357 + }, + { + "epoch": 0.6927476170741815, + "grad_norm": 0.23372849822044373, + "learning_rate": 1.5366763365105677e-06, + "loss": 0.4327, + "step": 8358 + }, + { + "epoch": 0.6928305014504765, + "grad_norm": 0.22896170616149902, + "learning_rate": 1.5362619146290926e-06, + "loss": 0.4806, + "step": 8359 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 0.2350788116455078, + "learning_rate": 1.5358474927476172e-06, + "loss": 0.5213, + "step": 8360 + }, + { + "epoch": 0.6929962702030668, + "grad_norm": 0.2538852095603943, + "learning_rate": 1.5354330708661418e-06, + "loss": 0.5383, + "step": 8361 + }, + { + "epoch": 0.6930791545793618, + "grad_norm": 0.24704349040985107, + "learning_rate": 1.5350186489846667e-06, + "loss": 0.5316, + "step": 8362 + }, + { + "epoch": 0.6931620389556569, + "grad_norm": 0.24367371201515198, + "learning_rate": 1.5346042271031911e-06, + "loss": 0.5179, + "step": 8363 + }, + { + "epoch": 0.693244923331952, + "grad_norm": 0.22632329165935516, + "learning_rate": 1.5341898052217158e-06, + "loss": 0.5021, + "step": 8364 + }, + { + "epoch": 0.693327807708247, + "grad_norm": 0.2505031228065491, + "learning_rate": 1.5337753833402404e-06, + "loss": 0.5406, + "step": 8365 + }, + { + "epoch": 0.6934106920845421, + "grad_norm": 0.27584734559059143, + "learning_rate": 1.5333609614587653e-06, + "loss": 0.5188, + "step": 8366 + }, + { + "epoch": 0.6934935764608371, + "grad_norm": 0.2562057375907898, + "learning_rate": 1.5329465395772897e-06, + "loss": 0.5163, + "step": 8367 + }, + { + "epoch": 0.6935764608371322, + "grad_norm": 0.247248113155365, + "learning_rate": 1.5325321176958144e-06, + "loss": 0.5281, + "step": 8368 + }, + { + "epoch": 0.6936593452134273, + "grad_norm": 0.26059481501579285, + "learning_rate": 1.5321176958143392e-06, + "loss": 0.5391, + "step": 8369 + }, + { + "epoch": 0.6937422295897223, + "grad_norm": 0.24925461411476135, + "learning_rate": 1.5317032739328639e-06, + "loss": 0.4703, + "step": 8370 + }, + { + "epoch": 0.6938251139660174, + "grad_norm": 0.2632819414138794, + "learning_rate": 1.5312888520513885e-06, + "loss": 0.4951, + "step": 8371 + }, + { + "epoch": 0.6939079983423124, + "grad_norm": 0.23716473579406738, + "learning_rate": 1.530874430169913e-06, + "loss": 0.5057, + "step": 8372 + }, + { + "epoch": 0.6939908827186075, + "grad_norm": 0.2439935952425003, + "learning_rate": 1.5304600082884378e-06, + "loss": 0.5153, + "step": 8373 + }, + { + "epoch": 0.6940737670949026, + "grad_norm": 0.2734087109565735, + "learning_rate": 1.5300455864069624e-06, + "loss": 0.5034, + "step": 8374 + }, + { + "epoch": 0.6941566514711976, + "grad_norm": 0.24319836497306824, + "learning_rate": 1.529631164525487e-06, + "loss": 0.4713, + "step": 8375 + }, + { + "epoch": 0.6942395358474928, + "grad_norm": 0.24756325781345367, + "learning_rate": 1.5292167426440115e-06, + "loss": 0.4741, + "step": 8376 + }, + { + "epoch": 0.6943224202237878, + "grad_norm": 0.2668731212615967, + "learning_rate": 1.5288023207625364e-06, + "loss": 0.5188, + "step": 8377 + }, + { + "epoch": 0.6944053046000829, + "grad_norm": 0.23461757600307465, + "learning_rate": 1.528387898881061e-06, + "loss": 0.4681, + "step": 8378 + }, + { + "epoch": 0.694488188976378, + "grad_norm": 0.28325915336608887, + "learning_rate": 1.5279734769995859e-06, + "loss": 0.4886, + "step": 8379 + }, + { + "epoch": 0.694571073352673, + "grad_norm": 0.24916474521160126, + "learning_rate": 1.5275590551181105e-06, + "loss": 0.524, + "step": 8380 + }, + { + "epoch": 0.6946539577289681, + "grad_norm": 0.22786204516887665, + "learning_rate": 1.527144633236635e-06, + "loss": 0.4658, + "step": 8381 + }, + { + "epoch": 0.6947368421052632, + "grad_norm": 0.2607201933860779, + "learning_rate": 1.5267302113551596e-06, + "loss": 0.5212, + "step": 8382 + }, + { + "epoch": 0.6948197264815582, + "grad_norm": 0.23249809443950653, + "learning_rate": 1.5263157894736844e-06, + "loss": 0.4983, + "step": 8383 + }, + { + "epoch": 0.6949026108578533, + "grad_norm": 0.2543908357620239, + "learning_rate": 1.525901367592209e-06, + "loss": 0.5101, + "step": 8384 + }, + { + "epoch": 0.6949854952341483, + "grad_norm": 0.23820289969444275, + "learning_rate": 1.5254869457107335e-06, + "loss": 0.5012, + "step": 8385 + }, + { + "epoch": 0.6950683796104434, + "grad_norm": 0.2323910892009735, + "learning_rate": 1.5250725238292582e-06, + "loss": 0.4603, + "step": 8386 + }, + { + "epoch": 0.6951512639867385, + "grad_norm": 0.23545654118061066, + "learning_rate": 1.524658101947783e-06, + "loss": 0.464, + "step": 8387 + }, + { + "epoch": 0.6952341483630335, + "grad_norm": 0.2413708120584488, + "learning_rate": 1.5242436800663076e-06, + "loss": 0.4759, + "step": 8388 + }, + { + "epoch": 0.6953170327393287, + "grad_norm": 0.2508584260940552, + "learning_rate": 1.5238292581848325e-06, + "loss": 0.4976, + "step": 8389 + }, + { + "epoch": 0.6953999171156237, + "grad_norm": 0.24265332520008087, + "learning_rate": 1.523414836303357e-06, + "loss": 0.5117, + "step": 8390 + }, + { + "epoch": 0.6954828014919188, + "grad_norm": 0.25561636686325073, + "learning_rate": 1.5230004144218816e-06, + "loss": 0.5253, + "step": 8391 + }, + { + "epoch": 0.6955656858682139, + "grad_norm": 0.2462255209684372, + "learning_rate": 1.5225859925404062e-06, + "loss": 0.4997, + "step": 8392 + }, + { + "epoch": 0.6956485702445089, + "grad_norm": 0.2521069645881653, + "learning_rate": 1.522171570658931e-06, + "loss": 0.4753, + "step": 8393 + }, + { + "epoch": 0.695731454620804, + "grad_norm": 0.24396608769893646, + "learning_rate": 1.5217571487774555e-06, + "loss": 0.4844, + "step": 8394 + }, + { + "epoch": 0.6958143389970991, + "grad_norm": 0.2316456139087677, + "learning_rate": 1.5213427268959802e-06, + "loss": 0.4727, + "step": 8395 + }, + { + "epoch": 0.6958972233733941, + "grad_norm": 0.24840569496154785, + "learning_rate": 1.5209283050145048e-06, + "loss": 0.5328, + "step": 8396 + }, + { + "epoch": 0.6959801077496892, + "grad_norm": 0.24838104844093323, + "learning_rate": 1.5205138831330297e-06, + "loss": 0.5228, + "step": 8397 + }, + { + "epoch": 0.6960629921259842, + "grad_norm": 0.22455810010433197, + "learning_rate": 1.5200994612515543e-06, + "loss": 0.4496, + "step": 8398 + }, + { + "epoch": 0.6961458765022793, + "grad_norm": 0.24708344042301178, + "learning_rate": 1.5196850393700787e-06, + "loss": 0.512, + "step": 8399 + }, + { + "epoch": 0.6962287608785744, + "grad_norm": 0.4508987069129944, + "learning_rate": 1.5192706174886036e-06, + "loss": 0.5135, + "step": 8400 + }, + { + "epoch": 0.6963116452548694, + "grad_norm": 0.261269211769104, + "learning_rate": 1.5188561956071282e-06, + "loss": 0.5334, + "step": 8401 + }, + { + "epoch": 0.6963945296311645, + "grad_norm": 0.25128188729286194, + "learning_rate": 1.5184417737256529e-06, + "loss": 0.5074, + "step": 8402 + }, + { + "epoch": 0.6964774140074595, + "grad_norm": 0.24411548674106598, + "learning_rate": 1.5180273518441773e-06, + "loss": 0.4779, + "step": 8403 + }, + { + "epoch": 0.6965602983837547, + "grad_norm": 0.21901147067546844, + "learning_rate": 1.5176129299627022e-06, + "loss": 0.4757, + "step": 8404 + }, + { + "epoch": 0.6966431827600498, + "grad_norm": 0.23926886916160583, + "learning_rate": 1.5171985080812268e-06, + "loss": 0.5105, + "step": 8405 + }, + { + "epoch": 0.6967260671363448, + "grad_norm": 0.24255181849002838, + "learning_rate": 1.5167840861997514e-06, + "loss": 0.4771, + "step": 8406 + }, + { + "epoch": 0.6968089515126399, + "grad_norm": 0.2470545768737793, + "learning_rate": 1.5163696643182763e-06, + "loss": 0.4768, + "step": 8407 + }, + { + "epoch": 0.6968918358889349, + "grad_norm": 0.24303194880485535, + "learning_rate": 1.5159552424368007e-06, + "loss": 0.4746, + "step": 8408 + }, + { + "epoch": 0.69697472026523, + "grad_norm": 0.2402554303407669, + "learning_rate": 1.5155408205553254e-06, + "loss": 0.4823, + "step": 8409 + }, + { + "epoch": 0.6970576046415251, + "grad_norm": 0.25641363859176636, + "learning_rate": 1.5151263986738502e-06, + "loss": 0.4758, + "step": 8410 + }, + { + "epoch": 0.6971404890178201, + "grad_norm": 0.24012036621570587, + "learning_rate": 1.5147119767923749e-06, + "loss": 0.4819, + "step": 8411 + }, + { + "epoch": 0.6972233733941152, + "grad_norm": 0.2456699162721634, + "learning_rate": 1.5142975549108993e-06, + "loss": 0.4955, + "step": 8412 + }, + { + "epoch": 0.6973062577704103, + "grad_norm": 0.2280259132385254, + "learning_rate": 1.513883133029424e-06, + "loss": 0.4783, + "step": 8413 + }, + { + "epoch": 0.6973891421467053, + "grad_norm": 0.2841648757457733, + "learning_rate": 1.5134687111479488e-06, + "loss": 0.4937, + "step": 8414 + }, + { + "epoch": 0.6974720265230004, + "grad_norm": 0.25400274991989136, + "learning_rate": 1.5130542892664734e-06, + "loss": 0.5239, + "step": 8415 + }, + { + "epoch": 0.6975549108992954, + "grad_norm": 0.2521774172782898, + "learning_rate": 1.512639867384998e-06, + "loss": 0.4968, + "step": 8416 + }, + { + "epoch": 0.6976377952755906, + "grad_norm": 0.23096849024295807, + "learning_rate": 1.5122254455035225e-06, + "loss": 0.5009, + "step": 8417 + }, + { + "epoch": 0.6977206796518857, + "grad_norm": 0.23432140052318573, + "learning_rate": 1.5118110236220474e-06, + "loss": 0.4939, + "step": 8418 + }, + { + "epoch": 0.6978035640281807, + "grad_norm": 0.24503222107887268, + "learning_rate": 1.511396601740572e-06, + "loss": 0.4827, + "step": 8419 + }, + { + "epoch": 0.6978864484044758, + "grad_norm": 0.24168303608894348, + "learning_rate": 1.5109821798590969e-06, + "loss": 0.5091, + "step": 8420 + }, + { + "epoch": 0.6979693327807708, + "grad_norm": 0.22889307141304016, + "learning_rate": 1.5105677579776213e-06, + "loss": 0.5106, + "step": 8421 + }, + { + "epoch": 0.6980522171570659, + "grad_norm": 0.23950761556625366, + "learning_rate": 1.510153336096146e-06, + "loss": 0.5026, + "step": 8422 + }, + { + "epoch": 0.698135101533361, + "grad_norm": 0.258231520652771, + "learning_rate": 1.5097389142146706e-06, + "loss": 0.5205, + "step": 8423 + }, + { + "epoch": 0.698217985909656, + "grad_norm": 0.23322635889053345, + "learning_rate": 1.5093244923331955e-06, + "loss": 0.4792, + "step": 8424 + }, + { + "epoch": 0.6983008702859511, + "grad_norm": 0.23953068256378174, + "learning_rate": 1.50891007045172e-06, + "loss": 0.475, + "step": 8425 + }, + { + "epoch": 0.6983837546622461, + "grad_norm": 0.25142961740493774, + "learning_rate": 1.5084956485702445e-06, + "loss": 0.5377, + "step": 8426 + }, + { + "epoch": 0.6984666390385412, + "grad_norm": 0.22327931225299835, + "learning_rate": 1.5080812266887692e-06, + "loss": 0.4461, + "step": 8427 + }, + { + "epoch": 0.6985495234148363, + "grad_norm": 0.25832805037498474, + "learning_rate": 1.507666804807294e-06, + "loss": 0.5015, + "step": 8428 + }, + { + "epoch": 0.6986324077911313, + "grad_norm": 0.22395361959934235, + "learning_rate": 1.5072523829258187e-06, + "loss": 0.504, + "step": 8429 + }, + { + "epoch": 0.6987152921674264, + "grad_norm": 0.23794467747211456, + "learning_rate": 1.506837961044343e-06, + "loss": 0.4947, + "step": 8430 + }, + { + "epoch": 0.6987981765437216, + "grad_norm": 0.23633109033107758, + "learning_rate": 1.506423539162868e-06, + "loss": 0.499, + "step": 8431 + }, + { + "epoch": 0.6988810609200166, + "grad_norm": 0.2484033703804016, + "learning_rate": 1.5060091172813926e-06, + "loss": 0.5132, + "step": 8432 + }, + { + "epoch": 0.6989639452963117, + "grad_norm": 0.2703877091407776, + "learning_rate": 1.5055946953999172e-06, + "loss": 0.5082, + "step": 8433 + }, + { + "epoch": 0.6990468296726067, + "grad_norm": 0.23613396286964417, + "learning_rate": 1.5051802735184417e-06, + "loss": 0.5139, + "step": 8434 + }, + { + "epoch": 0.6991297140489018, + "grad_norm": 0.2516641318798065, + "learning_rate": 1.5047658516369665e-06, + "loss": 0.4793, + "step": 8435 + }, + { + "epoch": 0.6992125984251969, + "grad_norm": 0.22615085542201996, + "learning_rate": 1.5043514297554912e-06, + "loss": 0.4667, + "step": 8436 + }, + { + "epoch": 0.6992954828014919, + "grad_norm": 0.23646114766597748, + "learning_rate": 1.5039370078740158e-06, + "loss": 0.472, + "step": 8437 + }, + { + "epoch": 0.699378367177787, + "grad_norm": 0.24718576669692993, + "learning_rate": 1.5035225859925407e-06, + "loss": 0.4813, + "step": 8438 + }, + { + "epoch": 0.699461251554082, + "grad_norm": 0.23602333664894104, + "learning_rate": 1.503108164111065e-06, + "loss": 0.4652, + "step": 8439 + }, + { + "epoch": 0.6995441359303771, + "grad_norm": 0.23740249872207642, + "learning_rate": 1.5026937422295897e-06, + "loss": 0.5213, + "step": 8440 + }, + { + "epoch": 0.6996270203066722, + "grad_norm": 0.2304675579071045, + "learning_rate": 1.5022793203481146e-06, + "loss": 0.4882, + "step": 8441 + }, + { + "epoch": 0.6997099046829672, + "grad_norm": 0.27595409750938416, + "learning_rate": 1.5018648984666392e-06, + "loss": 0.5267, + "step": 8442 + }, + { + "epoch": 0.6997927890592623, + "grad_norm": 0.24758248031139374, + "learning_rate": 1.5014504765851637e-06, + "loss": 0.4707, + "step": 8443 + }, + { + "epoch": 0.6998756734355573, + "grad_norm": 0.24388962984085083, + "learning_rate": 1.5010360547036883e-06, + "loss": 0.5295, + "step": 8444 + }, + { + "epoch": 0.6999585578118525, + "grad_norm": 0.23885446786880493, + "learning_rate": 1.5006216328222132e-06, + "loss": 0.5029, + "step": 8445 + }, + { + "epoch": 0.7000414421881476, + "grad_norm": 0.23660160601139069, + "learning_rate": 1.5002072109407378e-06, + "loss": 0.4978, + "step": 8446 + }, + { + "epoch": 0.7001243265644426, + "grad_norm": 0.2548794150352478, + "learning_rate": 1.4997927890592625e-06, + "loss": 0.5437, + "step": 8447 + }, + { + "epoch": 0.7002072109407377, + "grad_norm": 0.2446957230567932, + "learning_rate": 1.4993783671777871e-06, + "loss": 0.4907, + "step": 8448 + }, + { + "epoch": 0.7002900953170328, + "grad_norm": 0.26771944761276245, + "learning_rate": 1.4989639452963118e-06, + "loss": 0.5123, + "step": 8449 + }, + { + "epoch": 0.7003729796933278, + "grad_norm": 0.25026747584342957, + "learning_rate": 1.4985495234148364e-06, + "loss": 0.519, + "step": 8450 + }, + { + "epoch": 0.7004558640696229, + "grad_norm": 0.23133306205272675, + "learning_rate": 1.4981351015333613e-06, + "loss": 0.4577, + "step": 8451 + }, + { + "epoch": 0.7005387484459179, + "grad_norm": 0.23684480786323547, + "learning_rate": 1.4977206796518857e-06, + "loss": 0.4621, + "step": 8452 + }, + { + "epoch": 0.700621632822213, + "grad_norm": 0.2864174544811249, + "learning_rate": 1.4973062577704103e-06, + "loss": 0.5144, + "step": 8453 + }, + { + "epoch": 0.7007045171985081, + "grad_norm": 0.241298109292984, + "learning_rate": 1.496891835888935e-06, + "loss": 0.5015, + "step": 8454 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 0.2461838275194168, + "learning_rate": 1.4964774140074598e-06, + "loss": 0.4706, + "step": 8455 + }, + { + "epoch": 0.7008702859510982, + "grad_norm": 0.24209155142307281, + "learning_rate": 1.4960629921259845e-06, + "loss": 0.511, + "step": 8456 + }, + { + "epoch": 0.7009531703273932, + "grad_norm": 0.23693256080150604, + "learning_rate": 1.495648570244509e-06, + "loss": 0.514, + "step": 8457 + }, + { + "epoch": 0.7010360547036883, + "grad_norm": 0.23758763074874878, + "learning_rate": 1.4952341483630338e-06, + "loss": 0.5287, + "step": 8458 + }, + { + "epoch": 0.7011189390799835, + "grad_norm": 0.2288748174905777, + "learning_rate": 1.4948197264815584e-06, + "loss": 0.471, + "step": 8459 + }, + { + "epoch": 0.7012018234562785, + "grad_norm": 0.24685609340667725, + "learning_rate": 1.494405304600083e-06, + "loss": 0.5255, + "step": 8460 + }, + { + "epoch": 0.7012847078325736, + "grad_norm": 0.23398041725158691, + "learning_rate": 1.4939908827186075e-06, + "loss": 0.5081, + "step": 8461 + }, + { + "epoch": 0.7013675922088686, + "grad_norm": 0.24576526880264282, + "learning_rate": 1.4935764608371323e-06, + "loss": 0.4956, + "step": 8462 + }, + { + "epoch": 0.7014504765851637, + "grad_norm": 0.27074670791625977, + "learning_rate": 1.493162038955657e-06, + "loss": 0.5082, + "step": 8463 + }, + { + "epoch": 0.7015333609614588, + "grad_norm": 0.22627101838588715, + "learning_rate": 1.4927476170741816e-06, + "loss": 0.4271, + "step": 8464 + }, + { + "epoch": 0.7016162453377538, + "grad_norm": 0.23196810483932495, + "learning_rate": 1.4923331951927065e-06, + "loss": 0.4783, + "step": 8465 + }, + { + "epoch": 0.7016991297140489, + "grad_norm": 0.254933625459671, + "learning_rate": 1.491918773311231e-06, + "loss": 0.4886, + "step": 8466 + }, + { + "epoch": 0.701782014090344, + "grad_norm": 0.23569533228874207, + "learning_rate": 1.4915043514297555e-06, + "loss": 0.4734, + "step": 8467 + }, + { + "epoch": 0.701864898466639, + "grad_norm": 0.2369750589132309, + "learning_rate": 1.4910899295482804e-06, + "loss": 0.5096, + "step": 8468 + }, + { + "epoch": 0.7019477828429341, + "grad_norm": 0.2380780726671219, + "learning_rate": 1.490675507666805e-06, + "loss": 0.4523, + "step": 8469 + }, + { + "epoch": 0.7020306672192291, + "grad_norm": 0.2389364391565323, + "learning_rate": 1.4902610857853295e-06, + "loss": 0.4858, + "step": 8470 + }, + { + "epoch": 0.7021135515955242, + "grad_norm": 0.23384974896907806, + "learning_rate": 1.4898466639038541e-06, + "loss": 0.4794, + "step": 8471 + }, + { + "epoch": 0.7021964359718194, + "grad_norm": 0.2330332249403, + "learning_rate": 1.489432242022379e-06, + "loss": 0.5142, + "step": 8472 + }, + { + "epoch": 0.7022793203481144, + "grad_norm": 0.2703178822994232, + "learning_rate": 1.4890178201409036e-06, + "loss": 0.4753, + "step": 8473 + }, + { + "epoch": 0.7023622047244095, + "grad_norm": 0.24302196502685547, + "learning_rate": 1.4886033982594283e-06, + "loss": 0.5292, + "step": 8474 + }, + { + "epoch": 0.7024450891007045, + "grad_norm": 0.25824087858200073, + "learning_rate": 1.4881889763779527e-06, + "loss": 0.5378, + "step": 8475 + }, + { + "epoch": 0.7025279734769996, + "grad_norm": 0.2731492221355438, + "learning_rate": 1.4877745544964776e-06, + "loss": 0.5084, + "step": 8476 + }, + { + "epoch": 0.7026108578532947, + "grad_norm": 0.2371882200241089, + "learning_rate": 1.4873601326150022e-06, + "loss": 0.5034, + "step": 8477 + }, + { + "epoch": 0.7026937422295897, + "grad_norm": 0.23955635726451874, + "learning_rate": 1.486945710733527e-06, + "loss": 0.5255, + "step": 8478 + }, + { + "epoch": 0.7027766266058848, + "grad_norm": 0.2451208382844925, + "learning_rate": 1.4865312888520515e-06, + "loss": 0.5083, + "step": 8479 + }, + { + "epoch": 0.7028595109821799, + "grad_norm": 0.23578408360481262, + "learning_rate": 1.4861168669705761e-06, + "loss": 0.5104, + "step": 8480 + }, + { + "epoch": 0.7029423953584749, + "grad_norm": 0.2528175115585327, + "learning_rate": 1.4857024450891008e-06, + "loss": 0.5185, + "step": 8481 + }, + { + "epoch": 0.70302527973477, + "grad_norm": 0.2526828348636627, + "learning_rate": 1.4852880232076256e-06, + "loss": 0.5322, + "step": 8482 + }, + { + "epoch": 0.703108164111065, + "grad_norm": 0.25708842277526855, + "learning_rate": 1.4848736013261503e-06, + "loss": 0.449, + "step": 8483 + }, + { + "epoch": 0.7031910484873601, + "grad_norm": 0.275142103433609, + "learning_rate": 1.4844591794446747e-06, + "loss": 0.5198, + "step": 8484 + }, + { + "epoch": 0.7032739328636552, + "grad_norm": 0.2556648850440979, + "learning_rate": 1.4840447575631993e-06, + "loss": 0.5107, + "step": 8485 + }, + { + "epoch": 0.7033568172399502, + "grad_norm": 0.2346121072769165, + "learning_rate": 1.4836303356817242e-06, + "loss": 0.4933, + "step": 8486 + }, + { + "epoch": 0.7034397016162454, + "grad_norm": 0.23165251314640045, + "learning_rate": 1.4832159138002488e-06, + "loss": 0.4995, + "step": 8487 + }, + { + "epoch": 0.7035225859925404, + "grad_norm": 0.22874577343463898, + "learning_rate": 1.4828014919187733e-06, + "loss": 0.4772, + "step": 8488 + }, + { + "epoch": 0.7036054703688355, + "grad_norm": 0.24079668521881104, + "learning_rate": 1.4823870700372981e-06, + "loss": 0.5045, + "step": 8489 + }, + { + "epoch": 0.7036883547451306, + "grad_norm": 0.2592758536338806, + "learning_rate": 1.4819726481558228e-06, + "loss": 0.4819, + "step": 8490 + }, + { + "epoch": 0.7037712391214256, + "grad_norm": 0.24797825515270233, + "learning_rate": 1.4815582262743474e-06, + "loss": 0.5108, + "step": 8491 + }, + { + "epoch": 0.7038541234977207, + "grad_norm": 0.24840524792671204, + "learning_rate": 1.4811438043928723e-06, + "loss": 0.4753, + "step": 8492 + }, + { + "epoch": 0.7039370078740157, + "grad_norm": 0.24021302163600922, + "learning_rate": 1.4807293825113967e-06, + "loss": 0.486, + "step": 8493 + }, + { + "epoch": 0.7040198922503108, + "grad_norm": 0.22576092183589935, + "learning_rate": 1.4803149606299213e-06, + "loss": 0.4572, + "step": 8494 + }, + { + "epoch": 0.7041027766266059, + "grad_norm": 0.2626515030860901, + "learning_rate": 1.479900538748446e-06, + "loss": 0.5285, + "step": 8495 + }, + { + "epoch": 0.7041856610029009, + "grad_norm": 0.24324116110801697, + "learning_rate": 1.4794861168669708e-06, + "loss": 0.5385, + "step": 8496 + }, + { + "epoch": 0.704268545379196, + "grad_norm": 0.2450491040945053, + "learning_rate": 1.4790716949854953e-06, + "loss": 0.4967, + "step": 8497 + }, + { + "epoch": 0.7043514297554911, + "grad_norm": 0.25503966212272644, + "learning_rate": 1.47865727310402e-06, + "loss": 0.499, + "step": 8498 + }, + { + "epoch": 0.7044343141317861, + "grad_norm": 0.22867895662784576, + "learning_rate": 1.4782428512225448e-06, + "loss": 0.4827, + "step": 8499 + }, + { + "epoch": 0.7045171985080813, + "grad_norm": 0.25253990292549133, + "learning_rate": 1.4778284293410694e-06, + "loss": 0.4962, + "step": 8500 + }, + { + "epoch": 0.7046000828843763, + "grad_norm": 0.23030179738998413, + "learning_rate": 1.477414007459594e-06, + "loss": 0.4641, + "step": 8501 + }, + { + "epoch": 0.7046829672606714, + "grad_norm": 0.3303874135017395, + "learning_rate": 1.4769995855781185e-06, + "loss": 0.4877, + "step": 8502 + }, + { + "epoch": 0.7047658516369665, + "grad_norm": 0.26138436794281006, + "learning_rate": 1.4765851636966434e-06, + "loss": 0.4993, + "step": 8503 + }, + { + "epoch": 0.7048487360132615, + "grad_norm": 0.24164626002311707, + "learning_rate": 1.476170741815168e-06, + "loss": 0.4755, + "step": 8504 + }, + { + "epoch": 0.7049316203895566, + "grad_norm": 0.2549307942390442, + "learning_rate": 1.4757563199336926e-06, + "loss": 0.5345, + "step": 8505 + }, + { + "epoch": 0.7050145047658516, + "grad_norm": 0.23955599963665009, + "learning_rate": 1.475341898052217e-06, + "loss": 0.513, + "step": 8506 + }, + { + "epoch": 0.7050973891421467, + "grad_norm": 0.23878774046897888, + "learning_rate": 1.474927476170742e-06, + "loss": 0.4694, + "step": 8507 + }, + { + "epoch": 0.7051802735184418, + "grad_norm": 0.24608634412288666, + "learning_rate": 1.4745130542892666e-06, + "loss": 0.4684, + "step": 8508 + }, + { + "epoch": 0.7052631578947368, + "grad_norm": 0.24802447855472565, + "learning_rate": 1.4740986324077914e-06, + "loss": 0.4917, + "step": 8509 + }, + { + "epoch": 0.7053460422710319, + "grad_norm": 0.25912633538246155, + "learning_rate": 1.4736842105263159e-06, + "loss": 0.4835, + "step": 8510 + }, + { + "epoch": 0.7054289266473269, + "grad_norm": 0.25147220492362976, + "learning_rate": 1.4732697886448405e-06, + "loss": 0.5112, + "step": 8511 + }, + { + "epoch": 0.705511811023622, + "grad_norm": 0.24066859483718872, + "learning_rate": 1.4728553667633651e-06, + "loss": 0.4724, + "step": 8512 + }, + { + "epoch": 0.7055946953999171, + "grad_norm": 0.2495303452014923, + "learning_rate": 1.47244094488189e-06, + "loss": 0.4846, + "step": 8513 + }, + { + "epoch": 0.7056775797762121, + "grad_norm": 0.27521640062332153, + "learning_rate": 1.4720265230004146e-06, + "loss": 0.5707, + "step": 8514 + }, + { + "epoch": 0.7057604641525073, + "grad_norm": 0.2537860870361328, + "learning_rate": 1.471612101118939e-06, + "loss": 0.5103, + "step": 8515 + }, + { + "epoch": 0.7058433485288024, + "grad_norm": 0.24069590866565704, + "learning_rate": 1.4711976792374637e-06, + "loss": 0.5142, + "step": 8516 + }, + { + "epoch": 0.7059262329050974, + "grad_norm": 0.2247067093849182, + "learning_rate": 1.4707832573559886e-06, + "loss": 0.4482, + "step": 8517 + }, + { + "epoch": 0.7060091172813925, + "grad_norm": 0.2599788010120392, + "learning_rate": 1.4703688354745132e-06, + "loss": 0.4636, + "step": 8518 + }, + { + "epoch": 0.7060920016576875, + "grad_norm": 0.2592338025569916, + "learning_rate": 1.4699544135930376e-06, + "loss": 0.5019, + "step": 8519 + }, + { + "epoch": 0.7061748860339826, + "grad_norm": 0.2454696148633957, + "learning_rate": 1.4695399917115625e-06, + "loss": 0.4985, + "step": 8520 + }, + { + "epoch": 0.7062577704102777, + "grad_norm": 0.2432490438222885, + "learning_rate": 1.4691255698300871e-06, + "loss": 0.5021, + "step": 8521 + }, + { + "epoch": 0.7063406547865727, + "grad_norm": 0.22915400564670563, + "learning_rate": 1.4687111479486118e-06, + "loss": 0.4971, + "step": 8522 + }, + { + "epoch": 0.7064235391628678, + "grad_norm": 0.25342264771461487, + "learning_rate": 1.4682967260671366e-06, + "loss": 0.4942, + "step": 8523 + }, + { + "epoch": 0.7065064235391628, + "grad_norm": 0.23674166202545166, + "learning_rate": 1.467882304185661e-06, + "loss": 0.5491, + "step": 8524 + }, + { + "epoch": 0.7065893079154579, + "grad_norm": 0.22896523773670197, + "learning_rate": 1.4674678823041857e-06, + "loss": 0.5031, + "step": 8525 + }, + { + "epoch": 0.706672192291753, + "grad_norm": 0.23406754434108734, + "learning_rate": 1.4670534604227104e-06, + "loss": 0.4954, + "step": 8526 + }, + { + "epoch": 0.706755076668048, + "grad_norm": 0.24271178245544434, + "learning_rate": 1.4666390385412352e-06, + "loss": 0.4766, + "step": 8527 + }, + { + "epoch": 0.7068379610443432, + "grad_norm": 0.25465360283851624, + "learning_rate": 1.4662246166597597e-06, + "loss": 0.5423, + "step": 8528 + }, + { + "epoch": 0.7069208454206382, + "grad_norm": 0.23883527517318726, + "learning_rate": 1.4658101947782843e-06, + "loss": 0.5236, + "step": 8529 + }, + { + "epoch": 0.7070037297969333, + "grad_norm": 0.24184322357177734, + "learning_rate": 1.4653957728968092e-06, + "loss": 0.4863, + "step": 8530 + }, + { + "epoch": 0.7070866141732284, + "grad_norm": 0.24556291103363037, + "learning_rate": 1.4649813510153338e-06, + "loss": 0.4935, + "step": 8531 + }, + { + "epoch": 0.7071694985495234, + "grad_norm": 0.2457484006881714, + "learning_rate": 1.4645669291338584e-06, + "loss": 0.4873, + "step": 8532 + }, + { + "epoch": 0.7072523829258185, + "grad_norm": 0.238223597407341, + "learning_rate": 1.4641525072523829e-06, + "loss": 0.4741, + "step": 8533 + }, + { + "epoch": 0.7073352673021136, + "grad_norm": 0.2518101930618286, + "learning_rate": 1.4637380853709077e-06, + "loss": 0.5163, + "step": 8534 + }, + { + "epoch": 0.7074181516784086, + "grad_norm": 0.2559610903263092, + "learning_rate": 1.4633236634894324e-06, + "loss": 0.5408, + "step": 8535 + }, + { + "epoch": 0.7075010360547037, + "grad_norm": 0.2757256031036377, + "learning_rate": 1.462909241607957e-06, + "loss": 0.5315, + "step": 8536 + }, + { + "epoch": 0.7075839204309987, + "grad_norm": 0.24302177131175995, + "learning_rate": 1.4624948197264817e-06, + "loss": 0.5218, + "step": 8537 + }, + { + "epoch": 0.7076668048072938, + "grad_norm": 0.24641838669776917, + "learning_rate": 1.4620803978450063e-06, + "loss": 0.5147, + "step": 8538 + }, + { + "epoch": 0.7077496891835889, + "grad_norm": 0.23914003372192383, + "learning_rate": 1.461665975963531e-06, + "loss": 0.4902, + "step": 8539 + }, + { + "epoch": 0.7078325735598839, + "grad_norm": 0.24187055230140686, + "learning_rate": 1.4612515540820558e-06, + "loss": 0.5227, + "step": 8540 + }, + { + "epoch": 0.707915457936179, + "grad_norm": 0.24758659303188324, + "learning_rate": 1.4608371322005804e-06, + "loss": 0.5005, + "step": 8541 + }, + { + "epoch": 0.707998342312474, + "grad_norm": 0.22631829977035522, + "learning_rate": 1.4604227103191049e-06, + "loss": 0.4585, + "step": 8542 + }, + { + "epoch": 0.7080812266887692, + "grad_norm": 0.2663765549659729, + "learning_rate": 1.4600082884376295e-06, + "loss": 0.4901, + "step": 8543 + }, + { + "epoch": 0.7081641110650643, + "grad_norm": 0.2506657540798187, + "learning_rate": 1.4595938665561544e-06, + "loss": 0.5092, + "step": 8544 + }, + { + "epoch": 0.7082469954413593, + "grad_norm": 0.2374826818704605, + "learning_rate": 1.459179444674679e-06, + "loss": 0.483, + "step": 8545 + }, + { + "epoch": 0.7083298798176544, + "grad_norm": 0.24383142590522766, + "learning_rate": 1.4587650227932034e-06, + "loss": 0.4517, + "step": 8546 + }, + { + "epoch": 0.7084127641939495, + "grad_norm": 0.2534315884113312, + "learning_rate": 1.4583506009117283e-06, + "loss": 0.5, + "step": 8547 + }, + { + "epoch": 0.7084956485702445, + "grad_norm": 0.26466652750968933, + "learning_rate": 1.457936179030253e-06, + "loss": 0.5258, + "step": 8548 + }, + { + "epoch": 0.7085785329465396, + "grad_norm": 0.24726761877536774, + "learning_rate": 1.4575217571487776e-06, + "loss": 0.493, + "step": 8549 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 0.23872365057468414, + "learning_rate": 1.4571073352673024e-06, + "loss": 0.4991, + "step": 8550 + }, + { + "epoch": 0.7087443016991297, + "grad_norm": 0.2299206256866455, + "learning_rate": 1.4566929133858269e-06, + "loss": 0.4424, + "step": 8551 + }, + { + "epoch": 0.7088271860754248, + "grad_norm": 0.21521694958209991, + "learning_rate": 1.4562784915043515e-06, + "loss": 0.443, + "step": 8552 + }, + { + "epoch": 0.7089100704517198, + "grad_norm": 0.2617732882499695, + "learning_rate": 1.4558640696228762e-06, + "loss": 0.527, + "step": 8553 + }, + { + "epoch": 0.7089929548280149, + "grad_norm": 0.24927756190299988, + "learning_rate": 1.455449647741401e-06, + "loss": 0.4979, + "step": 8554 + }, + { + "epoch": 0.7090758392043099, + "grad_norm": 0.270692378282547, + "learning_rate": 1.4550352258599255e-06, + "loss": 0.5072, + "step": 8555 + }, + { + "epoch": 0.709158723580605, + "grad_norm": 0.23276378214359283, + "learning_rate": 1.45462080397845e-06, + "loss": 0.4916, + "step": 8556 + }, + { + "epoch": 0.7092416079569002, + "grad_norm": 0.25130319595336914, + "learning_rate": 1.4542063820969747e-06, + "loss": 0.5143, + "step": 8557 + }, + { + "epoch": 0.7093244923331952, + "grad_norm": 0.23895779252052307, + "learning_rate": 1.4537919602154996e-06, + "loss": 0.4892, + "step": 8558 + }, + { + "epoch": 0.7094073767094903, + "grad_norm": 0.2405640035867691, + "learning_rate": 1.4533775383340242e-06, + "loss": 0.4482, + "step": 8559 + }, + { + "epoch": 0.7094902610857853, + "grad_norm": 0.250596821308136, + "learning_rate": 1.4529631164525487e-06, + "loss": 0.5433, + "step": 8560 + }, + { + "epoch": 0.7095731454620804, + "grad_norm": 0.25519198179244995, + "learning_rate": 1.4525486945710735e-06, + "loss": 0.4851, + "step": 8561 + }, + { + "epoch": 0.7096560298383755, + "grad_norm": 0.23664671182632446, + "learning_rate": 1.4521342726895982e-06, + "loss": 0.4891, + "step": 8562 + }, + { + "epoch": 0.7097389142146705, + "grad_norm": 0.23781026899814606, + "learning_rate": 1.4517198508081228e-06, + "loss": 0.4878, + "step": 8563 + }, + { + "epoch": 0.7098217985909656, + "grad_norm": 0.2425031065940857, + "learning_rate": 1.4513054289266472e-06, + "loss": 0.4809, + "step": 8564 + }, + { + "epoch": 0.7099046829672607, + "grad_norm": 0.2650316655635834, + "learning_rate": 1.450891007045172e-06, + "loss": 0.5197, + "step": 8565 + }, + { + "epoch": 0.7099875673435557, + "grad_norm": 0.24599523842334747, + "learning_rate": 1.4504765851636967e-06, + "loss": 0.4893, + "step": 8566 + }, + { + "epoch": 0.7100704517198508, + "grad_norm": 0.22113832831382751, + "learning_rate": 1.4500621632822214e-06, + "loss": 0.4825, + "step": 8567 + }, + { + "epoch": 0.7101533360961458, + "grad_norm": 0.23113299906253815, + "learning_rate": 1.4496477414007462e-06, + "loss": 0.4955, + "step": 8568 + }, + { + "epoch": 0.710236220472441, + "grad_norm": 0.24717682600021362, + "learning_rate": 1.4492333195192707e-06, + "loss": 0.4994, + "step": 8569 + }, + { + "epoch": 0.7103191048487361, + "grad_norm": 0.25693178176879883, + "learning_rate": 1.4488188976377953e-06, + "loss": 0.5237, + "step": 8570 + }, + { + "epoch": 0.7104019892250311, + "grad_norm": 0.2542356550693512, + "learning_rate": 1.4484044757563202e-06, + "loss": 0.5245, + "step": 8571 + }, + { + "epoch": 0.7104848736013262, + "grad_norm": 0.2502613961696625, + "learning_rate": 1.4479900538748448e-06, + "loss": 0.5091, + "step": 8572 + }, + { + "epoch": 0.7105677579776212, + "grad_norm": 0.23184414207935333, + "learning_rate": 1.4475756319933692e-06, + "loss": 0.5108, + "step": 8573 + }, + { + "epoch": 0.7106506423539163, + "grad_norm": 0.2582687735557556, + "learning_rate": 1.4471612101118939e-06, + "loss": 0.5756, + "step": 8574 + }, + { + "epoch": 0.7107335267302114, + "grad_norm": 0.25204113125801086, + "learning_rate": 1.4467467882304187e-06, + "loss": 0.5251, + "step": 8575 + }, + { + "epoch": 0.7108164111065064, + "grad_norm": 0.22884295880794525, + "learning_rate": 1.4463323663489434e-06, + "loss": 0.4867, + "step": 8576 + }, + { + "epoch": 0.7108992954828015, + "grad_norm": 0.2716238498687744, + "learning_rate": 1.445917944467468e-06, + "loss": 0.493, + "step": 8577 + }, + { + "epoch": 0.7109821798590965, + "grad_norm": 0.24772800505161285, + "learning_rate": 1.4455035225859927e-06, + "loss": 0.5318, + "step": 8578 + }, + { + "epoch": 0.7110650642353916, + "grad_norm": 0.25086092948913574, + "learning_rate": 1.4450891007045173e-06, + "loss": 0.5311, + "step": 8579 + }, + { + "epoch": 0.7111479486116867, + "grad_norm": 0.24432384967803955, + "learning_rate": 1.444674678823042e-06, + "loss": 0.5527, + "step": 8580 + }, + { + "epoch": 0.7112308329879817, + "grad_norm": 0.2487839013338089, + "learning_rate": 1.4442602569415668e-06, + "loss": 0.5195, + "step": 8581 + }, + { + "epoch": 0.7113137173642768, + "grad_norm": 0.2970665991306305, + "learning_rate": 1.4438458350600913e-06, + "loss": 0.5198, + "step": 8582 + }, + { + "epoch": 0.711396601740572, + "grad_norm": 0.2406621128320694, + "learning_rate": 1.443431413178616e-06, + "loss": 0.4892, + "step": 8583 + }, + { + "epoch": 0.711479486116867, + "grad_norm": 0.23739312589168549, + "learning_rate": 1.4430169912971405e-06, + "loss": 0.5198, + "step": 8584 + }, + { + "epoch": 0.7115623704931621, + "grad_norm": 0.25688570737838745, + "learning_rate": 1.4426025694156654e-06, + "loss": 0.5021, + "step": 8585 + }, + { + "epoch": 0.7116452548694571, + "grad_norm": 0.2729821503162384, + "learning_rate": 1.4421881475341898e-06, + "loss": 0.5468, + "step": 8586 + }, + { + "epoch": 0.7117281392457522, + "grad_norm": 0.23702508211135864, + "learning_rate": 1.4417737256527145e-06, + "loss": 0.4815, + "step": 8587 + }, + { + "epoch": 0.7118110236220473, + "grad_norm": 0.273916631937027, + "learning_rate": 1.4413593037712393e-06, + "loss": 0.4954, + "step": 8588 + }, + { + "epoch": 0.7118939079983423, + "grad_norm": 0.23420047760009766, + "learning_rate": 1.440944881889764e-06, + "loss": 0.498, + "step": 8589 + }, + { + "epoch": 0.7119767923746374, + "grad_norm": 0.2503052353858948, + "learning_rate": 1.4405304600082886e-06, + "loss": 0.4828, + "step": 8590 + }, + { + "epoch": 0.7120596767509324, + "grad_norm": 0.24778102338314056, + "learning_rate": 1.440116038126813e-06, + "loss": 0.4805, + "step": 8591 + }, + { + "epoch": 0.7121425611272275, + "grad_norm": 0.2666938304901123, + "learning_rate": 1.439701616245338e-06, + "loss": 0.5255, + "step": 8592 + }, + { + "epoch": 0.7122254455035226, + "grad_norm": 0.2309558093547821, + "learning_rate": 1.4392871943638625e-06, + "loss": 0.4898, + "step": 8593 + }, + { + "epoch": 0.7123083298798176, + "grad_norm": 0.2585700750350952, + "learning_rate": 1.4388727724823872e-06, + "loss": 0.4995, + "step": 8594 + }, + { + "epoch": 0.7123912142561127, + "grad_norm": 0.24335183203220367, + "learning_rate": 1.4384583506009116e-06, + "loss": 0.4654, + "step": 8595 + }, + { + "epoch": 0.7124740986324077, + "grad_norm": 0.24218983948230743, + "learning_rate": 1.4380439287194365e-06, + "loss": 0.5266, + "step": 8596 + }, + { + "epoch": 0.7125569830087028, + "grad_norm": 0.24543385207653046, + "learning_rate": 1.4376295068379611e-06, + "loss": 0.5219, + "step": 8597 + }, + { + "epoch": 0.712639867384998, + "grad_norm": 0.2407945990562439, + "learning_rate": 1.437215084956486e-06, + "loss": 0.4789, + "step": 8598 + }, + { + "epoch": 0.712722751761293, + "grad_norm": 0.23732717335224152, + "learning_rate": 1.4368006630750106e-06, + "loss": 0.4719, + "step": 8599 + }, + { + "epoch": 0.7128056361375881, + "grad_norm": 0.23688261210918427, + "learning_rate": 1.436386241193535e-06, + "loss": 0.5109, + "step": 8600 + }, + { + "epoch": 0.7128885205138832, + "grad_norm": 0.2501402795314789, + "learning_rate": 1.4359718193120597e-06, + "loss": 0.5472, + "step": 8601 + }, + { + "epoch": 0.7129714048901782, + "grad_norm": 0.23888307809829712, + "learning_rate": 1.4355573974305845e-06, + "loss": 0.4852, + "step": 8602 + }, + { + "epoch": 0.7130542892664733, + "grad_norm": 0.25531211495399475, + "learning_rate": 1.4351429755491092e-06, + "loss": 0.5089, + "step": 8603 + }, + { + "epoch": 0.7131371736427683, + "grad_norm": 0.24283672869205475, + "learning_rate": 1.4347285536676336e-06, + "loss": 0.4961, + "step": 8604 + }, + { + "epoch": 0.7132200580190634, + "grad_norm": 0.24875158071517944, + "learning_rate": 1.4343141317861583e-06, + "loss": 0.4735, + "step": 8605 + }, + { + "epoch": 0.7133029423953585, + "grad_norm": 0.23499397933483124, + "learning_rate": 1.4338997099046831e-06, + "loss": 0.4728, + "step": 8606 + }, + { + "epoch": 0.7133858267716535, + "grad_norm": 0.25186216831207275, + "learning_rate": 1.4334852880232078e-06, + "loss": 0.5393, + "step": 8607 + }, + { + "epoch": 0.7134687111479486, + "grad_norm": 0.2403862178325653, + "learning_rate": 1.4330708661417326e-06, + "loss": 0.5009, + "step": 8608 + }, + { + "epoch": 0.7135515955242436, + "grad_norm": 0.23796327412128448, + "learning_rate": 1.432656444260257e-06, + "loss": 0.5023, + "step": 8609 + }, + { + "epoch": 0.7136344799005387, + "grad_norm": 0.23705889284610748, + "learning_rate": 1.4322420223787817e-06, + "loss": 0.5335, + "step": 8610 + }, + { + "epoch": 0.7137173642768339, + "grad_norm": 0.2561888098716736, + "learning_rate": 1.4318276004973063e-06, + "loss": 0.5122, + "step": 8611 + }, + { + "epoch": 0.7138002486531289, + "grad_norm": 0.23384413123130798, + "learning_rate": 1.4314131786158312e-06, + "loss": 0.5052, + "step": 8612 + }, + { + "epoch": 0.713883133029424, + "grad_norm": 0.24679772555828094, + "learning_rate": 1.4309987567343556e-06, + "loss": 0.4881, + "step": 8613 + }, + { + "epoch": 0.713966017405719, + "grad_norm": 0.25904321670532227, + "learning_rate": 1.4305843348528803e-06, + "loss": 0.5277, + "step": 8614 + }, + { + "epoch": 0.7140489017820141, + "grad_norm": 0.2461068034172058, + "learning_rate": 1.430169912971405e-06, + "loss": 0.486, + "step": 8615 + }, + { + "epoch": 0.7141317861583092, + "grad_norm": 0.24692943692207336, + "learning_rate": 1.4297554910899298e-06, + "loss": 0.4952, + "step": 8616 + }, + { + "epoch": 0.7142146705346042, + "grad_norm": 0.24665188789367676, + "learning_rate": 1.4293410692084544e-06, + "loss": 0.4872, + "step": 8617 + }, + { + "epoch": 0.7142975549108993, + "grad_norm": 0.23328977823257446, + "learning_rate": 1.4289266473269788e-06, + "loss": 0.5023, + "step": 8618 + }, + { + "epoch": 0.7143804392871944, + "grad_norm": 0.23311160504817963, + "learning_rate": 1.4285122254455037e-06, + "loss": 0.4699, + "step": 8619 + }, + { + "epoch": 0.7144633236634894, + "grad_norm": 0.23193682730197906, + "learning_rate": 1.4280978035640283e-06, + "loss": 0.4691, + "step": 8620 + }, + { + "epoch": 0.7145462080397845, + "grad_norm": 0.2722693979740143, + "learning_rate": 1.427683381682553e-06, + "loss": 0.4564, + "step": 8621 + }, + { + "epoch": 0.7146290924160795, + "grad_norm": 0.24448418617248535, + "learning_rate": 1.4272689598010774e-06, + "loss": 0.4868, + "step": 8622 + }, + { + "epoch": 0.7147119767923746, + "grad_norm": 0.24539028108119965, + "learning_rate": 1.4268545379196023e-06, + "loss": 0.4764, + "step": 8623 + }, + { + "epoch": 0.7147948611686697, + "grad_norm": 0.23124291002750397, + "learning_rate": 1.426440116038127e-06, + "loss": 0.499, + "step": 8624 + }, + { + "epoch": 0.7148777455449647, + "grad_norm": 0.25085940957069397, + "learning_rate": 1.4260256941566516e-06, + "loss": 0.5511, + "step": 8625 + }, + { + "epoch": 0.7149606299212599, + "grad_norm": 0.23569589853286743, + "learning_rate": 1.4256112722751764e-06, + "loss": 0.4775, + "step": 8626 + }, + { + "epoch": 0.7150435142975549, + "grad_norm": 0.2673863470554352, + "learning_rate": 1.4251968503937008e-06, + "loss": 0.5207, + "step": 8627 + }, + { + "epoch": 0.71512639867385, + "grad_norm": 0.24088577926158905, + "learning_rate": 1.4247824285122255e-06, + "loss": 0.4724, + "step": 8628 + }, + { + "epoch": 0.7152092830501451, + "grad_norm": 0.23540116846561432, + "learning_rate": 1.4243680066307503e-06, + "loss": 0.4985, + "step": 8629 + }, + { + "epoch": 0.7152921674264401, + "grad_norm": 0.25157803297042847, + "learning_rate": 1.423953584749275e-06, + "loss": 0.5094, + "step": 8630 + }, + { + "epoch": 0.7153750518027352, + "grad_norm": 0.2722606956958771, + "learning_rate": 1.4235391628677994e-06, + "loss": 0.5216, + "step": 8631 + }, + { + "epoch": 0.7154579361790303, + "grad_norm": 0.23383471369743347, + "learning_rate": 1.423124740986324e-06, + "loss": 0.5057, + "step": 8632 + }, + { + "epoch": 0.7155408205553253, + "grad_norm": 0.25570788979530334, + "learning_rate": 1.422710319104849e-06, + "loss": 0.4929, + "step": 8633 + }, + { + "epoch": 0.7156237049316204, + "grad_norm": 0.2314014583826065, + "learning_rate": 1.4222958972233736e-06, + "loss": 0.4832, + "step": 8634 + }, + { + "epoch": 0.7157065893079154, + "grad_norm": 0.23632146418094635, + "learning_rate": 1.4218814753418982e-06, + "loss": 0.4849, + "step": 8635 + }, + { + "epoch": 0.7157894736842105, + "grad_norm": 0.2407282590866089, + "learning_rate": 1.4214670534604226e-06, + "loss": 0.4886, + "step": 8636 + }, + { + "epoch": 0.7158723580605056, + "grad_norm": 0.2682712972164154, + "learning_rate": 1.4210526315789475e-06, + "loss": 0.5113, + "step": 8637 + }, + { + "epoch": 0.7159552424368006, + "grad_norm": 0.2609236240386963, + "learning_rate": 1.4206382096974721e-06, + "loss": 0.4943, + "step": 8638 + }, + { + "epoch": 0.7160381268130958, + "grad_norm": 0.25318682193756104, + "learning_rate": 1.420223787815997e-06, + "loss": 0.5123, + "step": 8639 + }, + { + "epoch": 0.7161210111893908, + "grad_norm": 0.24802446365356445, + "learning_rate": 1.4198093659345214e-06, + "loss": 0.4969, + "step": 8640 + }, + { + "epoch": 0.7162038955656859, + "grad_norm": 0.25130435824394226, + "learning_rate": 1.419394944053046e-06, + "loss": 0.4737, + "step": 8641 + }, + { + "epoch": 0.716286779941981, + "grad_norm": 0.23842760920524597, + "learning_rate": 1.4189805221715707e-06, + "loss": 0.4947, + "step": 8642 + }, + { + "epoch": 0.716369664318276, + "grad_norm": 0.24390877783298492, + "learning_rate": 1.4185661002900956e-06, + "loss": 0.5149, + "step": 8643 + }, + { + "epoch": 0.7164525486945711, + "grad_norm": 0.294960618019104, + "learning_rate": 1.4181516784086202e-06, + "loss": 0.5375, + "step": 8644 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 0.25333136320114136, + "learning_rate": 1.4177372565271446e-06, + "loss": 0.5277, + "step": 8645 + }, + { + "epoch": 0.7166183174471612, + "grad_norm": 0.26154348254203796, + "learning_rate": 1.4173228346456693e-06, + "loss": 0.5342, + "step": 8646 + }, + { + "epoch": 0.7167012018234563, + "grad_norm": 0.2510151267051697, + "learning_rate": 1.4169084127641941e-06, + "loss": 0.5088, + "step": 8647 + }, + { + "epoch": 0.7167840861997513, + "grad_norm": 0.26431557536125183, + "learning_rate": 1.4164939908827188e-06, + "loss": 0.5104, + "step": 8648 + }, + { + "epoch": 0.7168669705760464, + "grad_norm": 0.23203791677951813, + "learning_rate": 1.4160795690012432e-06, + "loss": 0.482, + "step": 8649 + }, + { + "epoch": 0.7169498549523415, + "grad_norm": 0.24355606734752655, + "learning_rate": 1.415665147119768e-06, + "loss": 0.5053, + "step": 8650 + }, + { + "epoch": 0.7170327393286365, + "grad_norm": 0.2431698590517044, + "learning_rate": 1.4152507252382927e-06, + "loss": 0.5059, + "step": 8651 + }, + { + "epoch": 0.7171156237049316, + "grad_norm": 0.23764149844646454, + "learning_rate": 1.4148363033568174e-06, + "loss": 0.4648, + "step": 8652 + }, + { + "epoch": 0.7171985080812266, + "grad_norm": 0.23791177570819855, + "learning_rate": 1.4144218814753422e-06, + "loss": 0.5017, + "step": 8653 + }, + { + "epoch": 0.7172813924575218, + "grad_norm": 0.2303919643163681, + "learning_rate": 1.4140074595938666e-06, + "loss": 0.4725, + "step": 8654 + }, + { + "epoch": 0.7173642768338169, + "grad_norm": 0.24475106596946716, + "learning_rate": 1.4135930377123913e-06, + "loss": 0.4774, + "step": 8655 + }, + { + "epoch": 0.7174471612101119, + "grad_norm": 0.250332772731781, + "learning_rate": 1.413178615830916e-06, + "loss": 0.4737, + "step": 8656 + }, + { + "epoch": 0.717530045586407, + "grad_norm": 0.23375676572322845, + "learning_rate": 1.4127641939494408e-06, + "loss": 0.476, + "step": 8657 + }, + { + "epoch": 0.717612929962702, + "grad_norm": 0.2346441149711609, + "learning_rate": 1.4123497720679652e-06, + "loss": 0.4755, + "step": 8658 + }, + { + "epoch": 0.7176958143389971, + "grad_norm": 0.294277548789978, + "learning_rate": 1.4119353501864899e-06, + "loss": 0.4983, + "step": 8659 + }, + { + "epoch": 0.7177786987152922, + "grad_norm": 0.2231309562921524, + "learning_rate": 1.4115209283050147e-06, + "loss": 0.4612, + "step": 8660 + }, + { + "epoch": 0.7178615830915872, + "grad_norm": 0.24761752784252167, + "learning_rate": 1.4111065064235394e-06, + "loss": 0.5439, + "step": 8661 + }, + { + "epoch": 0.7179444674678823, + "grad_norm": 0.2590859830379486, + "learning_rate": 1.4106920845420638e-06, + "loss": 0.5245, + "step": 8662 + }, + { + "epoch": 0.7180273518441773, + "grad_norm": 0.24187147617340088, + "learning_rate": 1.4102776626605884e-06, + "loss": 0.4926, + "step": 8663 + }, + { + "epoch": 0.7181102362204724, + "grad_norm": 0.2436944991350174, + "learning_rate": 1.4098632407791133e-06, + "loss": 0.4944, + "step": 8664 + }, + { + "epoch": 0.7181931205967675, + "grad_norm": 0.2413419485092163, + "learning_rate": 1.409448818897638e-06, + "loss": 0.4938, + "step": 8665 + }, + { + "epoch": 0.7182760049730625, + "grad_norm": 0.2365182638168335, + "learning_rate": 1.4090343970161626e-06, + "loss": 0.4852, + "step": 8666 + }, + { + "epoch": 0.7183588893493577, + "grad_norm": 0.26264292001724243, + "learning_rate": 1.4086199751346872e-06, + "loss": 0.5785, + "step": 8667 + }, + { + "epoch": 0.7184417737256528, + "grad_norm": 0.2514819800853729, + "learning_rate": 1.4082055532532119e-06, + "loss": 0.5148, + "step": 8668 + }, + { + "epoch": 0.7185246581019478, + "grad_norm": 0.2271403819322586, + "learning_rate": 1.4077911313717365e-06, + "loss": 0.4689, + "step": 8669 + }, + { + "epoch": 0.7186075424782429, + "grad_norm": 0.24428170919418335, + "learning_rate": 1.4073767094902614e-06, + "loss": 0.4939, + "step": 8670 + }, + { + "epoch": 0.7186904268545379, + "grad_norm": 0.22111356258392334, + "learning_rate": 1.4069622876087858e-06, + "loss": 0.4823, + "step": 8671 + }, + { + "epoch": 0.718773311230833, + "grad_norm": 0.2432151734828949, + "learning_rate": 1.4065478657273104e-06, + "loss": 0.5137, + "step": 8672 + }, + { + "epoch": 0.7188561956071281, + "grad_norm": 0.24187307059764862, + "learning_rate": 1.406133443845835e-06, + "loss": 0.5269, + "step": 8673 + }, + { + "epoch": 0.7189390799834231, + "grad_norm": 0.2384885549545288, + "learning_rate": 1.40571902196436e-06, + "loss": 0.516, + "step": 8674 + }, + { + "epoch": 0.7190219643597182, + "grad_norm": 0.21536117792129517, + "learning_rate": 1.4053046000828846e-06, + "loss": 0.46, + "step": 8675 + }, + { + "epoch": 0.7191048487360132, + "grad_norm": 0.24453186988830566, + "learning_rate": 1.404890178201409e-06, + "loss": 0.5097, + "step": 8676 + }, + { + "epoch": 0.7191877331123083, + "grad_norm": 0.2370661050081253, + "learning_rate": 1.4044757563199339e-06, + "loss": 0.4866, + "step": 8677 + }, + { + "epoch": 0.7192706174886034, + "grad_norm": 0.2487431764602661, + "learning_rate": 1.4040613344384585e-06, + "loss": 0.4784, + "step": 8678 + }, + { + "epoch": 0.7193535018648984, + "grad_norm": 0.25238052010536194, + "learning_rate": 1.4036469125569832e-06, + "loss": 0.5176, + "step": 8679 + }, + { + "epoch": 0.7194363862411935, + "grad_norm": 0.27273181080818176, + "learning_rate": 1.4032324906755076e-06, + "loss": 0.5264, + "step": 8680 + }, + { + "epoch": 0.7195192706174885, + "grad_norm": 0.23666752874851227, + "learning_rate": 1.4028180687940324e-06, + "loss": 0.5073, + "step": 8681 + }, + { + "epoch": 0.7196021549937837, + "grad_norm": 0.25552964210510254, + "learning_rate": 1.402403646912557e-06, + "loss": 0.4971, + "step": 8682 + }, + { + "epoch": 0.7196850393700788, + "grad_norm": 0.2497948855161667, + "learning_rate": 1.4019892250310817e-06, + "loss": 0.5679, + "step": 8683 + }, + { + "epoch": 0.7197679237463738, + "grad_norm": 0.2529717981815338, + "learning_rate": 1.4015748031496066e-06, + "loss": 0.4904, + "step": 8684 + }, + { + "epoch": 0.7198508081226689, + "grad_norm": 0.25152015686035156, + "learning_rate": 1.401160381268131e-06, + "loss": 0.5339, + "step": 8685 + }, + { + "epoch": 0.719933692498964, + "grad_norm": 0.23127183318138123, + "learning_rate": 1.4007459593866557e-06, + "loss": 0.457, + "step": 8686 + }, + { + "epoch": 0.720016576875259, + "grad_norm": 0.23667961359024048, + "learning_rate": 1.4003315375051805e-06, + "loss": 0.4786, + "step": 8687 + }, + { + "epoch": 0.7200994612515541, + "grad_norm": 0.25520822405815125, + "learning_rate": 1.3999171156237052e-06, + "loss": 0.5104, + "step": 8688 + }, + { + "epoch": 0.7201823456278491, + "grad_norm": 0.24367137253284454, + "learning_rate": 1.3995026937422296e-06, + "loss": 0.5108, + "step": 8689 + }, + { + "epoch": 0.7202652300041442, + "grad_norm": 0.2518264949321747, + "learning_rate": 1.3990882718607542e-06, + "loss": 0.5043, + "step": 8690 + }, + { + "epoch": 0.7203481143804393, + "grad_norm": 0.24926787614822388, + "learning_rate": 1.398673849979279e-06, + "loss": 0.4886, + "step": 8691 + }, + { + "epoch": 0.7204309987567343, + "grad_norm": 0.2477668970823288, + "learning_rate": 1.3982594280978037e-06, + "loss": 0.517, + "step": 8692 + }, + { + "epoch": 0.7205138831330294, + "grad_norm": 0.24163593351840973, + "learning_rate": 1.3978450062163284e-06, + "loss": 0.48, + "step": 8693 + }, + { + "epoch": 0.7205967675093244, + "grad_norm": 0.23968781530857086, + "learning_rate": 1.3974305843348528e-06, + "loss": 0.4832, + "step": 8694 + }, + { + "epoch": 0.7206796518856196, + "grad_norm": 0.2842702269554138, + "learning_rate": 1.3970161624533777e-06, + "loss": 0.5193, + "step": 8695 + }, + { + "epoch": 0.7207625362619147, + "grad_norm": 0.24437043070793152, + "learning_rate": 1.3966017405719023e-06, + "loss": 0.5357, + "step": 8696 + }, + { + "epoch": 0.7208454206382097, + "grad_norm": 0.25950172543525696, + "learning_rate": 1.3961873186904272e-06, + "loss": 0.5068, + "step": 8697 + }, + { + "epoch": 0.7209283050145048, + "grad_norm": 0.23638342320919037, + "learning_rate": 1.3957728968089516e-06, + "loss": 0.5016, + "step": 8698 + }, + { + "epoch": 0.7210111893907999, + "grad_norm": 0.23324614763259888, + "learning_rate": 1.3953584749274762e-06, + "loss": 0.4904, + "step": 8699 + }, + { + "epoch": 0.7210940737670949, + "grad_norm": 0.2316834181547165, + "learning_rate": 1.3949440530460009e-06, + "loss": 0.5101, + "step": 8700 + }, + { + "epoch": 0.72117695814339, + "grad_norm": 0.23823389410972595, + "learning_rate": 1.3945296311645257e-06, + "loss": 0.5024, + "step": 8701 + }, + { + "epoch": 0.721259842519685, + "grad_norm": 0.21971191465854645, + "learning_rate": 1.3941152092830504e-06, + "loss": 0.4345, + "step": 8702 + }, + { + "epoch": 0.7213427268959801, + "grad_norm": 0.25077566504478455, + "learning_rate": 1.3937007874015748e-06, + "loss": 0.4551, + "step": 8703 + }, + { + "epoch": 0.7214256112722752, + "grad_norm": 0.24162372946739197, + "learning_rate": 1.3932863655200995e-06, + "loss": 0.5186, + "step": 8704 + }, + { + "epoch": 0.7215084956485702, + "grad_norm": 0.22907458245754242, + "learning_rate": 1.3928719436386243e-06, + "loss": 0.4861, + "step": 8705 + }, + { + "epoch": 0.7215913800248653, + "grad_norm": 0.2559436857700348, + "learning_rate": 1.392457521757149e-06, + "loss": 0.5017, + "step": 8706 + }, + { + "epoch": 0.7216742644011603, + "grad_norm": 0.24453973770141602, + "learning_rate": 1.3920430998756734e-06, + "loss": 0.4655, + "step": 8707 + }, + { + "epoch": 0.7217571487774554, + "grad_norm": 0.3490285873413086, + "learning_rate": 1.3916286779941982e-06, + "loss": 0.5555, + "step": 8708 + }, + { + "epoch": 0.7218400331537506, + "grad_norm": 0.22958436608314514, + "learning_rate": 1.3912142561127229e-06, + "loss": 0.446, + "step": 8709 + }, + { + "epoch": 0.7219229175300456, + "grad_norm": 0.2533676326274872, + "learning_rate": 1.3907998342312475e-06, + "loss": 0.509, + "step": 8710 + }, + { + "epoch": 0.7220058019063407, + "grad_norm": 0.2172214239835739, + "learning_rate": 1.3903854123497724e-06, + "loss": 0.4883, + "step": 8711 + }, + { + "epoch": 0.7220886862826357, + "grad_norm": 0.24694694578647614, + "learning_rate": 1.3899709904682968e-06, + "loss": 0.4525, + "step": 8712 + }, + { + "epoch": 0.7221715706589308, + "grad_norm": 0.2417561113834381, + "learning_rate": 1.3895565685868215e-06, + "loss": 0.4687, + "step": 8713 + }, + { + "epoch": 0.7222544550352259, + "grad_norm": 0.25511234998703003, + "learning_rate": 1.3891421467053461e-06, + "loss": 0.4945, + "step": 8714 + }, + { + "epoch": 0.7223373394115209, + "grad_norm": 0.25932708382606506, + "learning_rate": 1.388727724823871e-06, + "loss": 0.4933, + "step": 8715 + }, + { + "epoch": 0.722420223787816, + "grad_norm": 0.25038328766822815, + "learning_rate": 1.3883133029423954e-06, + "loss": 0.4985, + "step": 8716 + }, + { + "epoch": 0.7225031081641111, + "grad_norm": 0.2433251142501831, + "learning_rate": 1.38789888106092e-06, + "loss": 0.4962, + "step": 8717 + }, + { + "epoch": 0.7225859925404061, + "grad_norm": 0.2506362199783325, + "learning_rate": 1.3874844591794449e-06, + "loss": 0.5146, + "step": 8718 + }, + { + "epoch": 0.7226688769167012, + "grad_norm": 0.23364800214767456, + "learning_rate": 1.3870700372979695e-06, + "loss": 0.4984, + "step": 8719 + }, + { + "epoch": 0.7227517612929962, + "grad_norm": 0.2383718192577362, + "learning_rate": 1.3866556154164942e-06, + "loss": 0.496, + "step": 8720 + }, + { + "epoch": 0.7228346456692913, + "grad_norm": 0.2550661265850067, + "learning_rate": 1.3862411935350186e-06, + "loss": 0.5255, + "step": 8721 + }, + { + "epoch": 0.7229175300455865, + "grad_norm": 0.24547472596168518, + "learning_rate": 1.3858267716535435e-06, + "loss": 0.506, + "step": 8722 + }, + { + "epoch": 0.7230004144218815, + "grad_norm": 0.26070404052734375, + "learning_rate": 1.3854123497720681e-06, + "loss": 0.4877, + "step": 8723 + }, + { + "epoch": 0.7230832987981766, + "grad_norm": 0.23242610692977905, + "learning_rate": 1.3849979278905928e-06, + "loss": 0.5179, + "step": 8724 + }, + { + "epoch": 0.7231661831744716, + "grad_norm": 0.26508885622024536, + "learning_rate": 1.3845835060091172e-06, + "loss": 0.561, + "step": 8725 + }, + { + "epoch": 0.7232490675507667, + "grad_norm": 0.26122036576271057, + "learning_rate": 1.384169084127642e-06, + "loss": 0.5421, + "step": 8726 + }, + { + "epoch": 0.7233319519270618, + "grad_norm": 0.2558877468109131, + "learning_rate": 1.3837546622461667e-06, + "loss": 0.4771, + "step": 8727 + }, + { + "epoch": 0.7234148363033568, + "grad_norm": 0.23310914635658264, + "learning_rate": 1.3833402403646915e-06, + "loss": 0.4806, + "step": 8728 + }, + { + "epoch": 0.7234977206796519, + "grad_norm": 0.2529543340206146, + "learning_rate": 1.3829258184832162e-06, + "loss": 0.4962, + "step": 8729 + }, + { + "epoch": 0.7235806050559469, + "grad_norm": 0.2422959953546524, + "learning_rate": 1.3825113966017406e-06, + "loss": 0.5059, + "step": 8730 + }, + { + "epoch": 0.723663489432242, + "grad_norm": 0.21086810529232025, + "learning_rate": 1.3820969747202653e-06, + "loss": 0.4475, + "step": 8731 + }, + { + "epoch": 0.7237463738085371, + "grad_norm": 0.24581028521060944, + "learning_rate": 1.3816825528387901e-06, + "loss": 0.4874, + "step": 8732 + }, + { + "epoch": 0.7238292581848321, + "grad_norm": 0.23343153297901154, + "learning_rate": 1.3812681309573148e-06, + "loss": 0.4875, + "step": 8733 + }, + { + "epoch": 0.7239121425611272, + "grad_norm": 0.2652910053730011, + "learning_rate": 1.3808537090758392e-06, + "loss": 0.552, + "step": 8734 + }, + { + "epoch": 0.7239950269374223, + "grad_norm": 0.26380258798599243, + "learning_rate": 1.3804392871943638e-06, + "loss": 0.5149, + "step": 8735 + }, + { + "epoch": 0.7240779113137173, + "grad_norm": 0.2567080557346344, + "learning_rate": 1.3800248653128887e-06, + "loss": 0.5314, + "step": 8736 + }, + { + "epoch": 0.7241607956900125, + "grad_norm": 0.24573467671871185, + "learning_rate": 1.3796104434314133e-06, + "loss": 0.5359, + "step": 8737 + }, + { + "epoch": 0.7242436800663075, + "grad_norm": 0.24658076465129852, + "learning_rate": 1.3791960215499378e-06, + "loss": 0.4725, + "step": 8738 + }, + { + "epoch": 0.7243265644426026, + "grad_norm": 0.24195773899555206, + "learning_rate": 1.3787815996684626e-06, + "loss": 0.4357, + "step": 8739 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 0.21866434812545776, + "learning_rate": 1.3783671777869873e-06, + "loss": 0.4601, + "step": 8740 + }, + { + "epoch": 0.7244923331951927, + "grad_norm": 0.24174395203590393, + "learning_rate": 1.377952755905512e-06, + "loss": 0.4993, + "step": 8741 + }, + { + "epoch": 0.7245752175714878, + "grad_norm": 0.25888800621032715, + "learning_rate": 1.3775383340240368e-06, + "loss": 0.5033, + "step": 8742 + }, + { + "epoch": 0.7246581019477828, + "grad_norm": 0.23512357473373413, + "learning_rate": 1.3771239121425612e-06, + "loss": 0.5062, + "step": 8743 + }, + { + "epoch": 0.7247409863240779, + "grad_norm": 0.25838345289230347, + "learning_rate": 1.3767094902610858e-06, + "loss": 0.5031, + "step": 8744 + }, + { + "epoch": 0.724823870700373, + "grad_norm": 0.24652142822742462, + "learning_rate": 1.3762950683796105e-06, + "loss": 0.4803, + "step": 8745 + }, + { + "epoch": 0.724906755076668, + "grad_norm": 0.2634184658527374, + "learning_rate": 1.3758806464981353e-06, + "loss": 0.5439, + "step": 8746 + }, + { + "epoch": 0.7249896394529631, + "grad_norm": 0.25008806586265564, + "learning_rate": 1.3754662246166598e-06, + "loss": 0.4826, + "step": 8747 + }, + { + "epoch": 0.7250725238292581, + "grad_norm": 0.22647593915462494, + "learning_rate": 1.3750518027351844e-06, + "loss": 0.4723, + "step": 8748 + }, + { + "epoch": 0.7251554082055532, + "grad_norm": 0.2417553812265396, + "learning_rate": 1.3746373808537093e-06, + "loss": 0.5115, + "step": 8749 + }, + { + "epoch": 0.7252382925818484, + "grad_norm": 0.23567284643650055, + "learning_rate": 1.374222958972234e-06, + "loss": 0.4854, + "step": 8750 + }, + { + "epoch": 0.7253211769581434, + "grad_norm": 0.24211673438549042, + "learning_rate": 1.3738085370907586e-06, + "loss": 0.5121, + "step": 8751 + }, + { + "epoch": 0.7254040613344385, + "grad_norm": 0.23898205161094666, + "learning_rate": 1.373394115209283e-06, + "loss": 0.4843, + "step": 8752 + }, + { + "epoch": 0.7254869457107336, + "grad_norm": 0.23132379353046417, + "learning_rate": 1.3729796933278078e-06, + "loss": 0.4509, + "step": 8753 + }, + { + "epoch": 0.7255698300870286, + "grad_norm": 0.25949496030807495, + "learning_rate": 1.3725652714463325e-06, + "loss": 0.5317, + "step": 8754 + }, + { + "epoch": 0.7256527144633237, + "grad_norm": 0.26998820900917053, + "learning_rate": 1.3721508495648571e-06, + "loss": 0.5141, + "step": 8755 + }, + { + "epoch": 0.7257355988396187, + "grad_norm": 0.25158393383026123, + "learning_rate": 1.3717364276833818e-06, + "loss": 0.5022, + "step": 8756 + }, + { + "epoch": 0.7258184832159138, + "grad_norm": 0.27314436435699463, + "learning_rate": 1.3713220058019064e-06, + "loss": 0.5328, + "step": 8757 + }, + { + "epoch": 0.7259013675922089, + "grad_norm": 0.2311522662639618, + "learning_rate": 1.370907583920431e-06, + "loss": 0.4783, + "step": 8758 + }, + { + "epoch": 0.7259842519685039, + "grad_norm": 0.24010112881660461, + "learning_rate": 1.370493162038956e-06, + "loss": 0.491, + "step": 8759 + }, + { + "epoch": 0.726067136344799, + "grad_norm": 0.23552319407463074, + "learning_rate": 1.3700787401574806e-06, + "loss": 0.4952, + "step": 8760 + }, + { + "epoch": 0.726150020721094, + "grad_norm": 0.23590174317359924, + "learning_rate": 1.369664318276005e-06, + "loss": 0.505, + "step": 8761 + }, + { + "epoch": 0.7262329050973891, + "grad_norm": 0.271854966878891, + "learning_rate": 1.3692498963945296e-06, + "loss": 0.4655, + "step": 8762 + }, + { + "epoch": 0.7263157894736842, + "grad_norm": 0.24574267864227295, + "learning_rate": 1.3688354745130545e-06, + "loss": 0.5284, + "step": 8763 + }, + { + "epoch": 0.7263986738499792, + "grad_norm": 0.24071015417575836, + "learning_rate": 1.3684210526315791e-06, + "loss": 0.4852, + "step": 8764 + }, + { + "epoch": 0.7264815582262744, + "grad_norm": 0.25183457136154175, + "learning_rate": 1.3680066307501036e-06, + "loss": 0.4975, + "step": 8765 + }, + { + "epoch": 0.7265644426025694, + "grad_norm": 0.253787100315094, + "learning_rate": 1.3675922088686284e-06, + "loss": 0.5475, + "step": 8766 + }, + { + "epoch": 0.7266473269788645, + "grad_norm": 0.22197744250297546, + "learning_rate": 1.367177786987153e-06, + "loss": 0.4321, + "step": 8767 + }, + { + "epoch": 0.7267302113551596, + "grad_norm": 0.2514803111553192, + "learning_rate": 1.3667633651056777e-06, + "loss": 0.5141, + "step": 8768 + }, + { + "epoch": 0.7268130957314546, + "grad_norm": 0.24885106086730957, + "learning_rate": 1.3663489432242026e-06, + "loss": 0.5156, + "step": 8769 + }, + { + "epoch": 0.7268959801077497, + "grad_norm": 0.26322489976882935, + "learning_rate": 1.365934521342727e-06, + "loss": 0.5623, + "step": 8770 + }, + { + "epoch": 0.7269788644840448, + "grad_norm": 0.23137712478637695, + "learning_rate": 1.3655200994612516e-06, + "loss": 0.5011, + "step": 8771 + }, + { + "epoch": 0.7270617488603398, + "grad_norm": 0.2460624724626541, + "learning_rate": 1.3651056775797763e-06, + "loss": 0.5175, + "step": 8772 + }, + { + "epoch": 0.7271446332366349, + "grad_norm": 0.22291354835033417, + "learning_rate": 1.3646912556983011e-06, + "loss": 0.4685, + "step": 8773 + }, + { + "epoch": 0.7272275176129299, + "grad_norm": 0.24373279511928558, + "learning_rate": 1.3642768338168256e-06, + "loss": 0.4795, + "step": 8774 + }, + { + "epoch": 0.727310401989225, + "grad_norm": 0.24007532000541687, + "learning_rate": 1.3638624119353502e-06, + "loss": 0.4682, + "step": 8775 + }, + { + "epoch": 0.7273932863655201, + "grad_norm": 0.23031944036483765, + "learning_rate": 1.3634479900538749e-06, + "loss": 0.4743, + "step": 8776 + }, + { + "epoch": 0.7274761707418151, + "grad_norm": 0.22879743576049805, + "learning_rate": 1.3630335681723997e-06, + "loss": 0.4855, + "step": 8777 + }, + { + "epoch": 0.7275590551181103, + "grad_norm": 0.24998028576374054, + "learning_rate": 1.3626191462909244e-06, + "loss": 0.5203, + "step": 8778 + }, + { + "epoch": 0.7276419394944053, + "grad_norm": 0.2510838210582733, + "learning_rate": 1.3622047244094488e-06, + "loss": 0.5097, + "step": 8779 + }, + { + "epoch": 0.7277248238707004, + "grad_norm": 0.2359093874692917, + "learning_rate": 1.3617903025279736e-06, + "loss": 0.4894, + "step": 8780 + }, + { + "epoch": 0.7278077082469955, + "grad_norm": 0.23702828586101532, + "learning_rate": 1.3613758806464983e-06, + "loss": 0.4837, + "step": 8781 + }, + { + "epoch": 0.7278905926232905, + "grad_norm": 0.2525924742221832, + "learning_rate": 1.360961458765023e-06, + "loss": 0.5357, + "step": 8782 + }, + { + "epoch": 0.7279734769995856, + "grad_norm": 0.24305996298789978, + "learning_rate": 1.3605470368835474e-06, + "loss": 0.5243, + "step": 8783 + }, + { + "epoch": 0.7280563613758807, + "grad_norm": 0.25283390283584595, + "learning_rate": 1.3601326150020722e-06, + "loss": 0.5327, + "step": 8784 + }, + { + "epoch": 0.7281392457521757, + "grad_norm": 0.25532832741737366, + "learning_rate": 1.3597181931205969e-06, + "loss": 0.4944, + "step": 8785 + }, + { + "epoch": 0.7282221301284708, + "grad_norm": 0.2366204410791397, + "learning_rate": 1.3593037712391215e-06, + "loss": 0.463, + "step": 8786 + }, + { + "epoch": 0.7283050145047658, + "grad_norm": 0.24010741710662842, + "learning_rate": 1.3588893493576464e-06, + "loss": 0.4721, + "step": 8787 + }, + { + "epoch": 0.7283878988810609, + "grad_norm": 0.24602001905441284, + "learning_rate": 1.3584749274761708e-06, + "loss": 0.5381, + "step": 8788 + }, + { + "epoch": 0.728470783257356, + "grad_norm": 0.24733005464076996, + "learning_rate": 1.3580605055946954e-06, + "loss": 0.5072, + "step": 8789 + }, + { + "epoch": 0.728553667633651, + "grad_norm": 0.23640474677085876, + "learning_rate": 1.3576460837132203e-06, + "loss": 0.4669, + "step": 8790 + }, + { + "epoch": 0.7286365520099461, + "grad_norm": 0.23577895760536194, + "learning_rate": 1.357231661831745e-06, + "loss": 0.4788, + "step": 8791 + }, + { + "epoch": 0.7287194363862411, + "grad_norm": 0.22573216259479523, + "learning_rate": 1.3568172399502694e-06, + "loss": 0.455, + "step": 8792 + }, + { + "epoch": 0.7288023207625363, + "grad_norm": 0.2237868756055832, + "learning_rate": 1.356402818068794e-06, + "loss": 0.4875, + "step": 8793 + }, + { + "epoch": 0.7288852051388314, + "grad_norm": 0.24943694472312927, + "learning_rate": 1.3559883961873189e-06, + "loss": 0.4854, + "step": 8794 + }, + { + "epoch": 0.7289680895151264, + "grad_norm": 0.2588233947753906, + "learning_rate": 1.3555739743058435e-06, + "loss": 0.5095, + "step": 8795 + }, + { + "epoch": 0.7290509738914215, + "grad_norm": 0.23750101029872894, + "learning_rate": 1.3551595524243681e-06, + "loss": 0.4848, + "step": 8796 + }, + { + "epoch": 0.7291338582677165, + "grad_norm": 0.2632625997066498, + "learning_rate": 1.3547451305428928e-06, + "loss": 0.5024, + "step": 8797 + }, + { + "epoch": 0.7292167426440116, + "grad_norm": 0.2108021378517151, + "learning_rate": 1.3543307086614174e-06, + "loss": 0.4397, + "step": 8798 + }, + { + "epoch": 0.7292996270203067, + "grad_norm": 0.2354714274406433, + "learning_rate": 1.353916286779942e-06, + "loss": 0.4516, + "step": 8799 + }, + { + "epoch": 0.7293825113966017, + "grad_norm": 0.2579306960105896, + "learning_rate": 1.353501864898467e-06, + "loss": 0.5158, + "step": 8800 + }, + { + "epoch": 0.7294653957728968, + "grad_norm": 0.22710740566253662, + "learning_rate": 1.3530874430169914e-06, + "loss": 0.4443, + "step": 8801 + }, + { + "epoch": 0.7295482801491919, + "grad_norm": 0.2849856913089752, + "learning_rate": 1.352673021135516e-06, + "loss": 0.5316, + "step": 8802 + }, + { + "epoch": 0.7296311645254869, + "grad_norm": 0.25104543566703796, + "learning_rate": 1.3522585992540407e-06, + "loss": 0.5102, + "step": 8803 + }, + { + "epoch": 0.729714048901782, + "grad_norm": 0.2567070722579956, + "learning_rate": 1.3518441773725655e-06, + "loss": 0.5098, + "step": 8804 + }, + { + "epoch": 0.729796933278077, + "grad_norm": 0.2468426376581192, + "learning_rate": 1.3514297554910902e-06, + "loss": 0.5049, + "step": 8805 + }, + { + "epoch": 0.7298798176543722, + "grad_norm": 0.25078022480010986, + "learning_rate": 1.3510153336096146e-06, + "loss": 0.4601, + "step": 8806 + }, + { + "epoch": 0.7299627020306673, + "grad_norm": 0.24569517374038696, + "learning_rate": 1.3506009117281394e-06, + "loss": 0.5117, + "step": 8807 + }, + { + "epoch": 0.7300455864069623, + "grad_norm": 0.24365082383155823, + "learning_rate": 1.350186489846664e-06, + "loss": 0.468, + "step": 8808 + }, + { + "epoch": 0.7301284707832574, + "grad_norm": 0.26006367802619934, + "learning_rate": 1.3497720679651887e-06, + "loss": 0.519, + "step": 8809 + }, + { + "epoch": 0.7302113551595524, + "grad_norm": 0.24880172312259674, + "learning_rate": 1.3493576460837132e-06, + "loss": 0.5227, + "step": 8810 + }, + { + "epoch": 0.7302942395358475, + "grad_norm": 0.23972131311893463, + "learning_rate": 1.348943224202238e-06, + "loss": 0.5078, + "step": 8811 + }, + { + "epoch": 0.7303771239121426, + "grad_norm": 0.2375386506319046, + "learning_rate": 1.3485288023207627e-06, + "loss": 0.5242, + "step": 8812 + }, + { + "epoch": 0.7304600082884376, + "grad_norm": 0.23936019837856293, + "learning_rate": 1.3481143804392873e-06, + "loss": 0.5209, + "step": 8813 + }, + { + "epoch": 0.7305428926647327, + "grad_norm": 0.2428368777036667, + "learning_rate": 1.3476999585578117e-06, + "loss": 0.4949, + "step": 8814 + }, + { + "epoch": 0.7306257770410277, + "grad_norm": 0.3943721354007721, + "learning_rate": 1.3472855366763366e-06, + "loss": 0.4525, + "step": 8815 + }, + { + "epoch": 0.7307086614173228, + "grad_norm": 0.25262904167175293, + "learning_rate": 1.3468711147948612e-06, + "loss": 0.5095, + "step": 8816 + }, + { + "epoch": 0.7307915457936179, + "grad_norm": 0.23632606863975525, + "learning_rate": 1.346456692913386e-06, + "loss": 0.5078, + "step": 8817 + }, + { + "epoch": 0.7308744301699129, + "grad_norm": 0.2657429277896881, + "learning_rate": 1.3460422710319107e-06, + "loss": 0.5138, + "step": 8818 + }, + { + "epoch": 0.730957314546208, + "grad_norm": 0.22477862238883972, + "learning_rate": 1.3456278491504352e-06, + "loss": 0.4982, + "step": 8819 + }, + { + "epoch": 0.7310401989225032, + "grad_norm": 0.24090281128883362, + "learning_rate": 1.3452134272689598e-06, + "loss": 0.4878, + "step": 8820 + }, + { + "epoch": 0.7311230832987982, + "grad_norm": 0.23899467289447784, + "learning_rate": 1.3447990053874847e-06, + "loss": 0.4833, + "step": 8821 + }, + { + "epoch": 0.7312059676750933, + "grad_norm": 0.2389904260635376, + "learning_rate": 1.3443845835060093e-06, + "loss": 0.514, + "step": 8822 + }, + { + "epoch": 0.7312888520513883, + "grad_norm": 0.237071231007576, + "learning_rate": 1.3439701616245337e-06, + "loss": 0.5036, + "step": 8823 + }, + { + "epoch": 0.7313717364276834, + "grad_norm": 0.23345807194709778, + "learning_rate": 1.3435557397430584e-06, + "loss": 0.507, + "step": 8824 + }, + { + "epoch": 0.7314546208039785, + "grad_norm": 0.2533978819847107, + "learning_rate": 1.3431413178615832e-06, + "loss": 0.4877, + "step": 8825 + }, + { + "epoch": 0.7315375051802735, + "grad_norm": 0.23954494297504425, + "learning_rate": 1.3427268959801079e-06, + "loss": 0.5009, + "step": 8826 + }, + { + "epoch": 0.7316203895565686, + "grad_norm": 0.22912606596946716, + "learning_rate": 1.3423124740986327e-06, + "loss": 0.4786, + "step": 8827 + }, + { + "epoch": 0.7317032739328636, + "grad_norm": 0.25828057527542114, + "learning_rate": 1.3418980522171572e-06, + "loss": 0.5172, + "step": 8828 + }, + { + "epoch": 0.7317861583091587, + "grad_norm": 0.2533746063709259, + "learning_rate": 1.3414836303356818e-06, + "loss": 0.5497, + "step": 8829 + }, + { + "epoch": 0.7318690426854538, + "grad_norm": 0.23668748140335083, + "learning_rate": 1.3410692084542065e-06, + "loss": 0.4734, + "step": 8830 + }, + { + "epoch": 0.7319519270617488, + "grad_norm": 0.24152179062366486, + "learning_rate": 1.3406547865727313e-06, + "loss": 0.5171, + "step": 8831 + }, + { + "epoch": 0.7320348114380439, + "grad_norm": 0.2368352860212326, + "learning_rate": 1.3402403646912557e-06, + "loss": 0.4707, + "step": 8832 + }, + { + "epoch": 0.7321176958143389, + "grad_norm": 0.25087764859199524, + "learning_rate": 1.3398259428097804e-06, + "loss": 0.5276, + "step": 8833 + }, + { + "epoch": 0.732200580190634, + "grad_norm": 0.23859284818172455, + "learning_rate": 1.339411520928305e-06, + "loss": 0.511, + "step": 8834 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 0.25391310453414917, + "learning_rate": 1.3389970990468299e-06, + "loss": 0.5005, + "step": 8835 + }, + { + "epoch": 0.7323663489432242, + "grad_norm": 0.2379051148891449, + "learning_rate": 1.3385826771653545e-06, + "loss": 0.4983, + "step": 8836 + }, + { + "epoch": 0.7324492333195193, + "grad_norm": 0.24992942810058594, + "learning_rate": 1.338168255283879e-06, + "loss": 0.5077, + "step": 8837 + }, + { + "epoch": 0.7325321176958144, + "grad_norm": 0.23895709216594696, + "learning_rate": 1.3377538334024038e-06, + "loss": 0.5289, + "step": 8838 + }, + { + "epoch": 0.7326150020721094, + "grad_norm": 0.2508300244808197, + "learning_rate": 1.3373394115209285e-06, + "loss": 0.5468, + "step": 8839 + }, + { + "epoch": 0.7326978864484045, + "grad_norm": 0.24584732949733734, + "learning_rate": 1.336924989639453e-06, + "loss": 0.4612, + "step": 8840 + }, + { + "epoch": 0.7327807708246995, + "grad_norm": 0.2585107386112213, + "learning_rate": 1.3365105677579775e-06, + "loss": 0.4563, + "step": 8841 + }, + { + "epoch": 0.7328636552009946, + "grad_norm": 0.2477787286043167, + "learning_rate": 1.3360961458765024e-06, + "loss": 0.5436, + "step": 8842 + }, + { + "epoch": 0.7329465395772897, + "grad_norm": 0.25132477283477783, + "learning_rate": 1.335681723995027e-06, + "loss": 0.5306, + "step": 8843 + }, + { + "epoch": 0.7330294239535847, + "grad_norm": 0.2569427788257599, + "learning_rate": 1.3352673021135517e-06, + "loss": 0.4974, + "step": 8844 + }, + { + "epoch": 0.7331123083298798, + "grad_norm": 0.25904303789138794, + "learning_rate": 1.3348528802320765e-06, + "loss": 0.5134, + "step": 8845 + }, + { + "epoch": 0.7331951927061748, + "grad_norm": 0.24030375480651855, + "learning_rate": 1.334438458350601e-06, + "loss": 0.5228, + "step": 8846 + }, + { + "epoch": 0.73327807708247, + "grad_norm": 0.24731679260730743, + "learning_rate": 1.3340240364691256e-06, + "loss": 0.5128, + "step": 8847 + }, + { + "epoch": 0.7333609614587651, + "grad_norm": 0.24303966760635376, + "learning_rate": 1.3336096145876505e-06, + "loss": 0.5063, + "step": 8848 + }, + { + "epoch": 0.7334438458350601, + "grad_norm": 0.26844239234924316, + "learning_rate": 1.333195192706175e-06, + "loss": 0.5285, + "step": 8849 + }, + { + "epoch": 0.7335267302113552, + "grad_norm": 0.2248915731906891, + "learning_rate": 1.3327807708246995e-06, + "loss": 0.4431, + "step": 8850 + }, + { + "epoch": 0.7336096145876503, + "grad_norm": 0.24163562059402466, + "learning_rate": 1.3323663489432242e-06, + "loss": 0.4952, + "step": 8851 + }, + { + "epoch": 0.7336924989639453, + "grad_norm": 0.2271629273891449, + "learning_rate": 1.331951927061749e-06, + "loss": 0.4548, + "step": 8852 + }, + { + "epoch": 0.7337753833402404, + "grad_norm": 0.23227162659168243, + "learning_rate": 1.3315375051802737e-06, + "loss": 0.4998, + "step": 8853 + }, + { + "epoch": 0.7338582677165354, + "grad_norm": 0.2557642161846161, + "learning_rate": 1.3311230832987983e-06, + "loss": 0.5251, + "step": 8854 + }, + { + "epoch": 0.7339411520928305, + "grad_norm": 0.23281323909759521, + "learning_rate": 1.3307086614173228e-06, + "loss": 0.4818, + "step": 8855 + }, + { + "epoch": 0.7340240364691256, + "grad_norm": 0.2317448854446411, + "learning_rate": 1.3302942395358476e-06, + "loss": 0.4889, + "step": 8856 + }, + { + "epoch": 0.7341069208454206, + "grad_norm": 0.2410125434398651, + "learning_rate": 1.3298798176543723e-06, + "loss": 0.5114, + "step": 8857 + }, + { + "epoch": 0.7341898052217157, + "grad_norm": 0.24396269023418427, + "learning_rate": 1.3294653957728971e-06, + "loss": 0.502, + "step": 8858 + }, + { + "epoch": 0.7342726895980107, + "grad_norm": 0.2496923804283142, + "learning_rate": 1.3290509738914215e-06, + "loss": 0.4583, + "step": 8859 + }, + { + "epoch": 0.7343555739743058, + "grad_norm": 0.2483823299407959, + "learning_rate": 1.3286365520099462e-06, + "loss": 0.5332, + "step": 8860 + }, + { + "epoch": 0.734438458350601, + "grad_norm": 0.23357215523719788, + "learning_rate": 1.3282221301284708e-06, + "loss": 0.4922, + "step": 8861 + }, + { + "epoch": 0.734521342726896, + "grad_norm": 0.2413865178823471, + "learning_rate": 1.3278077082469957e-06, + "loss": 0.4676, + "step": 8862 + }, + { + "epoch": 0.7346042271031911, + "grad_norm": 0.22834433615207672, + "learning_rate": 1.3273932863655203e-06, + "loss": 0.4937, + "step": 8863 + }, + { + "epoch": 0.7346871114794861, + "grad_norm": 0.2578032910823822, + "learning_rate": 1.3269788644840448e-06, + "loss": 0.511, + "step": 8864 + }, + { + "epoch": 0.7347699958557812, + "grad_norm": 0.23028352856636047, + "learning_rate": 1.3265644426025694e-06, + "loss": 0.5073, + "step": 8865 + }, + { + "epoch": 0.7348528802320763, + "grad_norm": 0.30024072527885437, + "learning_rate": 1.3261500207210943e-06, + "loss": 0.4927, + "step": 8866 + }, + { + "epoch": 0.7349357646083713, + "grad_norm": 0.23485560715198517, + "learning_rate": 1.325735598839619e-06, + "loss": 0.5043, + "step": 8867 + }, + { + "epoch": 0.7350186489846664, + "grad_norm": 0.22290098667144775, + "learning_rate": 1.3253211769581433e-06, + "loss": 0.4674, + "step": 8868 + }, + { + "epoch": 0.7351015333609615, + "grad_norm": 0.24091051518917084, + "learning_rate": 1.3249067550766682e-06, + "loss": 0.4722, + "step": 8869 + }, + { + "epoch": 0.7351844177372565, + "grad_norm": 0.24653147161006927, + "learning_rate": 1.3244923331951928e-06, + "loss": 0.5004, + "step": 8870 + }, + { + "epoch": 0.7352673021135516, + "grad_norm": 0.2321300059556961, + "learning_rate": 1.3240779113137175e-06, + "loss": 0.477, + "step": 8871 + }, + { + "epoch": 0.7353501864898466, + "grad_norm": 0.23367860913276672, + "learning_rate": 1.3236634894322423e-06, + "loss": 0.4853, + "step": 8872 + }, + { + "epoch": 0.7354330708661417, + "grad_norm": 0.26250964403152466, + "learning_rate": 1.3232490675507668e-06, + "loss": 0.5069, + "step": 8873 + }, + { + "epoch": 0.7355159552424368, + "grad_norm": 0.25949540734291077, + "learning_rate": 1.3228346456692914e-06, + "loss": 0.5178, + "step": 8874 + }, + { + "epoch": 0.7355988396187318, + "grad_norm": 0.2511058747768402, + "learning_rate": 1.322420223787816e-06, + "loss": 0.5149, + "step": 8875 + }, + { + "epoch": 0.735681723995027, + "grad_norm": 0.24319173395633698, + "learning_rate": 1.322005801906341e-06, + "loss": 0.4928, + "step": 8876 + }, + { + "epoch": 0.735764608371322, + "grad_norm": 0.2726857364177704, + "learning_rate": 1.3215913800248653e-06, + "loss": 0.5157, + "step": 8877 + }, + { + "epoch": 0.7358474927476171, + "grad_norm": 0.24717044830322266, + "learning_rate": 1.32117695814339e-06, + "loss": 0.5064, + "step": 8878 + }, + { + "epoch": 0.7359303771239122, + "grad_norm": 0.25805824995040894, + "learning_rate": 1.3207625362619148e-06, + "loss": 0.4972, + "step": 8879 + }, + { + "epoch": 0.7360132615002072, + "grad_norm": 0.2473284900188446, + "learning_rate": 1.3203481143804395e-06, + "loss": 0.4766, + "step": 8880 + }, + { + "epoch": 0.7360961458765023, + "grad_norm": 0.23171748220920563, + "learning_rate": 1.3199336924989641e-06, + "loss": 0.4846, + "step": 8881 + }, + { + "epoch": 0.7361790302527973, + "grad_norm": 0.2522009313106537, + "learning_rate": 1.3195192706174886e-06, + "loss": 0.4924, + "step": 8882 + }, + { + "epoch": 0.7362619146290924, + "grad_norm": 0.2605189085006714, + "learning_rate": 1.3191048487360134e-06, + "loss": 0.5185, + "step": 8883 + }, + { + "epoch": 0.7363447990053875, + "grad_norm": 0.23036156594753265, + "learning_rate": 1.318690426854538e-06, + "loss": 0.4958, + "step": 8884 + }, + { + "epoch": 0.7364276833816825, + "grad_norm": 0.24141399562358856, + "learning_rate": 1.3182760049730627e-06, + "loss": 0.5166, + "step": 8885 + }, + { + "epoch": 0.7365105677579776, + "grad_norm": 0.2293413132429123, + "learning_rate": 1.3178615830915873e-06, + "loss": 0.4735, + "step": 8886 + }, + { + "epoch": 0.7365934521342727, + "grad_norm": 0.23184272646903992, + "learning_rate": 1.317447161210112e-06, + "loss": 0.4842, + "step": 8887 + }, + { + "epoch": 0.7366763365105677, + "grad_norm": 0.23969705402851105, + "learning_rate": 1.3170327393286366e-06, + "loss": 0.516, + "step": 8888 + }, + { + "epoch": 0.7367592208868629, + "grad_norm": 0.2603188753128052, + "learning_rate": 1.3166183174471615e-06, + "loss": 0.5538, + "step": 8889 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 0.2387988567352295, + "learning_rate": 1.316203895565686e-06, + "loss": 0.5157, + "step": 8890 + }, + { + "epoch": 0.736924989639453, + "grad_norm": 0.23925350606441498, + "learning_rate": 1.3157894736842106e-06, + "loss": 0.5061, + "step": 8891 + }, + { + "epoch": 0.7370078740157481, + "grad_norm": 0.2505304515361786, + "learning_rate": 1.3153750518027352e-06, + "loss": 0.4959, + "step": 8892 + }, + { + "epoch": 0.7370907583920431, + "grad_norm": 0.24507176876068115, + "learning_rate": 1.31496062992126e-06, + "loss": 0.4946, + "step": 8893 + }, + { + "epoch": 0.7371736427683382, + "grad_norm": 0.26935747265815735, + "learning_rate": 1.3145462080397847e-06, + "loss": 0.5683, + "step": 8894 + }, + { + "epoch": 0.7372565271446332, + "grad_norm": 0.23322004079818726, + "learning_rate": 1.3141317861583091e-06, + "loss": 0.4388, + "step": 8895 + }, + { + "epoch": 0.7373394115209283, + "grad_norm": 0.2309434711933136, + "learning_rate": 1.313717364276834e-06, + "loss": 0.4747, + "step": 8896 + }, + { + "epoch": 0.7374222958972234, + "grad_norm": 0.2451125979423523, + "learning_rate": 1.3133029423953586e-06, + "loss": 0.4837, + "step": 8897 + }, + { + "epoch": 0.7375051802735184, + "grad_norm": 0.23918043076992035, + "learning_rate": 1.3128885205138833e-06, + "loss": 0.5067, + "step": 8898 + }, + { + "epoch": 0.7375880646498135, + "grad_norm": 0.23201195895671844, + "learning_rate": 1.3124740986324077e-06, + "loss": 0.4936, + "step": 8899 + }, + { + "epoch": 0.7376709490261085, + "grad_norm": 0.2530789375305176, + "learning_rate": 1.3120596767509326e-06, + "loss": 0.4847, + "step": 8900 + }, + { + "epoch": 0.7377538334024036, + "grad_norm": 0.2681843638420105, + "learning_rate": 1.3116452548694572e-06, + "loss": 0.4963, + "step": 8901 + }, + { + "epoch": 0.7378367177786987, + "grad_norm": 0.268250435590744, + "learning_rate": 1.3112308329879818e-06, + "loss": 0.5168, + "step": 8902 + }, + { + "epoch": 0.7379196021549937, + "grad_norm": 0.23692594468593597, + "learning_rate": 1.3108164111065067e-06, + "loss": 0.4742, + "step": 8903 + }, + { + "epoch": 0.7380024865312889, + "grad_norm": 0.27911847829818726, + "learning_rate": 1.3104019892250311e-06, + "loss": 0.5356, + "step": 8904 + }, + { + "epoch": 0.738085370907584, + "grad_norm": 0.2531006634235382, + "learning_rate": 1.3099875673435558e-06, + "loss": 0.4934, + "step": 8905 + }, + { + "epoch": 0.738168255283879, + "grad_norm": 0.26254093647003174, + "learning_rate": 1.3095731454620806e-06, + "loss": 0.4978, + "step": 8906 + }, + { + "epoch": 0.7382511396601741, + "grad_norm": 0.22818733751773834, + "learning_rate": 1.3091587235806053e-06, + "loss": 0.4727, + "step": 8907 + }, + { + "epoch": 0.7383340240364691, + "grad_norm": 0.2536693215370178, + "learning_rate": 1.3087443016991297e-06, + "loss": 0.5283, + "step": 8908 + }, + { + "epoch": 0.7384169084127642, + "grad_norm": 0.259346604347229, + "learning_rate": 1.3083298798176544e-06, + "loss": 0.5126, + "step": 8909 + }, + { + "epoch": 0.7384997927890593, + "grad_norm": 0.23765160143375397, + "learning_rate": 1.3079154579361792e-06, + "loss": 0.4827, + "step": 8910 + }, + { + "epoch": 0.7385826771653543, + "grad_norm": 0.22990916669368744, + "learning_rate": 1.3075010360547039e-06, + "loss": 0.4651, + "step": 8911 + }, + { + "epoch": 0.7386655615416494, + "grad_norm": 0.247243732213974, + "learning_rate": 1.3070866141732285e-06, + "loss": 0.4828, + "step": 8912 + }, + { + "epoch": 0.7387484459179444, + "grad_norm": 0.24512691795825958, + "learning_rate": 1.306672192291753e-06, + "loss": 0.4736, + "step": 8913 + }, + { + "epoch": 0.7388313302942395, + "grad_norm": 0.24670717120170593, + "learning_rate": 1.3062577704102778e-06, + "loss": 0.4552, + "step": 8914 + }, + { + "epoch": 0.7389142146705346, + "grad_norm": 0.24987837672233582, + "learning_rate": 1.3058433485288024e-06, + "loss": 0.5388, + "step": 8915 + }, + { + "epoch": 0.7389970990468296, + "grad_norm": 0.2426641434431076, + "learning_rate": 1.3054289266473273e-06, + "loss": 0.5145, + "step": 8916 + }, + { + "epoch": 0.7390799834231248, + "grad_norm": 0.25330808758735657, + "learning_rate": 1.3050145047658517e-06, + "loss": 0.5255, + "step": 8917 + }, + { + "epoch": 0.7391628677994199, + "grad_norm": 0.23191629350185394, + "learning_rate": 1.3046000828843764e-06, + "loss": 0.4987, + "step": 8918 + }, + { + "epoch": 0.7392457521757149, + "grad_norm": 0.24102385342121124, + "learning_rate": 1.304185661002901e-06, + "loss": 0.5066, + "step": 8919 + }, + { + "epoch": 0.73932863655201, + "grad_norm": 0.2350786328315735, + "learning_rate": 1.3037712391214259e-06, + "loss": 0.4777, + "step": 8920 + }, + { + "epoch": 0.739411520928305, + "grad_norm": 0.2434397041797638, + "learning_rate": 1.3033568172399505e-06, + "loss": 0.4828, + "step": 8921 + }, + { + "epoch": 0.7394944053046001, + "grad_norm": 0.22661791741847992, + "learning_rate": 1.302942395358475e-06, + "loss": 0.4539, + "step": 8922 + }, + { + "epoch": 0.7395772896808952, + "grad_norm": 0.25815433263778687, + "learning_rate": 1.3025279734769996e-06, + "loss": 0.5484, + "step": 8923 + }, + { + "epoch": 0.7396601740571902, + "grad_norm": 0.24872098863124847, + "learning_rate": 1.3021135515955244e-06, + "loss": 0.5169, + "step": 8924 + }, + { + "epoch": 0.7397430584334853, + "grad_norm": 0.24624016880989075, + "learning_rate": 1.301699129714049e-06, + "loss": 0.5015, + "step": 8925 + }, + { + "epoch": 0.7398259428097803, + "grad_norm": 0.2623516321182251, + "learning_rate": 1.3012847078325735e-06, + "loss": 0.4993, + "step": 8926 + }, + { + "epoch": 0.7399088271860754, + "grad_norm": 0.2464321106672287, + "learning_rate": 1.3008702859510984e-06, + "loss": 0.4737, + "step": 8927 + }, + { + "epoch": 0.7399917115623705, + "grad_norm": 0.2485446184873581, + "learning_rate": 1.300455864069623e-06, + "loss": 0.4969, + "step": 8928 + }, + { + "epoch": 0.7400745959386655, + "grad_norm": 0.23874399065971375, + "learning_rate": 1.3000414421881476e-06, + "loss": 0.5231, + "step": 8929 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 0.252113938331604, + "learning_rate": 1.2996270203066725e-06, + "loss": 0.5224, + "step": 8930 + }, + { + "epoch": 0.7402403646912556, + "grad_norm": 0.26027384400367737, + "learning_rate": 1.299212598425197e-06, + "loss": 0.5054, + "step": 8931 + }, + { + "epoch": 0.7403232490675508, + "grad_norm": 0.2940969169139862, + "learning_rate": 1.2987981765437216e-06, + "loss": 0.4962, + "step": 8932 + }, + { + "epoch": 0.7404061334438459, + "grad_norm": 0.24097366631031036, + "learning_rate": 1.2983837546622462e-06, + "loss": 0.4724, + "step": 8933 + }, + { + "epoch": 0.7404890178201409, + "grad_norm": 0.27326756715774536, + "learning_rate": 1.297969332780771e-06, + "loss": 0.4895, + "step": 8934 + }, + { + "epoch": 0.740571902196436, + "grad_norm": 0.2468504011631012, + "learning_rate": 1.2975549108992955e-06, + "loss": 0.5051, + "step": 8935 + }, + { + "epoch": 0.7406547865727311, + "grad_norm": 0.2600789964199066, + "learning_rate": 1.2971404890178202e-06, + "loss": 0.546, + "step": 8936 + }, + { + "epoch": 0.7407376709490261, + "grad_norm": 0.22934766113758087, + "learning_rate": 1.296726067136345e-06, + "loss": 0.5062, + "step": 8937 + }, + { + "epoch": 0.7408205553253212, + "grad_norm": 0.27255043387413025, + "learning_rate": 1.2963116452548696e-06, + "loss": 0.4852, + "step": 8938 + }, + { + "epoch": 0.7409034397016162, + "grad_norm": 0.22799605131149292, + "learning_rate": 1.2958972233733943e-06, + "loss": 0.4665, + "step": 8939 + }, + { + "epoch": 0.7409863240779113, + "grad_norm": 0.24433350563049316, + "learning_rate": 1.2954828014919187e-06, + "loss": 0.532, + "step": 8940 + }, + { + "epoch": 0.7410692084542064, + "grad_norm": 0.2382727563381195, + "learning_rate": 1.2950683796104436e-06, + "loss": 0.4752, + "step": 8941 + }, + { + "epoch": 0.7411520928305014, + "grad_norm": 0.2574387788772583, + "learning_rate": 1.2946539577289682e-06, + "loss": 0.4826, + "step": 8942 + }, + { + "epoch": 0.7412349772067965, + "grad_norm": 0.2382907271385193, + "learning_rate": 1.2942395358474929e-06, + "loss": 0.4822, + "step": 8943 + }, + { + "epoch": 0.7413178615830915, + "grad_norm": 0.24860118329524994, + "learning_rate": 1.2938251139660173e-06, + "loss": 0.5176, + "step": 8944 + }, + { + "epoch": 0.7414007459593867, + "grad_norm": 0.24188539385795593, + "learning_rate": 1.2934106920845422e-06, + "loss": 0.5295, + "step": 8945 + }, + { + "epoch": 0.7414836303356818, + "grad_norm": 0.27551406621932983, + "learning_rate": 1.2929962702030668e-06, + "loss": 0.5319, + "step": 8946 + }, + { + "epoch": 0.7415665147119768, + "grad_norm": 0.25417906045913696, + "learning_rate": 1.2925818483215917e-06, + "loss": 0.5263, + "step": 8947 + }, + { + "epoch": 0.7416493990882719, + "grad_norm": 0.24139286577701569, + "learning_rate": 1.2921674264401163e-06, + "loss": 0.4872, + "step": 8948 + }, + { + "epoch": 0.7417322834645669, + "grad_norm": 0.25041335821151733, + "learning_rate": 1.2917530045586407e-06, + "loss": 0.4916, + "step": 8949 + }, + { + "epoch": 0.741815167840862, + "grad_norm": 0.24810190498828888, + "learning_rate": 1.2913385826771654e-06, + "loss": 0.5122, + "step": 8950 + }, + { + "epoch": 0.7418980522171571, + "grad_norm": 0.2399970144033432, + "learning_rate": 1.2909241607956902e-06, + "loss": 0.4848, + "step": 8951 + }, + { + "epoch": 0.7419809365934521, + "grad_norm": 0.25173959136009216, + "learning_rate": 1.2905097389142149e-06, + "loss": 0.5159, + "step": 8952 + }, + { + "epoch": 0.7420638209697472, + "grad_norm": 0.2261149287223816, + "learning_rate": 1.2900953170327393e-06, + "loss": 0.4413, + "step": 8953 + }, + { + "epoch": 0.7421467053460423, + "grad_norm": 0.2357540726661682, + "learning_rate": 1.289680895151264e-06, + "loss": 0.4733, + "step": 8954 + }, + { + "epoch": 0.7422295897223373, + "grad_norm": 0.2429419457912445, + "learning_rate": 1.2892664732697888e-06, + "loss": 0.474, + "step": 8955 + }, + { + "epoch": 0.7423124740986324, + "grad_norm": 0.2338411659002304, + "learning_rate": 1.2888520513883134e-06, + "loss": 0.4659, + "step": 8956 + }, + { + "epoch": 0.7423953584749274, + "grad_norm": 0.2700817584991455, + "learning_rate": 1.2884376295068379e-06, + "loss": 0.5494, + "step": 8957 + }, + { + "epoch": 0.7424782428512225, + "grad_norm": 0.23175737261772156, + "learning_rate": 1.2880232076253627e-06, + "loss": 0.4618, + "step": 8958 + }, + { + "epoch": 0.7425611272275177, + "grad_norm": 0.2609354853630066, + "learning_rate": 1.2876087857438874e-06, + "loss": 0.5274, + "step": 8959 + }, + { + "epoch": 0.7426440116038127, + "grad_norm": 0.2302442044019699, + "learning_rate": 1.287194363862412e-06, + "loss": 0.497, + "step": 8960 + }, + { + "epoch": 0.7427268959801078, + "grad_norm": 0.2919301986694336, + "learning_rate": 1.2867799419809369e-06, + "loss": 0.4973, + "step": 8961 + }, + { + "epoch": 0.7428097803564028, + "grad_norm": 0.25246113538742065, + "learning_rate": 1.2863655200994613e-06, + "loss": 0.5073, + "step": 8962 + }, + { + "epoch": 0.7428926647326979, + "grad_norm": 0.24709546566009521, + "learning_rate": 1.285951098217986e-06, + "loss": 0.4998, + "step": 8963 + }, + { + "epoch": 0.742975549108993, + "grad_norm": 0.23775774240493774, + "learning_rate": 1.2855366763365106e-06, + "loss": 0.487, + "step": 8964 + }, + { + "epoch": 0.743058433485288, + "grad_norm": 0.28472641110420227, + "learning_rate": 1.2851222544550354e-06, + "loss": 0.5067, + "step": 8965 + }, + { + "epoch": 0.7431413178615831, + "grad_norm": 0.2410329133272171, + "learning_rate": 1.2847078325735599e-06, + "loss": 0.4802, + "step": 8966 + }, + { + "epoch": 0.7432242022378781, + "grad_norm": 0.24159009754657745, + "learning_rate": 1.2842934106920845e-06, + "loss": 0.5161, + "step": 8967 + }, + { + "epoch": 0.7433070866141732, + "grad_norm": 0.25662556290626526, + "learning_rate": 1.2838789888106094e-06, + "loss": 0.5464, + "step": 8968 + }, + { + "epoch": 0.7433899709904683, + "grad_norm": 0.2423292100429535, + "learning_rate": 1.283464566929134e-06, + "loss": 0.4998, + "step": 8969 + }, + { + "epoch": 0.7434728553667633, + "grad_norm": 0.22662906348705292, + "learning_rate": 1.2830501450476587e-06, + "loss": 0.4569, + "step": 8970 + }, + { + "epoch": 0.7435557397430584, + "grad_norm": 0.23656485974788666, + "learning_rate": 1.282635723166183e-06, + "loss": 0.5132, + "step": 8971 + }, + { + "epoch": 0.7436386241193536, + "grad_norm": 0.24495424330234528, + "learning_rate": 1.282221301284708e-06, + "loss": 0.5231, + "step": 8972 + }, + { + "epoch": 0.7437215084956486, + "grad_norm": 0.24893556535243988, + "learning_rate": 1.2818068794032326e-06, + "loss": 0.522, + "step": 8973 + }, + { + "epoch": 0.7438043928719437, + "grad_norm": 0.23508214950561523, + "learning_rate": 1.2813924575217572e-06, + "loss": 0.471, + "step": 8974 + }, + { + "epoch": 0.7438872772482387, + "grad_norm": 0.2518083155155182, + "learning_rate": 1.2809780356402819e-06, + "loss": 0.4975, + "step": 8975 + }, + { + "epoch": 0.7439701616245338, + "grad_norm": 0.25646039843559265, + "learning_rate": 1.2805636137588065e-06, + "loss": 0.5242, + "step": 8976 + }, + { + "epoch": 0.7440530460008289, + "grad_norm": 0.23000703752040863, + "learning_rate": 1.2801491918773312e-06, + "loss": 0.491, + "step": 8977 + }, + { + "epoch": 0.7441359303771239, + "grad_norm": 0.24102075397968292, + "learning_rate": 1.279734769995856e-06, + "loss": 0.5204, + "step": 8978 + }, + { + "epoch": 0.744218814753419, + "grad_norm": 0.25294286012649536, + "learning_rate": 1.2793203481143807e-06, + "loss": 0.4734, + "step": 8979 + }, + { + "epoch": 0.744301699129714, + "grad_norm": 0.2463332712650299, + "learning_rate": 1.278905926232905e-06, + "loss": 0.4794, + "step": 8980 + }, + { + "epoch": 0.7443845835060091, + "grad_norm": 0.23765911161899567, + "learning_rate": 1.2784915043514297e-06, + "loss": 0.4961, + "step": 8981 + }, + { + "epoch": 0.7444674678823042, + "grad_norm": 0.2493434101343155, + "learning_rate": 1.2780770824699546e-06, + "loss": 0.4937, + "step": 8982 + }, + { + "epoch": 0.7445503522585992, + "grad_norm": 0.2482835054397583, + "learning_rate": 1.2776626605884792e-06, + "loss": 0.4798, + "step": 8983 + }, + { + "epoch": 0.7446332366348943, + "grad_norm": 0.24131125211715698, + "learning_rate": 1.2772482387070037e-06, + "loss": 0.5129, + "step": 8984 + }, + { + "epoch": 0.7447161210111893, + "grad_norm": 0.2469405084848404, + "learning_rate": 1.2768338168255285e-06, + "loss": 0.4858, + "step": 8985 + }, + { + "epoch": 0.7447990053874844, + "grad_norm": 0.23941510915756226, + "learning_rate": 1.2764193949440532e-06, + "loss": 0.5087, + "step": 8986 + }, + { + "epoch": 0.7448818897637796, + "grad_norm": 0.2577242851257324, + "learning_rate": 1.2760049730625778e-06, + "loss": 0.5216, + "step": 8987 + }, + { + "epoch": 0.7449647741400746, + "grad_norm": 0.23870986700057983, + "learning_rate": 1.2755905511811027e-06, + "loss": 0.4482, + "step": 8988 + }, + { + "epoch": 0.7450476585163697, + "grad_norm": 0.24633656442165375, + "learning_rate": 1.2751761292996271e-06, + "loss": 0.4752, + "step": 8989 + }, + { + "epoch": 0.7451305428926648, + "grad_norm": 0.23280534148216248, + "learning_rate": 1.2747617074181517e-06, + "loss": 0.5009, + "step": 8990 + }, + { + "epoch": 0.7452134272689598, + "grad_norm": 0.26524055004119873, + "learning_rate": 1.2743472855366764e-06, + "loss": 0.5146, + "step": 8991 + }, + { + "epoch": 0.7452963116452549, + "grad_norm": 0.24762777984142303, + "learning_rate": 1.2739328636552012e-06, + "loss": 0.5058, + "step": 8992 + }, + { + "epoch": 0.7453791960215499, + "grad_norm": 0.24499180912971497, + "learning_rate": 1.2735184417737257e-06, + "loss": 0.5212, + "step": 8993 + }, + { + "epoch": 0.745462080397845, + "grad_norm": 0.2387910932302475, + "learning_rate": 1.2731040198922503e-06, + "loss": 0.4802, + "step": 8994 + }, + { + "epoch": 0.7455449647741401, + "grad_norm": 0.2401602864265442, + "learning_rate": 1.272689598010775e-06, + "loss": 0.5134, + "step": 8995 + }, + { + "epoch": 0.7456278491504351, + "grad_norm": 0.2455013394355774, + "learning_rate": 1.2722751761292998e-06, + "loss": 0.4957, + "step": 8996 + }, + { + "epoch": 0.7457107335267302, + "grad_norm": 0.23541295528411865, + "learning_rate": 1.2718607542478245e-06, + "loss": 0.4464, + "step": 8997 + }, + { + "epoch": 0.7457936179030252, + "grad_norm": 0.23610997200012207, + "learning_rate": 1.271446332366349e-06, + "loss": 0.5234, + "step": 8998 + }, + { + "epoch": 0.7458765022793203, + "grad_norm": 0.24459779262542725, + "learning_rate": 1.2710319104848738e-06, + "loss": 0.5023, + "step": 8999 + }, + { + "epoch": 0.7459593866556155, + "grad_norm": 0.2411675900220871, + "learning_rate": 1.2706174886033984e-06, + "loss": 0.5072, + "step": 9000 + }, + { + "epoch": 0.7460422710319105, + "grad_norm": 0.24411973357200623, + "learning_rate": 1.270203066721923e-06, + "loss": 0.5303, + "step": 9001 + }, + { + "epoch": 0.7461251554082056, + "grad_norm": 0.24283970892429352, + "learning_rate": 1.2697886448404475e-06, + "loss": 0.4878, + "step": 9002 + }, + { + "epoch": 0.7462080397845007, + "grad_norm": 0.24160367250442505, + "learning_rate": 1.2693742229589723e-06, + "loss": 0.4765, + "step": 9003 + }, + { + "epoch": 0.7462909241607957, + "grad_norm": 0.2565033733844757, + "learning_rate": 1.268959801077497e-06, + "loss": 0.4775, + "step": 9004 + }, + { + "epoch": 0.7463738085370908, + "grad_norm": 0.25146883726119995, + "learning_rate": 1.2685453791960216e-06, + "loss": 0.4982, + "step": 9005 + }, + { + "epoch": 0.7464566929133858, + "grad_norm": 0.25057587027549744, + "learning_rate": 1.2681309573145465e-06, + "loss": 0.4996, + "step": 9006 + }, + { + "epoch": 0.7465395772896809, + "grad_norm": 0.23499523103237152, + "learning_rate": 1.267716535433071e-06, + "loss": 0.5018, + "step": 9007 + }, + { + "epoch": 0.746622461665976, + "grad_norm": 0.22997602820396423, + "learning_rate": 1.2673021135515955e-06, + "loss": 0.4913, + "step": 9008 + }, + { + "epoch": 0.746705346042271, + "grad_norm": 0.2303667962551117, + "learning_rate": 1.2668876916701204e-06, + "loss": 0.4896, + "step": 9009 + }, + { + "epoch": 0.7467882304185661, + "grad_norm": 0.2574867308139801, + "learning_rate": 1.266473269788645e-06, + "loss": 0.5515, + "step": 9010 + }, + { + "epoch": 0.7468711147948611, + "grad_norm": 0.24333350360393524, + "learning_rate": 1.2660588479071695e-06, + "loss": 0.4941, + "step": 9011 + }, + { + "epoch": 0.7469539991711562, + "grad_norm": 0.2547963559627533, + "learning_rate": 1.2656444260256941e-06, + "loss": 0.485, + "step": 9012 + }, + { + "epoch": 0.7470368835474513, + "grad_norm": 0.24535715579986572, + "learning_rate": 1.265230004144219e-06, + "loss": 0.4984, + "step": 9013 + }, + { + "epoch": 0.7471197679237463, + "grad_norm": 0.22037847340106964, + "learning_rate": 1.2648155822627436e-06, + "loss": 0.4743, + "step": 9014 + }, + { + "epoch": 0.7472026523000415, + "grad_norm": 0.24020710587501526, + "learning_rate": 1.2644011603812683e-06, + "loss": 0.5125, + "step": 9015 + }, + { + "epoch": 0.7472855366763365, + "grad_norm": 0.24732592701911926, + "learning_rate": 1.263986738499793e-06, + "loss": 0.5085, + "step": 9016 + }, + { + "epoch": 0.7473684210526316, + "grad_norm": 0.24801403284072876, + "learning_rate": 1.2635723166183175e-06, + "loss": 0.5052, + "step": 9017 + }, + { + "epoch": 0.7474513054289267, + "grad_norm": 0.24207425117492676, + "learning_rate": 1.2631578947368422e-06, + "loss": 0.5221, + "step": 9018 + }, + { + "epoch": 0.7475341898052217, + "grad_norm": 0.2430393099784851, + "learning_rate": 1.262743472855367e-06, + "loss": 0.5072, + "step": 9019 + }, + { + "epoch": 0.7476170741815168, + "grad_norm": 0.21499121189117432, + "learning_rate": 1.2623290509738915e-06, + "loss": 0.4918, + "step": 9020 + }, + { + "epoch": 0.7476999585578119, + "grad_norm": 0.24160154163837433, + "learning_rate": 1.2619146290924161e-06, + "loss": 0.4982, + "step": 9021 + }, + { + "epoch": 0.7477828429341069, + "grad_norm": 0.23159551620483398, + "learning_rate": 1.2615002072109408e-06, + "loss": 0.4895, + "step": 9022 + }, + { + "epoch": 0.747865727310402, + "grad_norm": 0.23968185484409332, + "learning_rate": 1.2610857853294656e-06, + "loss": 0.4927, + "step": 9023 + }, + { + "epoch": 0.747948611686697, + "grad_norm": 0.2589290738105774, + "learning_rate": 1.2606713634479903e-06, + "loss": 0.5397, + "step": 9024 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 0.2367514818906784, + "learning_rate": 1.2602569415665147e-06, + "loss": 0.5057, + "step": 9025 + }, + { + "epoch": 0.7481143804392872, + "grad_norm": 0.23001249134540558, + "learning_rate": 1.2598425196850396e-06, + "loss": 0.499, + "step": 9026 + }, + { + "epoch": 0.7481972648155822, + "grad_norm": 0.23180434107780457, + "learning_rate": 1.2594280978035642e-06, + "loss": 0.4769, + "step": 9027 + }, + { + "epoch": 0.7482801491918774, + "grad_norm": 0.23156492412090302, + "learning_rate": 1.2590136759220888e-06, + "loss": 0.4778, + "step": 9028 + }, + { + "epoch": 0.7483630335681724, + "grad_norm": 0.23247098922729492, + "learning_rate": 1.2585992540406133e-06, + "loss": 0.504, + "step": 9029 + }, + { + "epoch": 0.7484459179444675, + "grad_norm": 0.23233254253864288, + "learning_rate": 1.2581848321591381e-06, + "loss": 0.488, + "step": 9030 + }, + { + "epoch": 0.7485288023207626, + "grad_norm": 0.22248941659927368, + "learning_rate": 1.2577704102776628e-06, + "loss": 0.46, + "step": 9031 + }, + { + "epoch": 0.7486116866970576, + "grad_norm": 0.2502148747444153, + "learning_rate": 1.2573559883961874e-06, + "loss": 0.4674, + "step": 9032 + }, + { + "epoch": 0.7486945710733527, + "grad_norm": 0.24500887095928192, + "learning_rate": 1.2569415665147118e-06, + "loss": 0.5176, + "step": 9033 + }, + { + "epoch": 0.7487774554496477, + "grad_norm": 0.26742252707481384, + "learning_rate": 1.2565271446332367e-06, + "loss": 0.5187, + "step": 9034 + }, + { + "epoch": 0.7488603398259428, + "grad_norm": 0.2873620390892029, + "learning_rate": 1.2561127227517613e-06, + "loss": 0.5359, + "step": 9035 + }, + { + "epoch": 0.7489432242022379, + "grad_norm": 0.24826759099960327, + "learning_rate": 1.2556983008702862e-06, + "loss": 0.4996, + "step": 9036 + }, + { + "epoch": 0.7490261085785329, + "grad_norm": 0.2644975781440735, + "learning_rate": 1.2552838789888108e-06, + "loss": 0.5577, + "step": 9037 + }, + { + "epoch": 0.749108992954828, + "grad_norm": 0.24382662773132324, + "learning_rate": 1.2548694571073353e-06, + "loss": 0.5059, + "step": 9038 + }, + { + "epoch": 0.7491918773311231, + "grad_norm": 0.26213371753692627, + "learning_rate": 1.25445503522586e-06, + "loss": 0.4766, + "step": 9039 + }, + { + "epoch": 0.7492747617074181, + "grad_norm": 0.24298511445522308, + "learning_rate": 1.2540406133443848e-06, + "loss": 0.4864, + "step": 9040 + }, + { + "epoch": 0.7493576460837132, + "grad_norm": 0.25007325410842896, + "learning_rate": 1.2536261914629094e-06, + "loss": 0.4807, + "step": 9041 + }, + { + "epoch": 0.7494405304600082, + "grad_norm": 0.2417823225259781, + "learning_rate": 1.2532117695814339e-06, + "loss": 0.4765, + "step": 9042 + }, + { + "epoch": 0.7495234148363034, + "grad_norm": 0.24214598536491394, + "learning_rate": 1.2527973476999585e-06, + "loss": 0.4611, + "step": 9043 + }, + { + "epoch": 0.7496062992125985, + "grad_norm": 0.22952638566493988, + "learning_rate": 1.2523829258184833e-06, + "loss": 0.5046, + "step": 9044 + }, + { + "epoch": 0.7496891835888935, + "grad_norm": 0.24335628747940063, + "learning_rate": 1.251968503937008e-06, + "loss": 0.5139, + "step": 9045 + }, + { + "epoch": 0.7497720679651886, + "grad_norm": 0.23835542798042297, + "learning_rate": 1.2515540820555328e-06, + "loss": 0.5034, + "step": 9046 + }, + { + "epoch": 0.7498549523414836, + "grad_norm": 0.23944908380508423, + "learning_rate": 1.2511396601740573e-06, + "loss": 0.4973, + "step": 9047 + }, + { + "epoch": 0.7499378367177787, + "grad_norm": 0.2387368381023407, + "learning_rate": 1.250725238292582e-06, + "loss": 0.4456, + "step": 9048 + }, + { + "epoch": 0.7500207210940738, + "grad_norm": 0.23810747265815735, + "learning_rate": 1.2503108164111066e-06, + "loss": 0.4898, + "step": 9049 + }, + { + "epoch": 0.7501036054703688, + "grad_norm": 0.25897103548049927, + "learning_rate": 1.2498963945296312e-06, + "loss": 0.521, + "step": 9050 + }, + { + "epoch": 0.7501864898466639, + "grad_norm": 0.2504790723323822, + "learning_rate": 1.2494819726481559e-06, + "loss": 0.5397, + "step": 9051 + }, + { + "epoch": 0.7502693742229589, + "grad_norm": 0.2599237859249115, + "learning_rate": 1.2490675507666805e-06, + "loss": 0.4959, + "step": 9052 + }, + { + "epoch": 0.750352258599254, + "grad_norm": 0.273995578289032, + "learning_rate": 1.2486531288852051e-06, + "loss": 0.5124, + "step": 9053 + }, + { + "epoch": 0.7504351429755491, + "grad_norm": 0.2438417375087738, + "learning_rate": 1.24823870700373e-06, + "loss": 0.4805, + "step": 9054 + }, + { + "epoch": 0.7505180273518441, + "grad_norm": 0.2314801961183548, + "learning_rate": 1.2478242851222546e-06, + "loss": 0.4727, + "step": 9055 + }, + { + "epoch": 0.7506009117281393, + "grad_norm": 0.2396225929260254, + "learning_rate": 1.2474098632407793e-06, + "loss": 0.4568, + "step": 9056 + }, + { + "epoch": 0.7506837961044344, + "grad_norm": 0.23322929441928864, + "learning_rate": 1.246995441359304e-06, + "loss": 0.4866, + "step": 9057 + }, + { + "epoch": 0.7507666804807294, + "grad_norm": 0.2520090937614441, + "learning_rate": 1.2465810194778286e-06, + "loss": 0.4469, + "step": 9058 + }, + { + "epoch": 0.7508495648570245, + "grad_norm": 0.2612369954586029, + "learning_rate": 1.2461665975963532e-06, + "loss": 0.5081, + "step": 9059 + }, + { + "epoch": 0.7509324492333195, + "grad_norm": 0.24522066116333008, + "learning_rate": 1.2457521757148779e-06, + "loss": 0.493, + "step": 9060 + }, + { + "epoch": 0.7510153336096146, + "grad_norm": 0.24588586390018463, + "learning_rate": 1.2453377538334025e-06, + "loss": 0.48, + "step": 9061 + }, + { + "epoch": 0.7510982179859097, + "grad_norm": 0.23599819839000702, + "learning_rate": 1.2449233319519271e-06, + "loss": 0.4699, + "step": 9062 + }, + { + "epoch": 0.7511811023622047, + "grad_norm": 0.2560065686702728, + "learning_rate": 1.2445089100704518e-06, + "loss": 0.5645, + "step": 9063 + }, + { + "epoch": 0.7512639867384998, + "grad_norm": 0.24901354312896729, + "learning_rate": 1.2440944881889764e-06, + "loss": 0.4992, + "step": 9064 + }, + { + "epoch": 0.7513468711147948, + "grad_norm": 0.25830796360969543, + "learning_rate": 1.2436800663075013e-06, + "loss": 0.5267, + "step": 9065 + }, + { + "epoch": 0.7514297554910899, + "grad_norm": 0.2527487874031067, + "learning_rate": 1.2432656444260257e-06, + "loss": 0.5056, + "step": 9066 + }, + { + "epoch": 0.751512639867385, + "grad_norm": 0.25650763511657715, + "learning_rate": 1.2428512225445506e-06, + "loss": 0.4689, + "step": 9067 + }, + { + "epoch": 0.75159552424368, + "grad_norm": 0.25458472967147827, + "learning_rate": 1.242436800663075e-06, + "loss": 0.5026, + "step": 9068 + }, + { + "epoch": 0.7516784086199751, + "grad_norm": 0.2594769299030304, + "learning_rate": 1.2420223787815999e-06, + "loss": 0.538, + "step": 9069 + }, + { + "epoch": 0.7517612929962703, + "grad_norm": 0.2588931918144226, + "learning_rate": 1.2416079569001243e-06, + "loss": 0.509, + "step": 9070 + }, + { + "epoch": 0.7518441773725653, + "grad_norm": 0.24007928371429443, + "learning_rate": 1.2411935350186491e-06, + "loss": 0.481, + "step": 9071 + }, + { + "epoch": 0.7519270617488604, + "grad_norm": 0.26551467180252075, + "learning_rate": 1.2407791131371738e-06, + "loss": 0.472, + "step": 9072 + }, + { + "epoch": 0.7520099461251554, + "grad_norm": 0.24877677857875824, + "learning_rate": 1.2403646912556984e-06, + "loss": 0.4767, + "step": 9073 + }, + { + "epoch": 0.7520928305014505, + "grad_norm": 0.24811901152133942, + "learning_rate": 1.239950269374223e-06, + "loss": 0.519, + "step": 9074 + }, + { + "epoch": 0.7521757148777456, + "grad_norm": 0.23987990617752075, + "learning_rate": 1.2395358474927477e-06, + "loss": 0.491, + "step": 9075 + }, + { + "epoch": 0.7522585992540406, + "grad_norm": 0.23099908232688904, + "learning_rate": 1.2391214256112724e-06, + "loss": 0.4792, + "step": 9076 + }, + { + "epoch": 0.7523414836303357, + "grad_norm": 0.2401091456413269, + "learning_rate": 1.238707003729797e-06, + "loss": 0.5051, + "step": 9077 + }, + { + "epoch": 0.7524243680066307, + "grad_norm": 0.24814124405384064, + "learning_rate": 1.2382925818483217e-06, + "loss": 0.4976, + "step": 9078 + }, + { + "epoch": 0.7525072523829258, + "grad_norm": 0.22992414236068726, + "learning_rate": 1.2378781599668463e-06, + "loss": 0.4676, + "step": 9079 + }, + { + "epoch": 0.7525901367592209, + "grad_norm": 0.24717119336128235, + "learning_rate": 1.237463738085371e-06, + "loss": 0.4764, + "step": 9080 + }, + { + "epoch": 0.7526730211355159, + "grad_norm": 0.23090842366218567, + "learning_rate": 1.2370493162038958e-06, + "loss": 0.4414, + "step": 9081 + }, + { + "epoch": 0.752755905511811, + "grad_norm": 0.23694181442260742, + "learning_rate": 1.2366348943224202e-06, + "loss": 0.5162, + "step": 9082 + }, + { + "epoch": 0.752838789888106, + "grad_norm": 0.23412460088729858, + "learning_rate": 1.236220472440945e-06, + "loss": 0.5064, + "step": 9083 + }, + { + "epoch": 0.7529216742644012, + "grad_norm": 0.24593006074428558, + "learning_rate": 1.2358060505594695e-06, + "loss": 0.5235, + "step": 9084 + }, + { + "epoch": 0.7530045586406963, + "grad_norm": 0.2466410994529724, + "learning_rate": 1.2353916286779944e-06, + "loss": 0.4949, + "step": 9085 + }, + { + "epoch": 0.7530874430169913, + "grad_norm": 0.24457021057605743, + "learning_rate": 1.234977206796519e-06, + "loss": 0.5347, + "step": 9086 + }, + { + "epoch": 0.7531703273932864, + "grad_norm": 0.36113184690475464, + "learning_rate": 1.2345627849150437e-06, + "loss": 0.481, + "step": 9087 + }, + { + "epoch": 0.7532532117695815, + "grad_norm": 0.24006661772727966, + "learning_rate": 1.2341483630335683e-06, + "loss": 0.5034, + "step": 9088 + }, + { + "epoch": 0.7533360961458765, + "grad_norm": 0.23903138935565948, + "learning_rate": 1.233733941152093e-06, + "loss": 0.5151, + "step": 9089 + }, + { + "epoch": 0.7534189805221716, + "grad_norm": 0.24114297330379486, + "learning_rate": 1.2333195192706176e-06, + "loss": 0.503, + "step": 9090 + }, + { + "epoch": 0.7535018648984666, + "grad_norm": 0.23712772130966187, + "learning_rate": 1.2329050973891422e-06, + "loss": 0.444, + "step": 9091 + }, + { + "epoch": 0.7535847492747617, + "grad_norm": 0.2242259681224823, + "learning_rate": 1.2324906755076669e-06, + "loss": 0.4783, + "step": 9092 + }, + { + "epoch": 0.7536676336510568, + "grad_norm": 0.24108374118804932, + "learning_rate": 1.2320762536261915e-06, + "loss": 0.5325, + "step": 9093 + }, + { + "epoch": 0.7537505180273518, + "grad_norm": 0.276334285736084, + "learning_rate": 1.2316618317447162e-06, + "loss": 0.5443, + "step": 9094 + }, + { + "epoch": 0.7538334024036469, + "grad_norm": 0.23184460401535034, + "learning_rate": 1.2312474098632408e-06, + "loss": 0.4786, + "step": 9095 + }, + { + "epoch": 0.7539162867799419, + "grad_norm": 0.2410542070865631, + "learning_rate": 1.2308329879817657e-06, + "loss": 0.5203, + "step": 9096 + }, + { + "epoch": 0.753999171156237, + "grad_norm": 0.2531834542751312, + "learning_rate": 1.23041856610029e-06, + "loss": 0.5716, + "step": 9097 + }, + { + "epoch": 0.7540820555325322, + "grad_norm": 0.23831520974636078, + "learning_rate": 1.230004144218815e-06, + "loss": 0.5064, + "step": 9098 + }, + { + "epoch": 0.7541649399088272, + "grad_norm": 0.24094998836517334, + "learning_rate": 1.2295897223373394e-06, + "loss": 0.4949, + "step": 9099 + }, + { + "epoch": 0.7542478242851223, + "grad_norm": 0.24528558552265167, + "learning_rate": 1.2291753004558642e-06, + "loss": 0.5441, + "step": 9100 + }, + { + "epoch": 0.7543307086614173, + "grad_norm": 0.25183993577957153, + "learning_rate": 1.2287608785743889e-06, + "loss": 0.4796, + "step": 9101 + }, + { + "epoch": 0.7544135930377124, + "grad_norm": 0.24654942750930786, + "learning_rate": 1.2283464566929135e-06, + "loss": 0.4855, + "step": 9102 + }, + { + "epoch": 0.7544964774140075, + "grad_norm": 0.2484031766653061, + "learning_rate": 1.2279320348114382e-06, + "loss": 0.4913, + "step": 9103 + }, + { + "epoch": 0.7545793617903025, + "grad_norm": 0.25080206990242004, + "learning_rate": 1.2275176129299628e-06, + "loss": 0.5107, + "step": 9104 + }, + { + "epoch": 0.7546622461665976, + "grad_norm": 0.263879656791687, + "learning_rate": 1.2271031910484875e-06, + "loss": 0.5186, + "step": 9105 + }, + { + "epoch": 0.7547451305428927, + "grad_norm": 0.25774216651916504, + "learning_rate": 1.226688769167012e-06, + "loss": 0.5101, + "step": 9106 + }, + { + "epoch": 0.7548280149191877, + "grad_norm": 0.2425256371498108, + "learning_rate": 1.2262743472855367e-06, + "loss": 0.5016, + "step": 9107 + }, + { + "epoch": 0.7549108992954828, + "grad_norm": 0.24748824536800385, + "learning_rate": 1.2258599254040614e-06, + "loss": 0.5161, + "step": 9108 + }, + { + "epoch": 0.7549937836717778, + "grad_norm": 0.24029919505119324, + "learning_rate": 1.225445503522586e-06, + "loss": 0.4966, + "step": 9109 + }, + { + "epoch": 0.7550766680480729, + "grad_norm": 0.24649040400981903, + "learning_rate": 1.2250310816411109e-06, + "loss": 0.4873, + "step": 9110 + }, + { + "epoch": 0.755159552424368, + "grad_norm": 0.2377341389656067, + "learning_rate": 1.2246166597596353e-06, + "loss": 0.4825, + "step": 9111 + }, + { + "epoch": 0.755242436800663, + "grad_norm": 0.24163603782653809, + "learning_rate": 1.2242022378781602e-06, + "loss": 0.491, + "step": 9112 + }, + { + "epoch": 0.7553253211769582, + "grad_norm": 0.23585794866085052, + "learning_rate": 1.2237878159966846e-06, + "loss": 0.4708, + "step": 9113 + }, + { + "epoch": 0.7554082055532532, + "grad_norm": 0.24145448207855225, + "learning_rate": 1.2233733941152095e-06, + "loss": 0.4876, + "step": 9114 + }, + { + "epoch": 0.7554910899295483, + "grad_norm": 0.2597424387931824, + "learning_rate": 1.222958972233734e-06, + "loss": 0.4677, + "step": 9115 + }, + { + "epoch": 0.7555739743058434, + "grad_norm": 0.2319059520959854, + "learning_rate": 1.2225445503522587e-06, + "loss": 0.4655, + "step": 9116 + }, + { + "epoch": 0.7556568586821384, + "grad_norm": 0.2477930188179016, + "learning_rate": 1.2221301284707834e-06, + "loss": 0.5085, + "step": 9117 + }, + { + "epoch": 0.7557397430584335, + "grad_norm": 0.24574969708919525, + "learning_rate": 1.221715706589308e-06, + "loss": 0.517, + "step": 9118 + }, + { + "epoch": 0.7558226274347285, + "grad_norm": 0.23506680130958557, + "learning_rate": 1.2213012847078327e-06, + "loss": 0.4748, + "step": 9119 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 0.24336712062358856, + "learning_rate": 1.2208868628263573e-06, + "loss": 0.5011, + "step": 9120 + }, + { + "epoch": 0.7559883961873187, + "grad_norm": 0.2630930542945862, + "learning_rate": 1.220472440944882e-06, + "loss": 0.5314, + "step": 9121 + }, + { + "epoch": 0.7560712805636137, + "grad_norm": 0.2531511187553406, + "learning_rate": 1.2200580190634066e-06, + "loss": 0.5598, + "step": 9122 + }, + { + "epoch": 0.7561541649399088, + "grad_norm": 0.25447309017181396, + "learning_rate": 1.2196435971819312e-06, + "loss": 0.5146, + "step": 9123 + }, + { + "epoch": 0.756237049316204, + "grad_norm": 0.2747897803783417, + "learning_rate": 1.2192291753004559e-06, + "loss": 0.4974, + "step": 9124 + }, + { + "epoch": 0.756319933692499, + "grad_norm": 0.23604395985603333, + "learning_rate": 1.2188147534189807e-06, + "loss": 0.5077, + "step": 9125 + }, + { + "epoch": 0.7564028180687941, + "grad_norm": 0.24915669858455658, + "learning_rate": 1.2184003315375052e-06, + "loss": 0.5202, + "step": 9126 + }, + { + "epoch": 0.7564857024450891, + "grad_norm": 0.25037682056427, + "learning_rate": 1.21798590965603e-06, + "loss": 0.5416, + "step": 9127 + }, + { + "epoch": 0.7565685868213842, + "grad_norm": 0.2735596001148224, + "learning_rate": 1.2175714877745545e-06, + "loss": 0.509, + "step": 9128 + }, + { + "epoch": 0.7566514711976793, + "grad_norm": 0.2499213069677353, + "learning_rate": 1.2171570658930793e-06, + "loss": 0.5052, + "step": 9129 + }, + { + "epoch": 0.7567343555739743, + "grad_norm": 0.2510205805301666, + "learning_rate": 1.216742644011604e-06, + "loss": 0.513, + "step": 9130 + }, + { + "epoch": 0.7568172399502694, + "grad_norm": 0.24303700029850006, + "learning_rate": 1.2163282221301286e-06, + "loss": 0.5363, + "step": 9131 + }, + { + "epoch": 0.7569001243265644, + "grad_norm": 0.23950795829296112, + "learning_rate": 1.2159138002486533e-06, + "loss": 0.5309, + "step": 9132 + }, + { + "epoch": 0.7569830087028595, + "grad_norm": 0.2382308840751648, + "learning_rate": 1.215499378367178e-06, + "loss": 0.4901, + "step": 9133 + }, + { + "epoch": 0.7570658930791546, + "grad_norm": 0.24924960732460022, + "learning_rate": 1.2150849564857025e-06, + "loss": 0.4815, + "step": 9134 + }, + { + "epoch": 0.7571487774554496, + "grad_norm": 0.22440138459205627, + "learning_rate": 1.2146705346042272e-06, + "loss": 0.4379, + "step": 9135 + }, + { + "epoch": 0.7572316618317447, + "grad_norm": 0.25621283054351807, + "learning_rate": 1.2142561127227518e-06, + "loss": 0.4911, + "step": 9136 + }, + { + "epoch": 0.7573145462080397, + "grad_norm": 0.22619780898094177, + "learning_rate": 1.2138416908412765e-06, + "loss": 0.4898, + "step": 9137 + }, + { + "epoch": 0.7573974305843348, + "grad_norm": 0.25213927030563354, + "learning_rate": 1.2134272689598011e-06, + "loss": 0.5377, + "step": 9138 + }, + { + "epoch": 0.75748031496063, + "grad_norm": 0.2426757514476776, + "learning_rate": 1.213012847078326e-06, + "loss": 0.5244, + "step": 9139 + }, + { + "epoch": 0.757563199336925, + "grad_norm": 0.24859462678432465, + "learning_rate": 1.2125984251968504e-06, + "loss": 0.5178, + "step": 9140 + }, + { + "epoch": 0.7576460837132201, + "grad_norm": 0.23337869346141815, + "learning_rate": 1.2121840033153753e-06, + "loss": 0.4643, + "step": 9141 + }, + { + "epoch": 0.7577289680895152, + "grad_norm": 0.2340867966413498, + "learning_rate": 1.2117695814338997e-06, + "loss": 0.4937, + "step": 9142 + }, + { + "epoch": 0.7578118524658102, + "grad_norm": 0.2411988526582718, + "learning_rate": 1.2113551595524245e-06, + "loss": 0.4894, + "step": 9143 + }, + { + "epoch": 0.7578947368421053, + "grad_norm": 0.25208863615989685, + "learning_rate": 1.210940737670949e-06, + "loss": 0.5113, + "step": 9144 + }, + { + "epoch": 0.7579776212184003, + "grad_norm": 0.22629322111606598, + "learning_rate": 1.2105263157894738e-06, + "loss": 0.4866, + "step": 9145 + }, + { + "epoch": 0.7580605055946954, + "grad_norm": 0.23504644632339478, + "learning_rate": 1.2101118939079985e-06, + "loss": 0.4836, + "step": 9146 + }, + { + "epoch": 0.7581433899709905, + "grad_norm": 0.23828890919685364, + "learning_rate": 1.2096974720265231e-06, + "loss": 0.4762, + "step": 9147 + }, + { + "epoch": 0.7582262743472855, + "grad_norm": 0.256966233253479, + "learning_rate": 1.2092830501450478e-06, + "loss": 0.4741, + "step": 9148 + }, + { + "epoch": 0.7583091587235806, + "grad_norm": 0.24449549615383148, + "learning_rate": 1.2088686282635724e-06, + "loss": 0.4886, + "step": 9149 + }, + { + "epoch": 0.7583920430998756, + "grad_norm": 0.23073886334896088, + "learning_rate": 1.208454206382097e-06, + "loss": 0.4946, + "step": 9150 + }, + { + "epoch": 0.7584749274761707, + "grad_norm": 0.23285506665706635, + "learning_rate": 1.2080397845006217e-06, + "loss": 0.4942, + "step": 9151 + }, + { + "epoch": 0.7585578118524658, + "grad_norm": 0.22236786782741547, + "learning_rate": 1.2076253626191463e-06, + "loss": 0.4706, + "step": 9152 + }, + { + "epoch": 0.7586406962287608, + "grad_norm": 0.2710731625556946, + "learning_rate": 1.207210940737671e-06, + "loss": 0.5219, + "step": 9153 + }, + { + "epoch": 0.758723580605056, + "grad_norm": 0.2618817090988159, + "learning_rate": 1.2067965188561956e-06, + "loss": 0.5283, + "step": 9154 + }, + { + "epoch": 0.7588064649813511, + "grad_norm": 0.2367786467075348, + "learning_rate": 1.2063820969747203e-06, + "loss": 0.5063, + "step": 9155 + }, + { + "epoch": 0.7588893493576461, + "grad_norm": 0.23276427388191223, + "learning_rate": 1.2059676750932451e-06, + "loss": 0.4787, + "step": 9156 + }, + { + "epoch": 0.7589722337339412, + "grad_norm": 0.264835923910141, + "learning_rate": 1.2055532532117698e-06, + "loss": 0.493, + "step": 9157 + }, + { + "epoch": 0.7590551181102362, + "grad_norm": 0.2487620711326599, + "learning_rate": 1.2051388313302944e-06, + "loss": 0.5064, + "step": 9158 + }, + { + "epoch": 0.7591380024865313, + "grad_norm": 0.24085970222949982, + "learning_rate": 1.204724409448819e-06, + "loss": 0.5089, + "step": 9159 + }, + { + "epoch": 0.7592208868628264, + "grad_norm": 0.2655286490917206, + "learning_rate": 1.2043099875673437e-06, + "loss": 0.5165, + "step": 9160 + }, + { + "epoch": 0.7593037712391214, + "grad_norm": 0.23180903494358063, + "learning_rate": 1.2038955656858683e-06, + "loss": 0.4809, + "step": 9161 + }, + { + "epoch": 0.7593866556154165, + "grad_norm": 0.26156333088874817, + "learning_rate": 1.203481143804393e-06, + "loss": 0.5251, + "step": 9162 + }, + { + "epoch": 0.7594695399917115, + "grad_norm": 0.23601670563220978, + "learning_rate": 1.2030667219229176e-06, + "loss": 0.4686, + "step": 9163 + }, + { + "epoch": 0.7595524243680066, + "grad_norm": 0.23480355739593506, + "learning_rate": 1.2026523000414423e-06, + "loss": 0.4782, + "step": 9164 + }, + { + "epoch": 0.7596353087443017, + "grad_norm": 0.24882090091705322, + "learning_rate": 1.202237878159967e-06, + "loss": 0.5046, + "step": 9165 + }, + { + "epoch": 0.7597181931205967, + "grad_norm": 0.24670158326625824, + "learning_rate": 1.2018234562784916e-06, + "loss": 0.4873, + "step": 9166 + }, + { + "epoch": 0.7598010774968919, + "grad_norm": 0.2339836210012436, + "learning_rate": 1.2014090343970162e-06, + "loss": 0.4736, + "step": 9167 + }, + { + "epoch": 0.7598839618731869, + "grad_norm": 0.23856787383556366, + "learning_rate": 1.200994612515541e-06, + "loss": 0.4741, + "step": 9168 + }, + { + "epoch": 0.759966846249482, + "grad_norm": 0.2334277331829071, + "learning_rate": 1.2005801906340655e-06, + "loss": 0.4859, + "step": 9169 + }, + { + "epoch": 0.7600497306257771, + "grad_norm": 0.2585003674030304, + "learning_rate": 1.2001657687525903e-06, + "loss": 0.4988, + "step": 9170 + }, + { + "epoch": 0.7601326150020721, + "grad_norm": 0.23612497746944427, + "learning_rate": 1.1997513468711148e-06, + "loss": 0.4734, + "step": 9171 + }, + { + "epoch": 0.7602154993783672, + "grad_norm": 0.2423248291015625, + "learning_rate": 1.1993369249896396e-06, + "loss": 0.4868, + "step": 9172 + }, + { + "epoch": 0.7602983837546623, + "grad_norm": 0.2380315512418747, + "learning_rate": 1.198922503108164e-06, + "loss": 0.491, + "step": 9173 + }, + { + "epoch": 0.7603812681309573, + "grad_norm": 0.23880723118782043, + "learning_rate": 1.198508081226689e-06, + "loss": 0.5132, + "step": 9174 + }, + { + "epoch": 0.7604641525072524, + "grad_norm": 0.23505738377571106, + "learning_rate": 1.1980936593452136e-06, + "loss": 0.4979, + "step": 9175 + }, + { + "epoch": 0.7605470368835474, + "grad_norm": 0.2771330773830414, + "learning_rate": 1.1976792374637382e-06, + "loss": 0.5386, + "step": 9176 + }, + { + "epoch": 0.7606299212598425, + "grad_norm": 0.23802414536476135, + "learning_rate": 1.1972648155822628e-06, + "loss": 0.4672, + "step": 9177 + }, + { + "epoch": 0.7607128056361376, + "grad_norm": 0.23720955848693848, + "learning_rate": 1.1968503937007875e-06, + "loss": 0.4583, + "step": 9178 + }, + { + "epoch": 0.7607956900124326, + "grad_norm": 0.26452940702438354, + "learning_rate": 1.1964359718193121e-06, + "loss": 0.4965, + "step": 9179 + }, + { + "epoch": 0.7608785743887277, + "grad_norm": 0.23794585466384888, + "learning_rate": 1.1960215499378368e-06, + "loss": 0.4855, + "step": 9180 + }, + { + "epoch": 0.7609614587650227, + "grad_norm": 0.24616864323616028, + "learning_rate": 1.1956071280563614e-06, + "loss": 0.4784, + "step": 9181 + }, + { + "epoch": 0.7610443431413179, + "grad_norm": 0.23738454282283783, + "learning_rate": 1.195192706174886e-06, + "loss": 0.4843, + "step": 9182 + }, + { + "epoch": 0.761127227517613, + "grad_norm": 0.25731298327445984, + "learning_rate": 1.1947782842934107e-06, + "loss": 0.4981, + "step": 9183 + }, + { + "epoch": 0.761210111893908, + "grad_norm": 0.24687010049819946, + "learning_rate": 1.1943638624119354e-06, + "loss": 0.4785, + "step": 9184 + }, + { + "epoch": 0.7612929962702031, + "grad_norm": 0.24006077647209167, + "learning_rate": 1.1939494405304602e-06, + "loss": 0.4986, + "step": 9185 + }, + { + "epoch": 0.7613758806464981, + "grad_norm": 0.2559702396392822, + "learning_rate": 1.1935350186489849e-06, + "loss": 0.474, + "step": 9186 + }, + { + "epoch": 0.7614587650227932, + "grad_norm": 0.2580423355102539, + "learning_rate": 1.1931205967675095e-06, + "loss": 0.4867, + "step": 9187 + }, + { + "epoch": 0.7615416493990883, + "grad_norm": 0.24570243060588837, + "learning_rate": 1.1927061748860341e-06, + "loss": 0.5007, + "step": 9188 + }, + { + "epoch": 0.7616245337753833, + "grad_norm": 0.2378951907157898, + "learning_rate": 1.1922917530045588e-06, + "loss": 0.4733, + "step": 9189 + }, + { + "epoch": 0.7617074181516784, + "grad_norm": 0.22747890651226044, + "learning_rate": 1.1918773311230834e-06, + "loss": 0.4863, + "step": 9190 + }, + { + "epoch": 0.7617903025279735, + "grad_norm": 0.24329166114330292, + "learning_rate": 1.191462909241608e-06, + "loss": 0.4865, + "step": 9191 + }, + { + "epoch": 0.7618731869042685, + "grad_norm": 0.26068904995918274, + "learning_rate": 1.1910484873601327e-06, + "loss": 0.5016, + "step": 9192 + }, + { + "epoch": 0.7619560712805636, + "grad_norm": 0.25666338205337524, + "learning_rate": 1.1906340654786574e-06, + "loss": 0.4912, + "step": 9193 + }, + { + "epoch": 0.7620389556568586, + "grad_norm": 0.2628639042377472, + "learning_rate": 1.190219643597182e-06, + "loss": 0.5692, + "step": 9194 + }, + { + "epoch": 0.7621218400331538, + "grad_norm": 0.27018630504608154, + "learning_rate": 1.1898052217157069e-06, + "loss": 0.512, + "step": 9195 + }, + { + "epoch": 0.7622047244094489, + "grad_norm": 0.25288334488868713, + "learning_rate": 1.1893907998342313e-06, + "loss": 0.4689, + "step": 9196 + }, + { + "epoch": 0.7622876087857439, + "grad_norm": 0.2450505495071411, + "learning_rate": 1.1889763779527561e-06, + "loss": 0.5005, + "step": 9197 + }, + { + "epoch": 0.762370493162039, + "grad_norm": 0.24015545845031738, + "learning_rate": 1.1885619560712806e-06, + "loss": 0.4834, + "step": 9198 + }, + { + "epoch": 0.762453377538334, + "grad_norm": 0.23827053606510162, + "learning_rate": 1.1881475341898054e-06, + "loss": 0.4991, + "step": 9199 + }, + { + "epoch": 0.7625362619146291, + "grad_norm": 0.23692238330841064, + "learning_rate": 1.1877331123083299e-06, + "loss": 0.5135, + "step": 9200 + }, + { + "epoch": 0.7626191462909242, + "grad_norm": 0.247190460562706, + "learning_rate": 1.1873186904268547e-06, + "loss": 0.5226, + "step": 9201 + }, + { + "epoch": 0.7627020306672192, + "grad_norm": 0.2341517210006714, + "learning_rate": 1.1869042685453791e-06, + "loss": 0.4786, + "step": 9202 + }, + { + "epoch": 0.7627849150435143, + "grad_norm": 0.2350131720304489, + "learning_rate": 1.186489846663904e-06, + "loss": 0.4854, + "step": 9203 + }, + { + "epoch": 0.7628677994198093, + "grad_norm": 0.24948467314243317, + "learning_rate": 1.1860754247824286e-06, + "loss": 0.4858, + "step": 9204 + }, + { + "epoch": 0.7629506837961044, + "grad_norm": 0.24404959380626678, + "learning_rate": 1.1856610029009533e-06, + "loss": 0.5033, + "step": 9205 + }, + { + "epoch": 0.7630335681723995, + "grad_norm": 0.2487054169178009, + "learning_rate": 1.185246581019478e-06, + "loss": 0.5234, + "step": 9206 + }, + { + "epoch": 0.7631164525486945, + "grad_norm": 0.23707905411720276, + "learning_rate": 1.1848321591380026e-06, + "loss": 0.4764, + "step": 9207 + }, + { + "epoch": 0.7631993369249896, + "grad_norm": 0.24265176057815552, + "learning_rate": 1.1844177372565272e-06, + "loss": 0.5022, + "step": 9208 + }, + { + "epoch": 0.7632822213012848, + "grad_norm": 0.24104395508766174, + "learning_rate": 1.1840033153750519e-06, + "loss": 0.4965, + "step": 9209 + }, + { + "epoch": 0.7633651056775798, + "grad_norm": 0.27526411414146423, + "learning_rate": 1.1835888934935765e-06, + "loss": 0.5115, + "step": 9210 + }, + { + "epoch": 0.7634479900538749, + "grad_norm": 0.23372229933738708, + "learning_rate": 1.1831744716121012e-06, + "loss": 0.4886, + "step": 9211 + }, + { + "epoch": 0.7635308744301699, + "grad_norm": 0.2657921016216278, + "learning_rate": 1.1827600497306258e-06, + "loss": 0.5442, + "step": 9212 + }, + { + "epoch": 0.763613758806465, + "grad_norm": 0.24271360039710999, + "learning_rate": 1.1823456278491504e-06, + "loss": 0.4929, + "step": 9213 + }, + { + "epoch": 0.7636966431827601, + "grad_norm": 0.2877947688102722, + "learning_rate": 1.1819312059676753e-06, + "loss": 0.487, + "step": 9214 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 0.25094595551490784, + "learning_rate": 1.1815167840862e-06, + "loss": 0.5139, + "step": 9215 + }, + { + "epoch": 0.7638624119353502, + "grad_norm": 0.2509172558784485, + "learning_rate": 1.1811023622047246e-06, + "loss": 0.5109, + "step": 9216 + }, + { + "epoch": 0.7639452963116452, + "grad_norm": 0.25364288687705994, + "learning_rate": 1.1806879403232492e-06, + "loss": 0.51, + "step": 9217 + }, + { + "epoch": 0.7640281806879403, + "grad_norm": 0.2460564225912094, + "learning_rate": 1.1802735184417739e-06, + "loss": 0.5066, + "step": 9218 + }, + { + "epoch": 0.7641110650642354, + "grad_norm": 0.24663135409355164, + "learning_rate": 1.1798590965602985e-06, + "loss": 0.5096, + "step": 9219 + }, + { + "epoch": 0.7641939494405304, + "grad_norm": 0.24855755269527435, + "learning_rate": 1.1794446746788232e-06, + "loss": 0.5332, + "step": 9220 + }, + { + "epoch": 0.7642768338168255, + "grad_norm": 0.24642564356327057, + "learning_rate": 1.1790302527973478e-06, + "loss": 0.498, + "step": 9221 + }, + { + "epoch": 0.7643597181931207, + "grad_norm": 0.2530307471752167, + "learning_rate": 1.1786158309158724e-06, + "loss": 0.5237, + "step": 9222 + }, + { + "epoch": 0.7644426025694157, + "grad_norm": 0.24985016882419586, + "learning_rate": 1.178201409034397e-06, + "loss": 0.5295, + "step": 9223 + }, + { + "epoch": 0.7645254869457108, + "grad_norm": 0.22930946946144104, + "learning_rate": 1.1777869871529217e-06, + "loss": 0.5238, + "step": 9224 + }, + { + "epoch": 0.7646083713220058, + "grad_norm": 0.23174813389778137, + "learning_rate": 1.1773725652714464e-06, + "loss": 0.4647, + "step": 9225 + }, + { + "epoch": 0.7646912556983009, + "grad_norm": 0.24882082641124725, + "learning_rate": 1.1769581433899712e-06, + "loss": 0.4799, + "step": 9226 + }, + { + "epoch": 0.764774140074596, + "grad_norm": 0.24946004152297974, + "learning_rate": 1.1765437215084957e-06, + "loss": 0.4859, + "step": 9227 + }, + { + "epoch": 0.764857024450891, + "grad_norm": 0.25122207403182983, + "learning_rate": 1.1761292996270205e-06, + "loss": 0.4584, + "step": 9228 + }, + { + "epoch": 0.7649399088271861, + "grad_norm": 0.26788198947906494, + "learning_rate": 1.175714877745545e-06, + "loss": 0.564, + "step": 9229 + }, + { + "epoch": 0.7650227932034811, + "grad_norm": 0.2519264221191406, + "learning_rate": 1.1753004558640698e-06, + "loss": 0.5197, + "step": 9230 + }, + { + "epoch": 0.7651056775797762, + "grad_norm": 0.22881591320037842, + "learning_rate": 1.1748860339825942e-06, + "loss": 0.4581, + "step": 9231 + }, + { + "epoch": 0.7651885619560713, + "grad_norm": 0.22939641773700714, + "learning_rate": 1.174471612101119e-06, + "loss": 0.4774, + "step": 9232 + }, + { + "epoch": 0.7652714463323663, + "grad_norm": 0.24176281690597534, + "learning_rate": 1.1740571902196437e-06, + "loss": 0.4936, + "step": 9233 + }, + { + "epoch": 0.7653543307086614, + "grad_norm": 0.23821131885051727, + "learning_rate": 1.1736427683381684e-06, + "loss": 0.5164, + "step": 9234 + }, + { + "epoch": 0.7654372150849564, + "grad_norm": 0.25240060687065125, + "learning_rate": 1.173228346456693e-06, + "loss": 0.4575, + "step": 9235 + }, + { + "epoch": 0.7655200994612515, + "grad_norm": 0.2731255888938904, + "learning_rate": 1.1728139245752177e-06, + "loss": 0.5034, + "step": 9236 + }, + { + "epoch": 0.7656029838375467, + "grad_norm": 0.24872006475925446, + "learning_rate": 1.1723995026937423e-06, + "loss": 0.5054, + "step": 9237 + }, + { + "epoch": 0.7656858682138417, + "grad_norm": 0.25831592082977295, + "learning_rate": 1.171985080812267e-06, + "loss": 0.5062, + "step": 9238 + }, + { + "epoch": 0.7657687525901368, + "grad_norm": 0.242235004901886, + "learning_rate": 1.1715706589307916e-06, + "loss": 0.491, + "step": 9239 + }, + { + "epoch": 0.7658516369664319, + "grad_norm": 0.23987017571926117, + "learning_rate": 1.1711562370493162e-06, + "loss": 0.5124, + "step": 9240 + }, + { + "epoch": 0.7659345213427269, + "grad_norm": 0.23079460859298706, + "learning_rate": 1.1707418151678409e-06, + "loss": 0.4992, + "step": 9241 + }, + { + "epoch": 0.766017405719022, + "grad_norm": 0.24415886402130127, + "learning_rate": 1.1703273932863655e-06, + "loss": 0.5617, + "step": 9242 + }, + { + "epoch": 0.766100290095317, + "grad_norm": 0.22809173166751862, + "learning_rate": 1.1699129714048902e-06, + "loss": 0.5016, + "step": 9243 + }, + { + "epoch": 0.7661831744716121, + "grad_norm": 0.23956342041492462, + "learning_rate": 1.169498549523415e-06, + "loss": 0.4877, + "step": 9244 + }, + { + "epoch": 0.7662660588479072, + "grad_norm": 0.25007525086402893, + "learning_rate": 1.1690841276419397e-06, + "loss": 0.5099, + "step": 9245 + }, + { + "epoch": 0.7663489432242022, + "grad_norm": 0.25718286633491516, + "learning_rate": 1.1686697057604643e-06, + "loss": 0.4702, + "step": 9246 + }, + { + "epoch": 0.7664318276004973, + "grad_norm": 0.2524026334285736, + "learning_rate": 1.168255283878989e-06, + "loss": 0.4993, + "step": 9247 + }, + { + "epoch": 0.7665147119767923, + "grad_norm": 0.26174604892730713, + "learning_rate": 1.1678408619975136e-06, + "loss": 0.503, + "step": 9248 + }, + { + "epoch": 0.7665975963530874, + "grad_norm": 0.24215178191661835, + "learning_rate": 1.1674264401160382e-06, + "loss": 0.511, + "step": 9249 + }, + { + "epoch": 0.7666804807293826, + "grad_norm": 0.24811623990535736, + "learning_rate": 1.1670120182345629e-06, + "loss": 0.4964, + "step": 9250 + }, + { + "epoch": 0.7667633651056776, + "grad_norm": 0.24560591578483582, + "learning_rate": 1.1665975963530875e-06, + "loss": 0.5084, + "step": 9251 + }, + { + "epoch": 0.7668462494819727, + "grad_norm": 0.23888777196407318, + "learning_rate": 1.1661831744716122e-06, + "loss": 0.5017, + "step": 9252 + }, + { + "epoch": 0.7669291338582677, + "grad_norm": 0.24735493957996368, + "learning_rate": 1.1657687525901368e-06, + "loss": 0.5254, + "step": 9253 + }, + { + "epoch": 0.7670120182345628, + "grad_norm": 0.24391913414001465, + "learning_rate": 1.1653543307086615e-06, + "loss": 0.5003, + "step": 9254 + }, + { + "epoch": 0.7670949026108579, + "grad_norm": 0.2516050338745117, + "learning_rate": 1.1649399088271863e-06, + "loss": 0.4938, + "step": 9255 + }, + { + "epoch": 0.7671777869871529, + "grad_norm": 0.24988025426864624, + "learning_rate": 1.1645254869457107e-06, + "loss": 0.4295, + "step": 9256 + }, + { + "epoch": 0.767260671363448, + "grad_norm": 0.23794470727443695, + "learning_rate": 1.1641110650642356e-06, + "loss": 0.5051, + "step": 9257 + }, + { + "epoch": 0.7673435557397431, + "grad_norm": 0.2460934966802597, + "learning_rate": 1.16369664318276e-06, + "loss": 0.4839, + "step": 9258 + }, + { + "epoch": 0.7674264401160381, + "grad_norm": 0.2364417016506195, + "learning_rate": 1.1632822213012849e-06, + "loss": 0.4924, + "step": 9259 + }, + { + "epoch": 0.7675093244923332, + "grad_norm": 0.2309630811214447, + "learning_rate": 1.1628677994198093e-06, + "loss": 0.4685, + "step": 9260 + }, + { + "epoch": 0.7675922088686282, + "grad_norm": 0.24698850512504578, + "learning_rate": 1.1624533775383342e-06, + "loss": 0.4898, + "step": 9261 + }, + { + "epoch": 0.7676750932449233, + "grad_norm": 0.2344154268503189, + "learning_rate": 1.1620389556568588e-06, + "loss": 0.4615, + "step": 9262 + }, + { + "epoch": 0.7677579776212184, + "grad_norm": 0.23449325561523438, + "learning_rate": 1.1616245337753835e-06, + "loss": 0.4811, + "step": 9263 + }, + { + "epoch": 0.7678408619975134, + "grad_norm": 0.25522154569625854, + "learning_rate": 1.1612101118939081e-06, + "loss": 0.5303, + "step": 9264 + }, + { + "epoch": 0.7679237463738086, + "grad_norm": 0.23398564755916595, + "learning_rate": 1.1607956900124328e-06, + "loss": 0.5031, + "step": 9265 + }, + { + "epoch": 0.7680066307501036, + "grad_norm": 0.24071550369262695, + "learning_rate": 1.1603812681309574e-06, + "loss": 0.4686, + "step": 9266 + }, + { + "epoch": 0.7680895151263987, + "grad_norm": 0.2534340023994446, + "learning_rate": 1.159966846249482e-06, + "loss": 0.4944, + "step": 9267 + }, + { + "epoch": 0.7681723995026938, + "grad_norm": 0.24109332263469696, + "learning_rate": 1.1595524243680067e-06, + "loss": 0.4954, + "step": 9268 + }, + { + "epoch": 0.7682552838789888, + "grad_norm": 0.23931801319122314, + "learning_rate": 1.1591380024865313e-06, + "loss": 0.4988, + "step": 9269 + }, + { + "epoch": 0.7683381682552839, + "grad_norm": 0.2344961315393448, + "learning_rate": 1.158723580605056e-06, + "loss": 0.4624, + "step": 9270 + }, + { + "epoch": 0.7684210526315789, + "grad_norm": 0.25632956624031067, + "learning_rate": 1.1583091587235808e-06, + "loss": 0.5032, + "step": 9271 + }, + { + "epoch": 0.768503937007874, + "grad_norm": 0.2289954572916031, + "learning_rate": 1.1578947368421053e-06, + "loss": 0.4902, + "step": 9272 + }, + { + "epoch": 0.7685868213841691, + "grad_norm": 0.22268109023571014, + "learning_rate": 1.1574803149606301e-06, + "loss": 0.4593, + "step": 9273 + }, + { + "epoch": 0.7686697057604641, + "grad_norm": 0.2588769495487213, + "learning_rate": 1.1570658930791548e-06, + "loss": 0.5125, + "step": 9274 + }, + { + "epoch": 0.7687525901367592, + "grad_norm": 0.2516871392726898, + "learning_rate": 1.1566514711976794e-06, + "loss": 0.5108, + "step": 9275 + }, + { + "epoch": 0.7688354745130543, + "grad_norm": 0.23603768646717072, + "learning_rate": 1.156237049316204e-06, + "loss": 0.4735, + "step": 9276 + }, + { + "epoch": 0.7689183588893493, + "grad_norm": 0.233668252825737, + "learning_rate": 1.1558226274347287e-06, + "loss": 0.5035, + "step": 9277 + }, + { + "epoch": 0.7690012432656445, + "grad_norm": 0.22065499424934387, + "learning_rate": 1.1554082055532533e-06, + "loss": 0.4515, + "step": 9278 + }, + { + "epoch": 0.7690841276419395, + "grad_norm": 0.24007974565029144, + "learning_rate": 1.154993783671778e-06, + "loss": 0.496, + "step": 9279 + }, + { + "epoch": 0.7691670120182346, + "grad_norm": 0.2355131208896637, + "learning_rate": 1.1545793617903026e-06, + "loss": 0.4871, + "step": 9280 + }, + { + "epoch": 0.7692498963945297, + "grad_norm": 0.23993384838104248, + "learning_rate": 1.1541649399088273e-06, + "loss": 0.455, + "step": 9281 + }, + { + "epoch": 0.7693327807708247, + "grad_norm": 0.2532583475112915, + "learning_rate": 1.153750518027352e-06, + "loss": 0.5079, + "step": 9282 + }, + { + "epoch": 0.7694156651471198, + "grad_norm": 0.24801649153232574, + "learning_rate": 1.1533360961458765e-06, + "loss": 0.4825, + "step": 9283 + }, + { + "epoch": 0.7694985495234148, + "grad_norm": 0.23578152060508728, + "learning_rate": 1.1529216742644014e-06, + "loss": 0.5214, + "step": 9284 + }, + { + "epoch": 0.7695814338997099, + "grad_norm": 0.23959709703922272, + "learning_rate": 1.1525072523829258e-06, + "loss": 0.4952, + "step": 9285 + }, + { + "epoch": 0.769664318276005, + "grad_norm": 0.22971096634864807, + "learning_rate": 1.1520928305014507e-06, + "loss": 0.4854, + "step": 9286 + }, + { + "epoch": 0.7697472026523, + "grad_norm": 0.25272753834724426, + "learning_rate": 1.1516784086199751e-06, + "loss": 0.494, + "step": 9287 + }, + { + "epoch": 0.7698300870285951, + "grad_norm": 0.25268957018852234, + "learning_rate": 1.1512639867385e-06, + "loss": 0.5207, + "step": 9288 + }, + { + "epoch": 0.7699129714048901, + "grad_norm": 0.23029378056526184, + "learning_rate": 1.1508495648570244e-06, + "loss": 0.4781, + "step": 9289 + }, + { + "epoch": 0.7699958557811852, + "grad_norm": 0.2631181478500366, + "learning_rate": 1.1504351429755493e-06, + "loss": 0.5457, + "step": 9290 + }, + { + "epoch": 0.7700787401574803, + "grad_norm": 0.2499530017375946, + "learning_rate": 1.150020721094074e-06, + "loss": 0.4939, + "step": 9291 + }, + { + "epoch": 0.7701616245337753, + "grad_norm": 0.25667575001716614, + "learning_rate": 1.1496062992125985e-06, + "loss": 0.5236, + "step": 9292 + }, + { + "epoch": 0.7702445089100705, + "grad_norm": 0.24108105897903442, + "learning_rate": 1.1491918773311232e-06, + "loss": 0.4973, + "step": 9293 + }, + { + "epoch": 0.7703273932863656, + "grad_norm": 0.24807089567184448, + "learning_rate": 1.1487774554496478e-06, + "loss": 0.514, + "step": 9294 + }, + { + "epoch": 0.7704102776626606, + "grad_norm": 0.24497400224208832, + "learning_rate": 1.1483630335681725e-06, + "loss": 0.5328, + "step": 9295 + }, + { + "epoch": 0.7704931620389557, + "grad_norm": 0.23853799700737, + "learning_rate": 1.1479486116866971e-06, + "loss": 0.462, + "step": 9296 + }, + { + "epoch": 0.7705760464152507, + "grad_norm": 0.23615938425064087, + "learning_rate": 1.1475341898052218e-06, + "loss": 0.5107, + "step": 9297 + }, + { + "epoch": 0.7706589307915458, + "grad_norm": 0.24373123049736023, + "learning_rate": 1.1471197679237464e-06, + "loss": 0.5141, + "step": 9298 + }, + { + "epoch": 0.7707418151678409, + "grad_norm": 0.23930108547210693, + "learning_rate": 1.146705346042271e-06, + "loss": 0.4982, + "step": 9299 + }, + { + "epoch": 0.7708246995441359, + "grad_norm": 0.2357410341501236, + "learning_rate": 1.146290924160796e-06, + "loss": 0.5176, + "step": 9300 + }, + { + "epoch": 0.770907583920431, + "grad_norm": 0.25173816084861755, + "learning_rate": 1.1458765022793203e-06, + "loss": 0.5025, + "step": 9301 + }, + { + "epoch": 0.770990468296726, + "grad_norm": 0.2378874272108078, + "learning_rate": 1.1454620803978452e-06, + "loss": 0.4812, + "step": 9302 + }, + { + "epoch": 0.7710733526730211, + "grad_norm": 0.2428276240825653, + "learning_rate": 1.1450476585163696e-06, + "loss": 0.5002, + "step": 9303 + }, + { + "epoch": 0.7711562370493162, + "grad_norm": 0.2396981120109558, + "learning_rate": 1.1446332366348945e-06, + "loss": 0.502, + "step": 9304 + }, + { + "epoch": 0.7712391214256112, + "grad_norm": 0.23433353006839752, + "learning_rate": 1.1442188147534191e-06, + "loss": 0.4774, + "step": 9305 + }, + { + "epoch": 0.7713220058019064, + "grad_norm": 0.2289823740720749, + "learning_rate": 1.1438043928719438e-06, + "loss": 0.4554, + "step": 9306 + }, + { + "epoch": 0.7714048901782015, + "grad_norm": 0.25330281257629395, + "learning_rate": 1.1433899709904684e-06, + "loss": 0.5025, + "step": 9307 + }, + { + "epoch": 0.7714877745544965, + "grad_norm": 0.23620390892028809, + "learning_rate": 1.142975549108993e-06, + "loss": 0.5132, + "step": 9308 + }, + { + "epoch": 0.7715706589307916, + "grad_norm": 0.24487461149692535, + "learning_rate": 1.1425611272275177e-06, + "loss": 0.4757, + "step": 9309 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 0.25213509798049927, + "learning_rate": 1.1421467053460423e-06, + "loss": 0.5382, + "step": 9310 + }, + { + "epoch": 0.7717364276833817, + "grad_norm": 0.25285211205482483, + "learning_rate": 1.141732283464567e-06, + "loss": 0.5016, + "step": 9311 + }, + { + "epoch": 0.7718193120596768, + "grad_norm": 0.2517211139202118, + "learning_rate": 1.1413178615830916e-06, + "loss": 0.4965, + "step": 9312 + }, + { + "epoch": 0.7719021964359718, + "grad_norm": 0.25905048847198486, + "learning_rate": 1.1409034397016163e-06, + "loss": 0.5062, + "step": 9313 + }, + { + "epoch": 0.7719850808122669, + "grad_norm": 0.2642914950847626, + "learning_rate": 1.140489017820141e-06, + "loss": 0.5329, + "step": 9314 + }, + { + "epoch": 0.7720679651885619, + "grad_norm": 0.253807008266449, + "learning_rate": 1.1400745959386658e-06, + "loss": 0.5178, + "step": 9315 + }, + { + "epoch": 0.772150849564857, + "grad_norm": 0.250356525182724, + "learning_rate": 1.1396601740571902e-06, + "loss": 0.4683, + "step": 9316 + }, + { + "epoch": 0.7722337339411521, + "grad_norm": 0.24572667479515076, + "learning_rate": 1.139245752175715e-06, + "loss": 0.5093, + "step": 9317 + }, + { + "epoch": 0.7723166183174471, + "grad_norm": 0.24099791049957275, + "learning_rate": 1.1388313302942395e-06, + "loss": 0.5078, + "step": 9318 + }, + { + "epoch": 0.7723995026937422, + "grad_norm": 0.2362409085035324, + "learning_rate": 1.1384169084127643e-06, + "loss": 0.4803, + "step": 9319 + }, + { + "epoch": 0.7724823870700372, + "grad_norm": 0.24370311200618744, + "learning_rate": 1.138002486531289e-06, + "loss": 0.5318, + "step": 9320 + }, + { + "epoch": 0.7725652714463324, + "grad_norm": 0.23114939033985138, + "learning_rate": 1.1375880646498136e-06, + "loss": 0.484, + "step": 9321 + }, + { + "epoch": 0.7726481558226275, + "grad_norm": 0.24758726358413696, + "learning_rate": 1.1371736427683383e-06, + "loss": 0.4933, + "step": 9322 + }, + { + "epoch": 0.7727310401989225, + "grad_norm": 0.2617655396461487, + "learning_rate": 1.136759220886863e-06, + "loss": 0.5369, + "step": 9323 + }, + { + "epoch": 0.7728139245752176, + "grad_norm": 0.24861524999141693, + "learning_rate": 1.1363447990053876e-06, + "loss": 0.5101, + "step": 9324 + }, + { + "epoch": 0.7728968089515127, + "grad_norm": 0.23225034773349762, + "learning_rate": 1.1359303771239122e-06, + "loss": 0.4917, + "step": 9325 + }, + { + "epoch": 0.7729796933278077, + "grad_norm": 0.235310897231102, + "learning_rate": 1.1355159552424369e-06, + "loss": 0.5023, + "step": 9326 + }, + { + "epoch": 0.7730625777041028, + "grad_norm": 0.24161064624786377, + "learning_rate": 1.1351015333609615e-06, + "loss": 0.5126, + "step": 9327 + }, + { + "epoch": 0.7731454620803978, + "grad_norm": 0.2733026444911957, + "learning_rate": 1.1346871114794861e-06, + "loss": 0.5358, + "step": 9328 + }, + { + "epoch": 0.7732283464566929, + "grad_norm": 0.2382124811410904, + "learning_rate": 1.134272689598011e-06, + "loss": 0.5035, + "step": 9329 + }, + { + "epoch": 0.773311230832988, + "grad_norm": 0.24575550854206085, + "learning_rate": 1.1338582677165354e-06, + "loss": 0.4795, + "step": 9330 + }, + { + "epoch": 0.773394115209283, + "grad_norm": 0.25066816806793213, + "learning_rate": 1.1334438458350603e-06, + "loss": 0.5362, + "step": 9331 + }, + { + "epoch": 0.7734769995855781, + "grad_norm": 0.2611173391342163, + "learning_rate": 1.1330294239535847e-06, + "loss": 0.559, + "step": 9332 + }, + { + "epoch": 0.7735598839618731, + "grad_norm": 0.24492305517196655, + "learning_rate": 1.1326150020721096e-06, + "loss": 0.512, + "step": 9333 + }, + { + "epoch": 0.7736427683381683, + "grad_norm": 0.23265235126018524, + "learning_rate": 1.1322005801906342e-06, + "loss": 0.4942, + "step": 9334 + }, + { + "epoch": 0.7737256527144634, + "grad_norm": 0.24503543972969055, + "learning_rate": 1.1317861583091589e-06, + "loss": 0.4866, + "step": 9335 + }, + { + "epoch": 0.7738085370907584, + "grad_norm": 0.2604057490825653, + "learning_rate": 1.1313717364276835e-06, + "loss": 0.5265, + "step": 9336 + }, + { + "epoch": 0.7738914214670535, + "grad_norm": 0.25068771839141846, + "learning_rate": 1.1309573145462081e-06, + "loss": 0.497, + "step": 9337 + }, + { + "epoch": 0.7739743058433485, + "grad_norm": 0.24063058197498322, + "learning_rate": 1.1305428926647328e-06, + "loss": 0.451, + "step": 9338 + }, + { + "epoch": 0.7740571902196436, + "grad_norm": 0.24957042932510376, + "learning_rate": 1.1301284707832574e-06, + "loss": 0.5259, + "step": 9339 + }, + { + "epoch": 0.7741400745959387, + "grad_norm": 0.24133183062076569, + "learning_rate": 1.129714048901782e-06, + "loss": 0.5139, + "step": 9340 + }, + { + "epoch": 0.7742229589722337, + "grad_norm": 0.24416254460811615, + "learning_rate": 1.1292996270203067e-06, + "loss": 0.4852, + "step": 9341 + }, + { + "epoch": 0.7743058433485288, + "grad_norm": 0.24122752249240875, + "learning_rate": 1.1288852051388314e-06, + "loss": 0.4675, + "step": 9342 + }, + { + "epoch": 0.7743887277248239, + "grad_norm": 0.23473148047924042, + "learning_rate": 1.128470783257356e-06, + "loss": 0.4841, + "step": 9343 + }, + { + "epoch": 0.7744716121011189, + "grad_norm": 0.251323938369751, + "learning_rate": 1.1280563613758809e-06, + "loss": 0.5202, + "step": 9344 + }, + { + "epoch": 0.774554496477414, + "grad_norm": 0.23937253654003143, + "learning_rate": 1.1276419394944053e-06, + "loss": 0.4907, + "step": 9345 + }, + { + "epoch": 0.774637380853709, + "grad_norm": 0.23251225054264069, + "learning_rate": 1.1272275176129301e-06, + "loss": 0.4612, + "step": 9346 + }, + { + "epoch": 0.7747202652300041, + "grad_norm": 0.24167227745056152, + "learning_rate": 1.1268130957314548e-06, + "loss": 0.501, + "step": 9347 + }, + { + "epoch": 0.7748031496062993, + "grad_norm": 0.22805319726467133, + "learning_rate": 1.1263986738499794e-06, + "loss": 0.4734, + "step": 9348 + }, + { + "epoch": 0.7748860339825943, + "grad_norm": 0.23557323217391968, + "learning_rate": 1.125984251968504e-06, + "loss": 0.4798, + "step": 9349 + }, + { + "epoch": 0.7749689183588894, + "grad_norm": 0.24108010530471802, + "learning_rate": 1.1255698300870287e-06, + "loss": 0.4533, + "step": 9350 + }, + { + "epoch": 0.7750518027351844, + "grad_norm": 0.2704063057899475, + "learning_rate": 1.1251554082055534e-06, + "loss": 0.5188, + "step": 9351 + }, + { + "epoch": 0.7751346871114795, + "grad_norm": 0.2984687089920044, + "learning_rate": 1.124740986324078e-06, + "loss": 0.4724, + "step": 9352 + }, + { + "epoch": 0.7752175714877746, + "grad_norm": 0.22989103198051453, + "learning_rate": 1.1243265644426027e-06, + "loss": 0.4922, + "step": 9353 + }, + { + "epoch": 0.7753004558640696, + "grad_norm": 0.24795514345169067, + "learning_rate": 1.1239121425611273e-06, + "loss": 0.5557, + "step": 9354 + }, + { + "epoch": 0.7753833402403647, + "grad_norm": 0.25297367572784424, + "learning_rate": 1.123497720679652e-06, + "loss": 0.5138, + "step": 9355 + }, + { + "epoch": 0.7754662246166597, + "grad_norm": 0.2600744664669037, + "learning_rate": 1.1230832987981766e-06, + "loss": 0.5106, + "step": 9356 + }, + { + "epoch": 0.7755491089929548, + "grad_norm": 0.24846526980400085, + "learning_rate": 1.1226688769167012e-06, + "loss": 0.4965, + "step": 9357 + }, + { + "epoch": 0.7756319933692499, + "grad_norm": 0.25480952858924866, + "learning_rate": 1.122254455035226e-06, + "loss": 0.5018, + "step": 9358 + }, + { + "epoch": 0.7757148777455449, + "grad_norm": 0.2619912028312683, + "learning_rate": 1.1218400331537505e-06, + "loss": 0.5018, + "step": 9359 + }, + { + "epoch": 0.77579776212184, + "grad_norm": 0.25107595324516296, + "learning_rate": 1.1214256112722754e-06, + "loss": 0.4807, + "step": 9360 + }, + { + "epoch": 0.7758806464981352, + "grad_norm": 0.2565012276172638, + "learning_rate": 1.1210111893907998e-06, + "loss": 0.4943, + "step": 9361 + }, + { + "epoch": 0.7759635308744302, + "grad_norm": 0.25855308771133423, + "learning_rate": 1.1205967675093247e-06, + "loss": 0.516, + "step": 9362 + }, + { + "epoch": 0.7760464152507253, + "grad_norm": 0.24712830781936646, + "learning_rate": 1.120182345627849e-06, + "loss": 0.4994, + "step": 9363 + }, + { + "epoch": 0.7761292996270203, + "grad_norm": 0.24172236025333405, + "learning_rate": 1.119767923746374e-06, + "loss": 0.4968, + "step": 9364 + }, + { + "epoch": 0.7762121840033154, + "grad_norm": 0.2347486913204193, + "learning_rate": 1.1193535018648986e-06, + "loss": 0.4753, + "step": 9365 + }, + { + "epoch": 0.7762950683796105, + "grad_norm": 0.24524784088134766, + "learning_rate": 1.1189390799834232e-06, + "loss": 0.5024, + "step": 9366 + }, + { + "epoch": 0.7763779527559055, + "grad_norm": 0.23354105651378632, + "learning_rate": 1.1185246581019479e-06, + "loss": 0.4711, + "step": 9367 + }, + { + "epoch": 0.7764608371322006, + "grad_norm": 0.2502935826778412, + "learning_rate": 1.1181102362204725e-06, + "loss": 0.4851, + "step": 9368 + }, + { + "epoch": 0.7765437215084956, + "grad_norm": 0.24188871681690216, + "learning_rate": 1.1176958143389972e-06, + "loss": 0.535, + "step": 9369 + }, + { + "epoch": 0.7766266058847907, + "grad_norm": 0.36861759424209595, + "learning_rate": 1.1172813924575218e-06, + "loss": 0.4958, + "step": 9370 + }, + { + "epoch": 0.7767094902610858, + "grad_norm": 0.2618878185749054, + "learning_rate": 1.1168669705760464e-06, + "loss": 0.5639, + "step": 9371 + }, + { + "epoch": 0.7767923746373808, + "grad_norm": 0.2377711683511734, + "learning_rate": 1.116452548694571e-06, + "loss": 0.4796, + "step": 9372 + }, + { + "epoch": 0.7768752590136759, + "grad_norm": 0.25268691778182983, + "learning_rate": 1.1160381268130957e-06, + "loss": 0.5328, + "step": 9373 + }, + { + "epoch": 0.776958143389971, + "grad_norm": 0.22972913086414337, + "learning_rate": 1.1156237049316204e-06, + "loss": 0.5085, + "step": 9374 + }, + { + "epoch": 0.777041027766266, + "grad_norm": 0.2600421905517578, + "learning_rate": 1.1152092830501452e-06, + "loss": 0.482, + "step": 9375 + }, + { + "epoch": 0.7771239121425612, + "grad_norm": 0.25679200887680054, + "learning_rate": 1.1147948611686699e-06, + "loss": 0.5287, + "step": 9376 + }, + { + "epoch": 0.7772067965188562, + "grad_norm": 0.23455940186977386, + "learning_rate": 1.1143804392871945e-06, + "loss": 0.4651, + "step": 9377 + }, + { + "epoch": 0.7772896808951513, + "grad_norm": 0.25783565640449524, + "learning_rate": 1.1139660174057192e-06, + "loss": 0.4963, + "step": 9378 + }, + { + "epoch": 0.7773725652714464, + "grad_norm": 0.23500429093837738, + "learning_rate": 1.1135515955242438e-06, + "loss": 0.5215, + "step": 9379 + }, + { + "epoch": 0.7774554496477414, + "grad_norm": 0.22077834606170654, + "learning_rate": 1.1131371736427685e-06, + "loss": 0.467, + "step": 9380 + }, + { + "epoch": 0.7775383340240365, + "grad_norm": 0.2622964680194855, + "learning_rate": 1.112722751761293e-06, + "loss": 0.516, + "step": 9381 + }, + { + "epoch": 0.7776212184003315, + "grad_norm": 0.23177438974380493, + "learning_rate": 1.1123083298798177e-06, + "loss": 0.503, + "step": 9382 + }, + { + "epoch": 0.7777041027766266, + "grad_norm": 0.260322242975235, + "learning_rate": 1.1118939079983424e-06, + "loss": 0.4977, + "step": 9383 + }, + { + "epoch": 0.7777869871529217, + "grad_norm": 0.2385798543691635, + "learning_rate": 1.111479486116867e-06, + "loss": 0.4871, + "step": 9384 + }, + { + "epoch": 0.7778698715292167, + "grad_norm": 0.23662222921848297, + "learning_rate": 1.1110650642353919e-06, + "loss": 0.488, + "step": 9385 + }, + { + "epoch": 0.7779527559055118, + "grad_norm": 0.23049603402614594, + "learning_rate": 1.1106506423539163e-06, + "loss": 0.4791, + "step": 9386 + }, + { + "epoch": 0.7780356402818068, + "grad_norm": 0.22874951362609863, + "learning_rate": 1.1102362204724412e-06, + "loss": 0.4459, + "step": 9387 + }, + { + "epoch": 0.7781185246581019, + "grad_norm": 0.23516514897346497, + "learning_rate": 1.1098217985909656e-06, + "loss": 0.4899, + "step": 9388 + }, + { + "epoch": 0.778201409034397, + "grad_norm": 0.2553059756755829, + "learning_rate": 1.1094073767094905e-06, + "loss": 0.5166, + "step": 9389 + }, + { + "epoch": 0.778284293410692, + "grad_norm": 0.23194657266139984, + "learning_rate": 1.1089929548280149e-06, + "loss": 0.514, + "step": 9390 + }, + { + "epoch": 0.7783671777869872, + "grad_norm": 0.24083487689495087, + "learning_rate": 1.1085785329465397e-06, + "loss": 0.5036, + "step": 9391 + }, + { + "epoch": 0.7784500621632823, + "grad_norm": 0.23263990879058838, + "learning_rate": 1.1081641110650642e-06, + "loss": 0.4727, + "step": 9392 + }, + { + "epoch": 0.7785329465395773, + "grad_norm": 0.2586837410926819, + "learning_rate": 1.107749689183589e-06, + "loss": 0.5264, + "step": 9393 + }, + { + "epoch": 0.7786158309158724, + "grad_norm": 0.22952702641487122, + "learning_rate": 1.1073352673021137e-06, + "loss": 0.4971, + "step": 9394 + }, + { + "epoch": 0.7786987152921674, + "grad_norm": 0.25251197814941406, + "learning_rate": 1.1069208454206383e-06, + "loss": 0.4968, + "step": 9395 + }, + { + "epoch": 0.7787815996684625, + "grad_norm": 0.24437880516052246, + "learning_rate": 1.106506423539163e-06, + "loss": 0.4492, + "step": 9396 + }, + { + "epoch": 0.7788644840447576, + "grad_norm": 0.23018217086791992, + "learning_rate": 1.1060920016576876e-06, + "loss": 0.481, + "step": 9397 + }, + { + "epoch": 0.7789473684210526, + "grad_norm": 0.2580467164516449, + "learning_rate": 1.1056775797762122e-06, + "loss": 0.4824, + "step": 9398 + }, + { + "epoch": 0.7790302527973477, + "grad_norm": 0.24302345514297485, + "learning_rate": 1.1052631578947369e-06, + "loss": 0.4905, + "step": 9399 + }, + { + "epoch": 0.7791131371736427, + "grad_norm": 0.26237478852272034, + "learning_rate": 1.1048487360132615e-06, + "loss": 0.5161, + "step": 9400 + }, + { + "epoch": 0.7791960215499378, + "grad_norm": 0.2334851324558258, + "learning_rate": 1.1044343141317862e-06, + "loss": 0.4659, + "step": 9401 + }, + { + "epoch": 0.779278905926233, + "grad_norm": 0.22957265377044678, + "learning_rate": 1.1040198922503108e-06, + "loss": 0.4872, + "step": 9402 + }, + { + "epoch": 0.779361790302528, + "grad_norm": 0.2475137859582901, + "learning_rate": 1.1036054703688355e-06, + "loss": 0.4882, + "step": 9403 + }, + { + "epoch": 0.7794446746788231, + "grad_norm": 0.2550935745239258, + "learning_rate": 1.1031910484873603e-06, + "loss": 0.5354, + "step": 9404 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 0.2527272403240204, + "learning_rate": 1.102776626605885e-06, + "loss": 0.4772, + "step": 9405 + }, + { + "epoch": 0.7796104434314132, + "grad_norm": 0.24778637290000916, + "learning_rate": 1.1023622047244096e-06, + "loss": 0.5098, + "step": 9406 + }, + { + "epoch": 0.7796933278077083, + "grad_norm": 0.24567101895809174, + "learning_rate": 1.1019477828429343e-06, + "loss": 0.5004, + "step": 9407 + }, + { + "epoch": 0.7797762121840033, + "grad_norm": 0.22249937057495117, + "learning_rate": 1.101533360961459e-06, + "loss": 0.4668, + "step": 9408 + }, + { + "epoch": 0.7798590965602984, + "grad_norm": 0.24558347463607788, + "learning_rate": 1.1011189390799835e-06, + "loss": 0.5122, + "step": 9409 + }, + { + "epoch": 0.7799419809365935, + "grad_norm": 0.24470919370651245, + "learning_rate": 1.1007045171985082e-06, + "loss": 0.4997, + "step": 9410 + }, + { + "epoch": 0.7800248653128885, + "grad_norm": 0.30863237380981445, + "learning_rate": 1.1002900953170328e-06, + "loss": 0.513, + "step": 9411 + }, + { + "epoch": 0.7801077496891836, + "grad_norm": 0.2580946981906891, + "learning_rate": 1.0998756734355575e-06, + "loss": 0.4965, + "step": 9412 + }, + { + "epoch": 0.7801906340654786, + "grad_norm": 0.26538515090942383, + "learning_rate": 1.0994612515540821e-06, + "loss": 0.5209, + "step": 9413 + }, + { + "epoch": 0.7802735184417737, + "grad_norm": 0.2496361881494522, + "learning_rate": 1.099046829672607e-06, + "loss": 0.4766, + "step": 9414 + }, + { + "epoch": 0.7803564028180688, + "grad_norm": 0.2472662627696991, + "learning_rate": 1.0986324077911314e-06, + "loss": 0.5133, + "step": 9415 + }, + { + "epoch": 0.7804392871943638, + "grad_norm": 0.24601225554943085, + "learning_rate": 1.0982179859096563e-06, + "loss": 0.4738, + "step": 9416 + }, + { + "epoch": 0.780522171570659, + "grad_norm": 0.24806159734725952, + "learning_rate": 1.0978035640281807e-06, + "loss": 0.5391, + "step": 9417 + }, + { + "epoch": 0.780605055946954, + "grad_norm": 0.24653050303459167, + "learning_rate": 1.0973891421467055e-06, + "loss": 0.4954, + "step": 9418 + }, + { + "epoch": 0.7806879403232491, + "grad_norm": 0.23206061124801636, + "learning_rate": 1.09697472026523e-06, + "loss": 0.4544, + "step": 9419 + }, + { + "epoch": 0.7807708246995442, + "grad_norm": 0.23619823157787323, + "learning_rate": 1.0965602983837548e-06, + "loss": 0.4675, + "step": 9420 + }, + { + "epoch": 0.7808537090758392, + "grad_norm": 0.25516921281814575, + "learning_rate": 1.0961458765022793e-06, + "loss": 0.5238, + "step": 9421 + }, + { + "epoch": 0.7809365934521343, + "grad_norm": 0.2518298625946045, + "learning_rate": 1.0957314546208041e-06, + "loss": 0.519, + "step": 9422 + }, + { + "epoch": 0.7810194778284293, + "grad_norm": 0.2660287022590637, + "learning_rate": 1.0953170327393288e-06, + "loss": 0.4966, + "step": 9423 + }, + { + "epoch": 0.7811023622047244, + "grad_norm": 0.2332509607076645, + "learning_rate": 1.0949026108578534e-06, + "loss": 0.4779, + "step": 9424 + }, + { + "epoch": 0.7811852465810195, + "grad_norm": 0.22785040736198425, + "learning_rate": 1.094488188976378e-06, + "loss": 0.4638, + "step": 9425 + }, + { + "epoch": 0.7812681309573145, + "grad_norm": 0.22573694586753845, + "learning_rate": 1.0940737670949027e-06, + "loss": 0.4918, + "step": 9426 + }, + { + "epoch": 0.7813510153336096, + "grad_norm": 0.2523157000541687, + "learning_rate": 1.0936593452134273e-06, + "loss": 0.5455, + "step": 9427 + }, + { + "epoch": 0.7814338997099047, + "grad_norm": 0.22373197972774506, + "learning_rate": 1.093244923331952e-06, + "loss": 0.4362, + "step": 9428 + }, + { + "epoch": 0.7815167840861997, + "grad_norm": 0.24962212145328522, + "learning_rate": 1.0928305014504766e-06, + "loss": 0.5005, + "step": 9429 + }, + { + "epoch": 0.7815996684624948, + "grad_norm": 0.23159143328666687, + "learning_rate": 1.0924160795690013e-06, + "loss": 0.4865, + "step": 9430 + }, + { + "epoch": 0.7816825528387898, + "grad_norm": 0.24931122362613678, + "learning_rate": 1.092001657687526e-06, + "loss": 0.5036, + "step": 9431 + }, + { + "epoch": 0.781765437215085, + "grad_norm": 0.235525980591774, + "learning_rate": 1.0915872358060506e-06, + "loss": 0.5218, + "step": 9432 + }, + { + "epoch": 0.7818483215913801, + "grad_norm": 0.23536507785320282, + "learning_rate": 1.0911728139245754e-06, + "loss": 0.5017, + "step": 9433 + }, + { + "epoch": 0.7819312059676751, + "grad_norm": 0.24605682492256165, + "learning_rate": 1.0907583920431e-06, + "loss": 0.4963, + "step": 9434 + }, + { + "epoch": 0.7820140903439702, + "grad_norm": 0.26268288493156433, + "learning_rate": 1.0903439701616247e-06, + "loss": 0.5157, + "step": 9435 + }, + { + "epoch": 0.7820969747202652, + "grad_norm": 0.24542368948459625, + "learning_rate": 1.0899295482801493e-06, + "loss": 0.4904, + "step": 9436 + }, + { + "epoch": 0.7821798590965603, + "grad_norm": 0.23626968264579773, + "learning_rate": 1.089515126398674e-06, + "loss": 0.5124, + "step": 9437 + }, + { + "epoch": 0.7822627434728554, + "grad_norm": 0.24977955222129822, + "learning_rate": 1.0891007045171986e-06, + "loss": 0.5131, + "step": 9438 + }, + { + "epoch": 0.7823456278491504, + "grad_norm": 0.5830584168434143, + "learning_rate": 1.0886862826357233e-06, + "loss": 0.4982, + "step": 9439 + }, + { + "epoch": 0.7824285122254455, + "grad_norm": 0.2458753138780594, + "learning_rate": 1.088271860754248e-06, + "loss": 0.5135, + "step": 9440 + }, + { + "epoch": 0.7825113966017405, + "grad_norm": 0.2668566405773163, + "learning_rate": 1.0878574388727726e-06, + "loss": 0.5214, + "step": 9441 + }, + { + "epoch": 0.7825942809780356, + "grad_norm": 0.24915547668933868, + "learning_rate": 1.0874430169912972e-06, + "loss": 0.4985, + "step": 9442 + }, + { + "epoch": 0.7826771653543307, + "grad_norm": 0.24721285700798035, + "learning_rate": 1.0870285951098218e-06, + "loss": 0.5151, + "step": 9443 + }, + { + "epoch": 0.7827600497306257, + "grad_norm": 0.241340771317482, + "learning_rate": 1.0866141732283465e-06, + "loss": 0.4717, + "step": 9444 + }, + { + "epoch": 0.7828429341069209, + "grad_norm": 0.2554435729980469, + "learning_rate": 1.0861997513468713e-06, + "loss": 0.5249, + "step": 9445 + }, + { + "epoch": 0.782925818483216, + "grad_norm": 0.23739685118198395, + "learning_rate": 1.0857853294653958e-06, + "loss": 0.4742, + "step": 9446 + }, + { + "epoch": 0.783008702859511, + "grad_norm": 0.26630067825317383, + "learning_rate": 1.0853709075839206e-06, + "loss": 0.5143, + "step": 9447 + }, + { + "epoch": 0.7830915872358061, + "grad_norm": 0.24893026053905487, + "learning_rate": 1.084956485702445e-06, + "loss": 0.5155, + "step": 9448 + }, + { + "epoch": 0.7831744716121011, + "grad_norm": 0.2496057003736496, + "learning_rate": 1.08454206382097e-06, + "loss": 0.5353, + "step": 9449 + }, + { + "epoch": 0.7832573559883962, + "grad_norm": 0.23453526198863983, + "learning_rate": 1.0841276419394943e-06, + "loss": 0.5146, + "step": 9450 + }, + { + "epoch": 0.7833402403646913, + "grad_norm": 0.23638878762722015, + "learning_rate": 1.0837132200580192e-06, + "loss": 0.5084, + "step": 9451 + }, + { + "epoch": 0.7834231247409863, + "grad_norm": 0.23884332180023193, + "learning_rate": 1.0832987981765438e-06, + "loss": 0.4663, + "step": 9452 + }, + { + "epoch": 0.7835060091172814, + "grad_norm": 0.23199211061000824, + "learning_rate": 1.0828843762950685e-06, + "loss": 0.5189, + "step": 9453 + }, + { + "epoch": 0.7835888934935764, + "grad_norm": 0.2544392943382263, + "learning_rate": 1.0824699544135931e-06, + "loss": 0.5008, + "step": 9454 + }, + { + "epoch": 0.7836717778698715, + "grad_norm": 0.23723939061164856, + "learning_rate": 1.0820555325321178e-06, + "loss": 0.4968, + "step": 9455 + }, + { + "epoch": 0.7837546622461666, + "grad_norm": 0.23110778629779816, + "learning_rate": 1.0816411106506424e-06, + "loss": 0.5063, + "step": 9456 + }, + { + "epoch": 0.7838375466224616, + "grad_norm": 0.2630501091480255, + "learning_rate": 1.081226688769167e-06, + "loss": 0.5046, + "step": 9457 + }, + { + "epoch": 0.7839204309987567, + "grad_norm": 0.23275026679039001, + "learning_rate": 1.0808122668876917e-06, + "loss": 0.4812, + "step": 9458 + }, + { + "epoch": 0.7840033153750519, + "grad_norm": 0.2615645229816437, + "learning_rate": 1.0803978450062164e-06, + "loss": 0.4816, + "step": 9459 + }, + { + "epoch": 0.7840861997513469, + "grad_norm": 0.24115005135536194, + "learning_rate": 1.079983423124741e-06, + "loss": 0.5136, + "step": 9460 + }, + { + "epoch": 0.784169084127642, + "grad_norm": 0.22207212448120117, + "learning_rate": 1.0795690012432659e-06, + "loss": 0.4589, + "step": 9461 + }, + { + "epoch": 0.784251968503937, + "grad_norm": 0.2332577109336853, + "learning_rate": 1.0791545793617903e-06, + "loss": 0.4825, + "step": 9462 + }, + { + "epoch": 0.7843348528802321, + "grad_norm": 0.2201392650604248, + "learning_rate": 1.0787401574803151e-06, + "loss": 0.4844, + "step": 9463 + }, + { + "epoch": 0.7844177372565272, + "grad_norm": 0.24205030500888824, + "learning_rate": 1.0783257355988398e-06, + "loss": 0.5123, + "step": 9464 + }, + { + "epoch": 0.7845006216328222, + "grad_norm": 0.24981921911239624, + "learning_rate": 1.0779113137173644e-06, + "loss": 0.5326, + "step": 9465 + }, + { + "epoch": 0.7845835060091173, + "grad_norm": 0.235609233379364, + "learning_rate": 1.077496891835889e-06, + "loss": 0.4637, + "step": 9466 + }, + { + "epoch": 0.7846663903854123, + "grad_norm": 0.2459152638912201, + "learning_rate": 1.0770824699544137e-06, + "loss": 0.5229, + "step": 9467 + }, + { + "epoch": 0.7847492747617074, + "grad_norm": 0.25946012139320374, + "learning_rate": 1.0766680480729384e-06, + "loss": 0.4985, + "step": 9468 + }, + { + "epoch": 0.7848321591380025, + "grad_norm": 0.24855715036392212, + "learning_rate": 1.076253626191463e-06, + "loss": 0.5169, + "step": 9469 + }, + { + "epoch": 0.7849150435142975, + "grad_norm": 0.230390727519989, + "learning_rate": 1.0758392043099876e-06, + "loss": 0.4673, + "step": 9470 + }, + { + "epoch": 0.7849979278905926, + "grad_norm": 0.2255980223417282, + "learning_rate": 1.0754247824285123e-06, + "loss": 0.4406, + "step": 9471 + }, + { + "epoch": 0.7850808122668876, + "grad_norm": 0.2320244461297989, + "learning_rate": 1.075010360547037e-06, + "loss": 0.4852, + "step": 9472 + }, + { + "epoch": 0.7851636966431828, + "grad_norm": 0.24201859533786774, + "learning_rate": 1.0745959386655616e-06, + "loss": 0.4793, + "step": 9473 + }, + { + "epoch": 0.7852465810194779, + "grad_norm": 0.24735641479492188, + "learning_rate": 1.0741815167840864e-06, + "loss": 0.5007, + "step": 9474 + }, + { + "epoch": 0.7853294653957729, + "grad_norm": 0.2532861828804016, + "learning_rate": 1.0737670949026109e-06, + "loss": 0.4957, + "step": 9475 + }, + { + "epoch": 0.785412349772068, + "grad_norm": 0.27934107184410095, + "learning_rate": 1.0733526730211357e-06, + "loss": 0.4789, + "step": 9476 + }, + { + "epoch": 0.7854952341483631, + "grad_norm": 0.2452678233385086, + "learning_rate": 1.0729382511396601e-06, + "loss": 0.501, + "step": 9477 + }, + { + "epoch": 0.7855781185246581, + "grad_norm": 0.2655497193336487, + "learning_rate": 1.072523829258185e-06, + "loss": 0.5273, + "step": 9478 + }, + { + "epoch": 0.7856610029009532, + "grad_norm": 0.2868204414844513, + "learning_rate": 1.0721094073767094e-06, + "loss": 0.536, + "step": 9479 + }, + { + "epoch": 0.7857438872772482, + "grad_norm": 0.24370479583740234, + "learning_rate": 1.0716949854952343e-06, + "loss": 0.5417, + "step": 9480 + }, + { + "epoch": 0.7858267716535433, + "grad_norm": 0.24106013774871826, + "learning_rate": 1.071280563613759e-06, + "loss": 0.4741, + "step": 9481 + }, + { + "epoch": 0.7859096560298384, + "grad_norm": 0.22703416645526886, + "learning_rate": 1.0708661417322836e-06, + "loss": 0.4656, + "step": 9482 + }, + { + "epoch": 0.7859925404061334, + "grad_norm": 0.25150051712989807, + "learning_rate": 1.0704517198508082e-06, + "loss": 0.4846, + "step": 9483 + }, + { + "epoch": 0.7860754247824285, + "grad_norm": 0.23460352420806885, + "learning_rate": 1.0700372979693329e-06, + "loss": 0.5136, + "step": 9484 + }, + { + "epoch": 0.7861583091587235, + "grad_norm": 0.2333671599626541, + "learning_rate": 1.0696228760878575e-06, + "loss": 0.4975, + "step": 9485 + }, + { + "epoch": 0.7862411935350186, + "grad_norm": 0.25958240032196045, + "learning_rate": 1.0692084542063822e-06, + "loss": 0.4969, + "step": 9486 + }, + { + "epoch": 0.7863240779113138, + "grad_norm": 0.25628942251205444, + "learning_rate": 1.0687940323249068e-06, + "loss": 0.5408, + "step": 9487 + }, + { + "epoch": 0.7864069622876088, + "grad_norm": 0.24001488089561462, + "learning_rate": 1.0683796104434314e-06, + "loss": 0.4553, + "step": 9488 + }, + { + "epoch": 0.7864898466639039, + "grad_norm": 0.2436957061290741, + "learning_rate": 1.067965188561956e-06, + "loss": 0.484, + "step": 9489 + }, + { + "epoch": 0.7865727310401989, + "grad_norm": 0.26522010564804077, + "learning_rate": 1.067550766680481e-06, + "loss": 0.5112, + "step": 9490 + }, + { + "epoch": 0.786655615416494, + "grad_norm": 0.24022500216960907, + "learning_rate": 1.0671363447990054e-06, + "loss": 0.4984, + "step": 9491 + }, + { + "epoch": 0.7867384997927891, + "grad_norm": 0.24719855189323425, + "learning_rate": 1.0667219229175302e-06, + "loss": 0.5209, + "step": 9492 + }, + { + "epoch": 0.7868213841690841, + "grad_norm": 0.2434086948633194, + "learning_rate": 1.0663075010360549e-06, + "loss": 0.4355, + "step": 9493 + }, + { + "epoch": 0.7869042685453792, + "grad_norm": 0.2516203820705414, + "learning_rate": 1.0658930791545795e-06, + "loss": 0.5145, + "step": 9494 + }, + { + "epoch": 0.7869871529216743, + "grad_norm": 0.25072547793388367, + "learning_rate": 1.0654786572731042e-06, + "loss": 0.5858, + "step": 9495 + }, + { + "epoch": 0.7870700372979693, + "grad_norm": 0.235484316945076, + "learning_rate": 1.0650642353916288e-06, + "loss": 0.4769, + "step": 9496 + }, + { + "epoch": 0.7871529216742644, + "grad_norm": 0.2427503764629364, + "learning_rate": 1.0646498135101534e-06, + "loss": 0.508, + "step": 9497 + }, + { + "epoch": 0.7872358060505594, + "grad_norm": 0.36940476298332214, + "learning_rate": 1.064235391628678e-06, + "loss": 0.5651, + "step": 9498 + }, + { + "epoch": 0.7873186904268545, + "grad_norm": 0.2510269284248352, + "learning_rate": 1.0638209697472027e-06, + "loss": 0.4294, + "step": 9499 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 0.2512453496456146, + "learning_rate": 1.0634065478657274e-06, + "loss": 0.5405, + "step": 9500 + }, + { + "epoch": 0.7874844591794447, + "grad_norm": 0.24477136135101318, + "learning_rate": 1.062992125984252e-06, + "loss": 0.5105, + "step": 9501 + }, + { + "epoch": 0.7875673435557398, + "grad_norm": 0.23319292068481445, + "learning_rate": 1.0625777041027767e-06, + "loss": 0.4998, + "step": 9502 + }, + { + "epoch": 0.7876502279320348, + "grad_norm": 0.2441142052412033, + "learning_rate": 1.0621632822213015e-06, + "loss": 0.4996, + "step": 9503 + }, + { + "epoch": 0.7877331123083299, + "grad_norm": 0.25348275899887085, + "learning_rate": 1.061748860339826e-06, + "loss": 0.5313, + "step": 9504 + }, + { + "epoch": 0.787815996684625, + "grad_norm": 0.2438945472240448, + "learning_rate": 1.0613344384583508e-06, + "loss": 0.4929, + "step": 9505 + }, + { + "epoch": 0.78789888106092, + "grad_norm": 0.23676222562789917, + "learning_rate": 1.0609200165768752e-06, + "loss": 0.4794, + "step": 9506 + }, + { + "epoch": 0.7879817654372151, + "grad_norm": 0.2302626222372055, + "learning_rate": 1.0605055946954e-06, + "loss": 0.4596, + "step": 9507 + }, + { + "epoch": 0.7880646498135101, + "grad_norm": 0.24513769149780273, + "learning_rate": 1.0600911728139245e-06, + "loss": 0.5094, + "step": 9508 + }, + { + "epoch": 0.7881475341898052, + "grad_norm": 0.23363324999809265, + "learning_rate": 1.0596767509324494e-06, + "loss": 0.4587, + "step": 9509 + }, + { + "epoch": 0.7882304185661003, + "grad_norm": 0.25246503949165344, + "learning_rate": 1.059262329050974e-06, + "loss": 0.4891, + "step": 9510 + }, + { + "epoch": 0.7883133029423953, + "grad_norm": 0.2513149678707123, + "learning_rate": 1.0588479071694987e-06, + "loss": 0.5026, + "step": 9511 + }, + { + "epoch": 0.7883961873186904, + "grad_norm": 0.24286963045597076, + "learning_rate": 1.0584334852880233e-06, + "loss": 0.4539, + "step": 9512 + }, + { + "epoch": 0.7884790716949855, + "grad_norm": 0.24575145542621613, + "learning_rate": 1.058019063406548e-06, + "loss": 0.4918, + "step": 9513 + }, + { + "epoch": 0.7885619560712805, + "grad_norm": 0.22838261723518372, + "learning_rate": 1.0576046415250726e-06, + "loss": 0.4734, + "step": 9514 + }, + { + "epoch": 0.7886448404475757, + "grad_norm": 0.22449183464050293, + "learning_rate": 1.0571902196435972e-06, + "loss": 0.4732, + "step": 9515 + }, + { + "epoch": 0.7887277248238707, + "grad_norm": 0.22792084515094757, + "learning_rate": 1.0567757977621219e-06, + "loss": 0.4811, + "step": 9516 + }, + { + "epoch": 0.7888106092001658, + "grad_norm": 0.2659744322299957, + "learning_rate": 1.0563613758806465e-06, + "loss": 0.5034, + "step": 9517 + }, + { + "epoch": 0.7888934935764609, + "grad_norm": 0.24253520369529724, + "learning_rate": 1.0559469539991712e-06, + "loss": 0.4749, + "step": 9518 + }, + { + "epoch": 0.7889763779527559, + "grad_norm": 0.2595040202140808, + "learning_rate": 1.055532532117696e-06, + "loss": 0.5511, + "step": 9519 + }, + { + "epoch": 0.789059262329051, + "grad_norm": 0.23747502267360687, + "learning_rate": 1.0551181102362205e-06, + "loss": 0.4771, + "step": 9520 + }, + { + "epoch": 0.789142146705346, + "grad_norm": 0.23082000017166138, + "learning_rate": 1.0547036883547453e-06, + "loss": 0.5057, + "step": 9521 + }, + { + "epoch": 0.7892250310816411, + "grad_norm": 0.23673014342784882, + "learning_rate": 1.0542892664732697e-06, + "loss": 0.4946, + "step": 9522 + }, + { + "epoch": 0.7893079154579362, + "grad_norm": 0.2325560748577118, + "learning_rate": 1.0538748445917946e-06, + "loss": 0.4803, + "step": 9523 + }, + { + "epoch": 0.7893907998342312, + "grad_norm": 0.24319271743297577, + "learning_rate": 1.0534604227103192e-06, + "loss": 0.5143, + "step": 9524 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 0.24055473506450653, + "learning_rate": 1.0530460008288439e-06, + "loss": 0.4764, + "step": 9525 + }, + { + "epoch": 0.7895565685868214, + "grad_norm": 0.2522287368774414, + "learning_rate": 1.0526315789473685e-06, + "loss": 0.4932, + "step": 9526 + }, + { + "epoch": 0.7896394529631164, + "grad_norm": 0.24395644664764404, + "learning_rate": 1.0522171570658932e-06, + "loss": 0.4935, + "step": 9527 + }, + { + "epoch": 0.7897223373394116, + "grad_norm": 0.243342325091362, + "learning_rate": 1.0518027351844178e-06, + "loss": 0.5191, + "step": 9528 + }, + { + "epoch": 0.7898052217157066, + "grad_norm": 0.23587879538536072, + "learning_rate": 1.0513883133029425e-06, + "loss": 0.4703, + "step": 9529 + }, + { + "epoch": 0.7898881060920017, + "grad_norm": 0.2414507269859314, + "learning_rate": 1.050973891421467e-06, + "loss": 0.4833, + "step": 9530 + }, + { + "epoch": 0.7899709904682968, + "grad_norm": 0.24065285921096802, + "learning_rate": 1.0505594695399917e-06, + "loss": 0.4706, + "step": 9531 + }, + { + "epoch": 0.7900538748445918, + "grad_norm": 0.24813444912433624, + "learning_rate": 1.0501450476585164e-06, + "loss": 0.5092, + "step": 9532 + }, + { + "epoch": 0.7901367592208869, + "grad_norm": 0.2441878467798233, + "learning_rate": 1.049730625777041e-06, + "loss": 0.5226, + "step": 9533 + }, + { + "epoch": 0.7902196435971819, + "grad_norm": 0.24762405455112457, + "learning_rate": 1.0493162038955659e-06, + "loss": 0.4657, + "step": 9534 + }, + { + "epoch": 0.790302527973477, + "grad_norm": 0.23476485908031464, + "learning_rate": 1.0489017820140903e-06, + "loss": 0.4879, + "step": 9535 + }, + { + "epoch": 0.7903854123497721, + "grad_norm": 0.23440231382846832, + "learning_rate": 1.0484873601326152e-06, + "loss": 0.4957, + "step": 9536 + }, + { + "epoch": 0.7904682967260671, + "grad_norm": 0.24788798391819, + "learning_rate": 1.0480729382511396e-06, + "loss": 0.4822, + "step": 9537 + }, + { + "epoch": 0.7905511811023622, + "grad_norm": 0.2373180389404297, + "learning_rate": 1.0476585163696645e-06, + "loss": 0.5007, + "step": 9538 + }, + { + "epoch": 0.7906340654786572, + "grad_norm": 0.23580434918403625, + "learning_rate": 1.0472440944881891e-06, + "loss": 0.5229, + "step": 9539 + }, + { + "epoch": 0.7907169498549523, + "grad_norm": 0.2422599047422409, + "learning_rate": 1.0468296726067138e-06, + "loss": 0.5165, + "step": 9540 + }, + { + "epoch": 0.7907998342312474, + "grad_norm": 0.2312830537557602, + "learning_rate": 1.0464152507252384e-06, + "loss": 0.4442, + "step": 9541 + }, + { + "epoch": 0.7908827186075424, + "grad_norm": 0.23898622393608093, + "learning_rate": 1.046000828843763e-06, + "loss": 0.485, + "step": 9542 + }, + { + "epoch": 0.7909656029838376, + "grad_norm": 0.24106670916080475, + "learning_rate": 1.0455864069622877e-06, + "loss": 0.5059, + "step": 9543 + }, + { + "epoch": 0.7910484873601327, + "grad_norm": 0.22422778606414795, + "learning_rate": 1.0451719850808123e-06, + "loss": 0.472, + "step": 9544 + }, + { + "epoch": 0.7911313717364277, + "grad_norm": 0.2712263762950897, + "learning_rate": 1.044757563199337e-06, + "loss": 0.5346, + "step": 9545 + }, + { + "epoch": 0.7912142561127228, + "grad_norm": 0.255456805229187, + "learning_rate": 1.0443431413178616e-06, + "loss": 0.4884, + "step": 9546 + }, + { + "epoch": 0.7912971404890178, + "grad_norm": 0.22760020196437836, + "learning_rate": 1.0439287194363863e-06, + "loss": 0.5173, + "step": 9547 + }, + { + "epoch": 0.7913800248653129, + "grad_norm": 0.23475681245326996, + "learning_rate": 1.0435142975549111e-06, + "loss": 0.5071, + "step": 9548 + }, + { + "epoch": 0.791462909241608, + "grad_norm": 0.26404011249542236, + "learning_rate": 1.0430998756734355e-06, + "loss": 0.5186, + "step": 9549 + }, + { + "epoch": 0.791545793617903, + "grad_norm": 0.23365427553653717, + "learning_rate": 1.0426854537919604e-06, + "loss": 0.4937, + "step": 9550 + }, + { + "epoch": 0.7916286779941981, + "grad_norm": 0.2392355501651764, + "learning_rate": 1.0422710319104848e-06, + "loss": 0.4954, + "step": 9551 + }, + { + "epoch": 0.7917115623704931, + "grad_norm": 0.24015729129314423, + "learning_rate": 1.0418566100290097e-06, + "loss": 0.5071, + "step": 9552 + }, + { + "epoch": 0.7917944467467882, + "grad_norm": 0.22467583417892456, + "learning_rate": 1.0414421881475343e-06, + "loss": 0.4339, + "step": 9553 + }, + { + "epoch": 0.7918773311230833, + "grad_norm": 0.26446473598480225, + "learning_rate": 1.041027766266059e-06, + "loss": 0.5378, + "step": 9554 + }, + { + "epoch": 0.7919602154993783, + "grad_norm": 0.22540156543254852, + "learning_rate": 1.0406133443845836e-06, + "loss": 0.4644, + "step": 9555 + }, + { + "epoch": 0.7920430998756735, + "grad_norm": 0.228639155626297, + "learning_rate": 1.0401989225031083e-06, + "loss": 0.4945, + "step": 9556 + }, + { + "epoch": 0.7921259842519685, + "grad_norm": 0.2670897841453552, + "learning_rate": 1.039784500621633e-06, + "loss": 0.5276, + "step": 9557 + }, + { + "epoch": 0.7922088686282636, + "grad_norm": 0.2420656979084015, + "learning_rate": 1.0393700787401575e-06, + "loss": 0.5081, + "step": 9558 + }, + { + "epoch": 0.7922917530045587, + "grad_norm": 0.23179297149181366, + "learning_rate": 1.0389556568586822e-06, + "loss": 0.4599, + "step": 9559 + }, + { + "epoch": 0.7923746373808537, + "grad_norm": 0.26570314168930054, + "learning_rate": 1.0385412349772068e-06, + "loss": 0.5001, + "step": 9560 + }, + { + "epoch": 0.7924575217571488, + "grad_norm": 0.24129991233348846, + "learning_rate": 1.0381268130957315e-06, + "loss": 0.5291, + "step": 9561 + }, + { + "epoch": 0.7925404061334439, + "grad_norm": 0.2490924447774887, + "learning_rate": 1.0377123912142561e-06, + "loss": 0.4623, + "step": 9562 + }, + { + "epoch": 0.7926232905097389, + "grad_norm": 0.24586401879787445, + "learning_rate": 1.037297969332781e-06, + "loss": 0.5384, + "step": 9563 + }, + { + "epoch": 0.792706174886034, + "grad_norm": 0.2480650544166565, + "learning_rate": 1.0368835474513054e-06, + "loss": 0.4862, + "step": 9564 + }, + { + "epoch": 0.792789059262329, + "grad_norm": 0.23736916482448578, + "learning_rate": 1.0364691255698303e-06, + "loss": 0.4714, + "step": 9565 + }, + { + "epoch": 0.7928719436386241, + "grad_norm": 0.24992893636226654, + "learning_rate": 1.036054703688355e-06, + "loss": 0.4991, + "step": 9566 + }, + { + "epoch": 0.7929548280149192, + "grad_norm": 0.23623839020729065, + "learning_rate": 1.0356402818068796e-06, + "loss": 0.4883, + "step": 9567 + }, + { + "epoch": 0.7930377123912142, + "grad_norm": 0.23473979532718658, + "learning_rate": 1.0352258599254042e-06, + "loss": 0.4547, + "step": 9568 + }, + { + "epoch": 0.7931205967675093, + "grad_norm": 0.21919402480125427, + "learning_rate": 1.0348114380439288e-06, + "loss": 0.476, + "step": 9569 + }, + { + "epoch": 0.7932034811438043, + "grad_norm": 0.24400803446769714, + "learning_rate": 1.0343970161624535e-06, + "loss": 0.4439, + "step": 9570 + }, + { + "epoch": 0.7932863655200995, + "grad_norm": 0.2368060052394867, + "learning_rate": 1.0339825942809781e-06, + "loss": 0.4908, + "step": 9571 + }, + { + "epoch": 0.7933692498963946, + "grad_norm": 0.2172282487154007, + "learning_rate": 1.0335681723995028e-06, + "loss": 0.4838, + "step": 9572 + }, + { + "epoch": 0.7934521342726896, + "grad_norm": 0.22814907133579254, + "learning_rate": 1.0331537505180274e-06, + "loss": 0.4986, + "step": 9573 + }, + { + "epoch": 0.7935350186489847, + "grad_norm": 0.2689914107322693, + "learning_rate": 1.032739328636552e-06, + "loss": 0.5186, + "step": 9574 + }, + { + "epoch": 0.7936179030252797, + "grad_norm": 0.23218876123428345, + "learning_rate": 1.0323249067550767e-06, + "loss": 0.4663, + "step": 9575 + }, + { + "epoch": 0.7937007874015748, + "grad_norm": 0.25390350818634033, + "learning_rate": 1.0319104848736013e-06, + "loss": 0.5045, + "step": 9576 + }, + { + "epoch": 0.7937836717778699, + "grad_norm": 0.2531627118587494, + "learning_rate": 1.0314960629921262e-06, + "loss": 0.5242, + "step": 9577 + }, + { + "epoch": 0.7938665561541649, + "grad_norm": 0.227661594748497, + "learning_rate": 1.0310816411106506e-06, + "loss": 0.5359, + "step": 9578 + }, + { + "epoch": 0.79394944053046, + "grad_norm": 0.23596392571926117, + "learning_rate": 1.0306672192291755e-06, + "loss": 0.473, + "step": 9579 + }, + { + "epoch": 0.7940323249067551, + "grad_norm": 0.2430090457201004, + "learning_rate": 1.0302527973477e-06, + "loss": 0.4891, + "step": 9580 + }, + { + "epoch": 0.7941152092830501, + "grad_norm": 0.22393928468227386, + "learning_rate": 1.0298383754662248e-06, + "loss": 0.4875, + "step": 9581 + }, + { + "epoch": 0.7941980936593452, + "grad_norm": 0.24988672137260437, + "learning_rate": 1.0294239535847492e-06, + "loss": 0.5036, + "step": 9582 + }, + { + "epoch": 0.7942809780356402, + "grad_norm": 0.2294778674840927, + "learning_rate": 1.029009531703274e-06, + "loss": 0.4613, + "step": 9583 + }, + { + "epoch": 0.7943638624119354, + "grad_norm": 0.2572665810585022, + "learning_rate": 1.0285951098217987e-06, + "loss": 0.5411, + "step": 9584 + }, + { + "epoch": 0.7944467467882305, + "grad_norm": 0.2589656710624695, + "learning_rate": 1.0281806879403233e-06, + "loss": 0.5338, + "step": 9585 + }, + { + "epoch": 0.7945296311645255, + "grad_norm": 0.2570745050907135, + "learning_rate": 1.027766266058848e-06, + "loss": 0.5339, + "step": 9586 + }, + { + "epoch": 0.7946125155408206, + "grad_norm": 0.23294652998447418, + "learning_rate": 1.0273518441773726e-06, + "loss": 0.4496, + "step": 9587 + }, + { + "epoch": 0.7946953999171156, + "grad_norm": 0.26186782121658325, + "learning_rate": 1.0269374222958973e-06, + "loss": 0.4986, + "step": 9588 + }, + { + "epoch": 0.7947782842934107, + "grad_norm": 0.23765544593334198, + "learning_rate": 1.026523000414422e-06, + "loss": 0.4944, + "step": 9589 + }, + { + "epoch": 0.7948611686697058, + "grad_norm": 0.2582770586013794, + "learning_rate": 1.0261085785329466e-06, + "loss": 0.5337, + "step": 9590 + }, + { + "epoch": 0.7949440530460008, + "grad_norm": 0.23419904708862305, + "learning_rate": 1.0256941566514712e-06, + "loss": 0.4568, + "step": 9591 + }, + { + "epoch": 0.7950269374222959, + "grad_norm": 0.26143208146095276, + "learning_rate": 1.0252797347699959e-06, + "loss": 0.5137, + "step": 9592 + }, + { + "epoch": 0.795109821798591, + "grad_norm": 0.2563678026199341, + "learning_rate": 1.0248653128885205e-06, + "loss": 0.4924, + "step": 9593 + }, + { + "epoch": 0.795192706174886, + "grad_norm": 0.23539942502975464, + "learning_rate": 1.0244508910070453e-06, + "loss": 0.4842, + "step": 9594 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 0.2564922273159027, + "learning_rate": 1.02403646912557e-06, + "loss": 0.4984, + "step": 9595 + }, + { + "epoch": 0.7953584749274761, + "grad_norm": 0.2387603372335434, + "learning_rate": 1.0236220472440946e-06, + "loss": 0.4856, + "step": 9596 + }, + { + "epoch": 0.7954413593037712, + "grad_norm": 0.26423853635787964, + "learning_rate": 1.0232076253626193e-06, + "loss": 0.4743, + "step": 9597 + }, + { + "epoch": 0.7955242436800664, + "grad_norm": 0.22231408953666687, + "learning_rate": 1.022793203481144e-06, + "loss": 0.4677, + "step": 9598 + }, + { + "epoch": 0.7956071280563614, + "grad_norm": 0.23172757029533386, + "learning_rate": 1.0223787815996686e-06, + "loss": 0.5129, + "step": 9599 + }, + { + "epoch": 0.7956900124326565, + "grad_norm": 0.2514471113681793, + "learning_rate": 1.0219643597181932e-06, + "loss": 0.4883, + "step": 9600 + }, + { + "epoch": 0.7957728968089515, + "grad_norm": 0.23631136119365692, + "learning_rate": 1.0215499378367179e-06, + "loss": 0.4993, + "step": 9601 + }, + { + "epoch": 0.7958557811852466, + "grad_norm": 0.23042762279510498, + "learning_rate": 1.0211355159552425e-06, + "loss": 0.4772, + "step": 9602 + }, + { + "epoch": 0.7959386655615417, + "grad_norm": 0.24369828402996063, + "learning_rate": 1.0207210940737671e-06, + "loss": 0.5358, + "step": 9603 + }, + { + "epoch": 0.7960215499378367, + "grad_norm": 0.21955378353595734, + "learning_rate": 1.020306672192292e-06, + "loss": 0.4535, + "step": 9604 + }, + { + "epoch": 0.7961044343141318, + "grad_norm": 0.24161259829998016, + "learning_rate": 1.0198922503108164e-06, + "loss": 0.4815, + "step": 9605 + }, + { + "epoch": 0.7961873186904268, + "grad_norm": 0.23430633544921875, + "learning_rate": 1.0194778284293413e-06, + "loss": 0.4638, + "step": 9606 + }, + { + "epoch": 0.7962702030667219, + "grad_norm": 0.24657845497131348, + "learning_rate": 1.0190634065478657e-06, + "loss": 0.5403, + "step": 9607 + }, + { + "epoch": 0.796353087443017, + "grad_norm": 0.2548978626728058, + "learning_rate": 1.0186489846663906e-06, + "loss": 0.5276, + "step": 9608 + }, + { + "epoch": 0.796435971819312, + "grad_norm": 0.24376943707466125, + "learning_rate": 1.018234562784915e-06, + "loss": 0.4608, + "step": 9609 + }, + { + "epoch": 0.7965188561956071, + "grad_norm": 0.23678022623062134, + "learning_rate": 1.0178201409034399e-06, + "loss": 0.5037, + "step": 9610 + }, + { + "epoch": 0.7966017405719023, + "grad_norm": 0.2544761300086975, + "learning_rate": 1.0174057190219643e-06, + "loss": 0.5377, + "step": 9611 + }, + { + "epoch": 0.7966846249481973, + "grad_norm": 0.2448352724313736, + "learning_rate": 1.0169912971404891e-06, + "loss": 0.5024, + "step": 9612 + }, + { + "epoch": 0.7967675093244924, + "grad_norm": 0.2468814253807068, + "learning_rate": 1.0165768752590138e-06, + "loss": 0.4941, + "step": 9613 + }, + { + "epoch": 0.7968503937007874, + "grad_norm": 0.23931603133678436, + "learning_rate": 1.0161624533775384e-06, + "loss": 0.5126, + "step": 9614 + }, + { + "epoch": 0.7969332780770825, + "grad_norm": 0.233881413936615, + "learning_rate": 1.015748031496063e-06, + "loss": 0.4784, + "step": 9615 + }, + { + "epoch": 0.7970161624533776, + "grad_norm": 0.22830887138843536, + "learning_rate": 1.0153336096145877e-06, + "loss": 0.4651, + "step": 9616 + }, + { + "epoch": 0.7970990468296726, + "grad_norm": 0.23105661571025848, + "learning_rate": 1.0149191877331124e-06, + "loss": 0.4559, + "step": 9617 + }, + { + "epoch": 0.7971819312059677, + "grad_norm": 0.25285759568214417, + "learning_rate": 1.014504765851637e-06, + "loss": 0.4975, + "step": 9618 + }, + { + "epoch": 0.7972648155822627, + "grad_norm": 0.2417861521244049, + "learning_rate": 1.0140903439701617e-06, + "loss": 0.5042, + "step": 9619 + }, + { + "epoch": 0.7973476999585578, + "grad_norm": 0.24411582946777344, + "learning_rate": 1.0136759220886863e-06, + "loss": 0.5081, + "step": 9620 + }, + { + "epoch": 0.7974305843348529, + "grad_norm": 0.24869146943092346, + "learning_rate": 1.013261500207211e-06, + "loss": 0.4934, + "step": 9621 + }, + { + "epoch": 0.7975134687111479, + "grad_norm": 0.2760137915611267, + "learning_rate": 1.0128470783257356e-06, + "loss": 0.5055, + "step": 9622 + }, + { + "epoch": 0.797596353087443, + "grad_norm": 0.24369937181472778, + "learning_rate": 1.0124326564442604e-06, + "loss": 0.4878, + "step": 9623 + }, + { + "epoch": 0.797679237463738, + "grad_norm": 0.23696282505989075, + "learning_rate": 1.012018234562785e-06, + "loss": 0.485, + "step": 9624 + }, + { + "epoch": 0.7977621218400331, + "grad_norm": 0.23912613093852997, + "learning_rate": 1.0116038126813097e-06, + "loss": 0.5161, + "step": 9625 + }, + { + "epoch": 0.7978450062163283, + "grad_norm": 0.22313541173934937, + "learning_rate": 1.0111893907998344e-06, + "loss": 0.4592, + "step": 9626 + }, + { + "epoch": 0.7979278905926233, + "grad_norm": 0.24444858729839325, + "learning_rate": 1.010774968918359e-06, + "loss": 0.4883, + "step": 9627 + }, + { + "epoch": 0.7980107749689184, + "grad_norm": 0.22541849315166473, + "learning_rate": 1.0103605470368837e-06, + "loss": 0.451, + "step": 9628 + }, + { + "epoch": 0.7980936593452135, + "grad_norm": 0.25942161679267883, + "learning_rate": 1.0099461251554083e-06, + "loss": 0.5345, + "step": 9629 + }, + { + "epoch": 0.7981765437215085, + "grad_norm": 0.2249785214662552, + "learning_rate": 1.009531703273933e-06, + "loss": 0.4894, + "step": 9630 + }, + { + "epoch": 0.7982594280978036, + "grad_norm": 0.26165035367012024, + "learning_rate": 1.0091172813924576e-06, + "loss": 0.4866, + "step": 9631 + }, + { + "epoch": 0.7983423124740986, + "grad_norm": 0.2569468021392822, + "learning_rate": 1.0087028595109822e-06, + "loss": 0.5131, + "step": 9632 + }, + { + "epoch": 0.7984251968503937, + "grad_norm": 0.2982080280780792, + "learning_rate": 1.008288437629507e-06, + "loss": 0.4772, + "step": 9633 + }, + { + "epoch": 0.7985080812266888, + "grad_norm": 0.24750882387161255, + "learning_rate": 1.0078740157480315e-06, + "loss": 0.5125, + "step": 9634 + }, + { + "epoch": 0.7985909656029838, + "grad_norm": 0.2580457627773285, + "learning_rate": 1.0074595938665564e-06, + "loss": 0.5045, + "step": 9635 + }, + { + "epoch": 0.7986738499792789, + "grad_norm": 0.23170968890190125, + "learning_rate": 1.0070451719850808e-06, + "loss": 0.4539, + "step": 9636 + }, + { + "epoch": 0.7987567343555739, + "grad_norm": 0.23763787746429443, + "learning_rate": 1.0066307501036057e-06, + "loss": 0.4772, + "step": 9637 + }, + { + "epoch": 0.798839618731869, + "grad_norm": 0.2447720468044281, + "learning_rate": 1.00621632822213e-06, + "loss": 0.5376, + "step": 9638 + }, + { + "epoch": 0.7989225031081642, + "grad_norm": 0.25821369886398315, + "learning_rate": 1.005801906340655e-06, + "loss": 0.4669, + "step": 9639 + }, + { + "epoch": 0.7990053874844592, + "grad_norm": 0.23921409249305725, + "learning_rate": 1.0053874844591794e-06, + "loss": 0.4949, + "step": 9640 + }, + { + "epoch": 0.7990882718607543, + "grad_norm": 0.24850398302078247, + "learning_rate": 1.0049730625777042e-06, + "loss": 0.4919, + "step": 9641 + }, + { + "epoch": 0.7991711562370493, + "grad_norm": 0.24836570024490356, + "learning_rate": 1.0045586406962289e-06, + "loss": 0.4983, + "step": 9642 + }, + { + "epoch": 0.7992540406133444, + "grad_norm": 0.21719248592853546, + "learning_rate": 1.0041442188147535e-06, + "loss": 0.45, + "step": 9643 + }, + { + "epoch": 0.7993369249896395, + "grad_norm": 0.230239599943161, + "learning_rate": 1.0037297969332782e-06, + "loss": 0.4545, + "step": 9644 + }, + { + "epoch": 0.7994198093659345, + "grad_norm": 0.23549872636795044, + "learning_rate": 1.0033153750518028e-06, + "loss": 0.4724, + "step": 9645 + }, + { + "epoch": 0.7995026937422296, + "grad_norm": 0.25402599573135376, + "learning_rate": 1.0029009531703274e-06, + "loss": 0.5085, + "step": 9646 + }, + { + "epoch": 0.7995855781185247, + "grad_norm": 0.23621249198913574, + "learning_rate": 1.002486531288852e-06, + "loss": 0.4745, + "step": 9647 + }, + { + "epoch": 0.7996684624948197, + "grad_norm": 0.248994842171669, + "learning_rate": 1.0020721094073767e-06, + "loss": 0.5217, + "step": 9648 + }, + { + "epoch": 0.7997513468711148, + "grad_norm": 0.2456979900598526, + "learning_rate": 1.0016576875259014e-06, + "loss": 0.4973, + "step": 9649 + }, + { + "epoch": 0.7998342312474098, + "grad_norm": 0.2331475168466568, + "learning_rate": 1.001243265644426e-06, + "loss": 0.4857, + "step": 9650 + }, + { + "epoch": 0.7999171156237049, + "grad_norm": 0.2447153478860855, + "learning_rate": 1.0008288437629507e-06, + "loss": 0.4773, + "step": 9651 + }, + { + "epoch": 0.8, + "grad_norm": 0.22694331407546997, + "learning_rate": 1.0004144218814755e-06, + "loss": 0.5126, + "step": 9652 + }, + { + "epoch": 0.800082884376295, + "grad_norm": 0.2611900269985199, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.5642, + "step": 9653 + }, + { + "epoch": 0.8001657687525902, + "grad_norm": 0.25103452801704407, + "learning_rate": 9.995855781185248e-07, + "loss": 0.4877, + "step": 9654 + }, + { + "epoch": 0.8002486531288852, + "grad_norm": 0.24912412464618683, + "learning_rate": 9.991711562370495e-07, + "loss": 0.4928, + "step": 9655 + }, + { + "epoch": 0.8003315375051803, + "grad_norm": 0.2426309436559677, + "learning_rate": 9.98756734355574e-07, + "loss": 0.488, + "step": 9656 + }, + { + "epoch": 0.8004144218814754, + "grad_norm": 0.23707252740859985, + "learning_rate": 9.983423124740987e-07, + "loss": 0.4981, + "step": 9657 + }, + { + "epoch": 0.8004973062577704, + "grad_norm": 0.23826564848423004, + "learning_rate": 9.979278905926234e-07, + "loss": 0.4772, + "step": 9658 + }, + { + "epoch": 0.8005801906340655, + "grad_norm": 0.23578156530857086, + "learning_rate": 9.97513468711148e-07, + "loss": 0.5077, + "step": 9659 + }, + { + "epoch": 0.8006630750103605, + "grad_norm": 0.2508103847503662, + "learning_rate": 9.970990468296727e-07, + "loss": 0.5266, + "step": 9660 + }, + { + "epoch": 0.8007459593866556, + "grad_norm": 0.2260228544473648, + "learning_rate": 9.966846249481973e-07, + "loss": 0.4755, + "step": 9661 + }, + { + "epoch": 0.8008288437629507, + "grad_norm": 0.24467425048351288, + "learning_rate": 9.96270203066722e-07, + "loss": 0.5261, + "step": 9662 + }, + { + "epoch": 0.8009117281392457, + "grad_norm": 0.23978449404239655, + "learning_rate": 9.958557811852466e-07, + "loss": 0.4979, + "step": 9663 + }, + { + "epoch": 0.8009946125155408, + "grad_norm": 0.22614862024784088, + "learning_rate": 9.954413593037715e-07, + "loss": 0.4502, + "step": 9664 + }, + { + "epoch": 0.8010774968918359, + "grad_norm": 0.2489856779575348, + "learning_rate": 9.950269374222959e-07, + "loss": 0.4875, + "step": 9665 + }, + { + "epoch": 0.8011603812681309, + "grad_norm": 0.24953968822956085, + "learning_rate": 9.946125155408207e-07, + "loss": 0.53, + "step": 9666 + }, + { + "epoch": 0.801243265644426, + "grad_norm": 0.24351705610752106, + "learning_rate": 9.941980936593452e-07, + "loss": 0.5151, + "step": 9667 + }, + { + "epoch": 0.801326150020721, + "grad_norm": 0.25376594066619873, + "learning_rate": 9.9378367177787e-07, + "loss": 0.5287, + "step": 9668 + }, + { + "epoch": 0.8014090343970162, + "grad_norm": 0.23249934613704681, + "learning_rate": 9.933692498963945e-07, + "loss": 0.4527, + "step": 9669 + }, + { + "epoch": 0.8014919187733113, + "grad_norm": 0.23421034216880798, + "learning_rate": 9.929548280149193e-07, + "loss": 0.4998, + "step": 9670 + }, + { + "epoch": 0.8015748031496063, + "grad_norm": 0.2469000518321991, + "learning_rate": 9.92540406133444e-07, + "loss": 0.5165, + "step": 9671 + }, + { + "epoch": 0.8016576875259014, + "grad_norm": 0.24287374317646027, + "learning_rate": 9.921259842519686e-07, + "loss": 0.5052, + "step": 9672 + }, + { + "epoch": 0.8017405719021964, + "grad_norm": 0.2456335723400116, + "learning_rate": 9.917115623704932e-07, + "loss": 0.5241, + "step": 9673 + }, + { + "epoch": 0.8018234562784915, + "grad_norm": 0.24865774810314178, + "learning_rate": 9.912971404890179e-07, + "loss": 0.4564, + "step": 9674 + }, + { + "epoch": 0.8019063406547866, + "grad_norm": 0.24011407792568207, + "learning_rate": 9.908827186075425e-07, + "loss": 0.485, + "step": 9675 + }, + { + "epoch": 0.8019892250310816, + "grad_norm": 0.23177197575569153, + "learning_rate": 9.904682967260672e-07, + "loss": 0.4245, + "step": 9676 + }, + { + "epoch": 0.8020721094073767, + "grad_norm": 0.2784504294395447, + "learning_rate": 9.900538748445918e-07, + "loss": 0.5295, + "step": 9677 + }, + { + "epoch": 0.8021549937836718, + "grad_norm": 0.2617378234863281, + "learning_rate": 9.896394529631165e-07, + "loss": 0.525, + "step": 9678 + }, + { + "epoch": 0.8022378781599668, + "grad_norm": 0.24348510801792145, + "learning_rate": 9.892250310816411e-07, + "loss": 0.4936, + "step": 9679 + }, + { + "epoch": 0.8023207625362619, + "grad_norm": 0.23731841146945953, + "learning_rate": 9.88810609200166e-07, + "loss": 0.4634, + "step": 9680 + }, + { + "epoch": 0.802403646912557, + "grad_norm": 0.2637873589992523, + "learning_rate": 9.883961873186904e-07, + "loss": 0.5005, + "step": 9681 + }, + { + "epoch": 0.8024865312888521, + "grad_norm": 0.23773974180221558, + "learning_rate": 9.879817654372153e-07, + "loss": 0.4638, + "step": 9682 + }, + { + "epoch": 0.8025694156651472, + "grad_norm": 0.23301319777965546, + "learning_rate": 9.8756734355574e-07, + "loss": 0.4548, + "step": 9683 + }, + { + "epoch": 0.8026523000414422, + "grad_norm": 0.23081330955028534, + "learning_rate": 9.871529216742645e-07, + "loss": 0.4641, + "step": 9684 + }, + { + "epoch": 0.8027351844177373, + "grad_norm": 0.25851836800575256, + "learning_rate": 9.867384997927892e-07, + "loss": 0.5131, + "step": 9685 + }, + { + "epoch": 0.8028180687940323, + "grad_norm": 0.240626260638237, + "learning_rate": 9.863240779113138e-07, + "loss": 0.4759, + "step": 9686 + }, + { + "epoch": 0.8029009531703274, + "grad_norm": 0.24423301219940186, + "learning_rate": 9.859096560298385e-07, + "loss": 0.506, + "step": 9687 + }, + { + "epoch": 0.8029838375466225, + "grad_norm": 0.23729662597179413, + "learning_rate": 9.854952341483631e-07, + "loss": 0.4943, + "step": 9688 + }, + { + "epoch": 0.8030667219229175, + "grad_norm": 0.268092542886734, + "learning_rate": 9.850808122668878e-07, + "loss": 0.4844, + "step": 9689 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 0.23433589935302734, + "learning_rate": 9.846663903854124e-07, + "loss": 0.4787, + "step": 9690 + }, + { + "epoch": 0.8032324906755076, + "grad_norm": 0.24303235113620758, + "learning_rate": 9.84251968503937e-07, + "loss": 0.4371, + "step": 9691 + }, + { + "epoch": 0.8033153750518027, + "grad_norm": 0.2517416179180145, + "learning_rate": 9.838375466224617e-07, + "loss": 0.4872, + "step": 9692 + }, + { + "epoch": 0.8033982594280978, + "grad_norm": 0.24960021674633026, + "learning_rate": 9.834231247409865e-07, + "loss": 0.5266, + "step": 9693 + }, + { + "epoch": 0.8034811438043928, + "grad_norm": 0.26666080951690674, + "learning_rate": 9.83008702859511e-07, + "loss": 0.4868, + "step": 9694 + }, + { + "epoch": 0.803564028180688, + "grad_norm": 0.2542940676212311, + "learning_rate": 9.825942809780358e-07, + "loss": 0.4919, + "step": 9695 + }, + { + "epoch": 0.8036469125569831, + "grad_norm": 0.24119047820568085, + "learning_rate": 9.821798590965603e-07, + "loss": 0.5168, + "step": 9696 + }, + { + "epoch": 0.8037297969332781, + "grad_norm": 0.22926852107048035, + "learning_rate": 9.817654372150851e-07, + "loss": 0.5398, + "step": 9697 + }, + { + "epoch": 0.8038126813095732, + "grad_norm": 0.2561147212982178, + "learning_rate": 9.813510153336095e-07, + "loss": 0.5187, + "step": 9698 + }, + { + "epoch": 0.8038955656858682, + "grad_norm": 0.27847030758857727, + "learning_rate": 9.809365934521344e-07, + "loss": 0.522, + "step": 9699 + }, + { + "epoch": 0.8039784500621633, + "grad_norm": 0.2425428181886673, + "learning_rate": 9.80522171570659e-07, + "loss": 0.5121, + "step": 9700 + }, + { + "epoch": 0.8040613344384584, + "grad_norm": 0.2533017098903656, + "learning_rate": 9.801077496891837e-07, + "loss": 0.5381, + "step": 9701 + }, + { + "epoch": 0.8041442188147534, + "grad_norm": 0.24633289873600006, + "learning_rate": 9.796933278077083e-07, + "loss": 0.4623, + "step": 9702 + }, + { + "epoch": 0.8042271031910485, + "grad_norm": 0.26208949089050293, + "learning_rate": 9.79278905926233e-07, + "loss": 0.5677, + "step": 9703 + }, + { + "epoch": 0.8043099875673435, + "grad_norm": 0.23387041687965393, + "learning_rate": 9.788644840447576e-07, + "loss": 0.5115, + "step": 9704 + }, + { + "epoch": 0.8043928719436386, + "grad_norm": 0.23268941044807434, + "learning_rate": 9.784500621632823e-07, + "loss": 0.4878, + "step": 9705 + }, + { + "epoch": 0.8044757563199337, + "grad_norm": 0.2314557433128357, + "learning_rate": 9.78035640281807e-07, + "loss": 0.4774, + "step": 9706 + }, + { + "epoch": 0.8045586406962287, + "grad_norm": 0.26107028126716614, + "learning_rate": 9.776212184003316e-07, + "loss": 0.5052, + "step": 9707 + }, + { + "epoch": 0.8046415250725238, + "grad_norm": 0.23641140758991241, + "learning_rate": 9.772067965188562e-07, + "loss": 0.5452, + "step": 9708 + }, + { + "epoch": 0.8047244094488188, + "grad_norm": 0.2363271564245224, + "learning_rate": 9.76792374637381e-07, + "loss": 0.4508, + "step": 9709 + }, + { + "epoch": 0.804807293825114, + "grad_norm": 0.2299196571111679, + "learning_rate": 9.763779527559055e-07, + "loss": 0.4791, + "step": 9710 + }, + { + "epoch": 0.8048901782014091, + "grad_norm": 0.23493559658527374, + "learning_rate": 9.759635308744303e-07, + "loss": 0.4599, + "step": 9711 + }, + { + "epoch": 0.8049730625777041, + "grad_norm": 0.23255354166030884, + "learning_rate": 9.75549108992955e-07, + "loss": 0.4647, + "step": 9712 + }, + { + "epoch": 0.8050559469539992, + "grad_norm": 0.2453741878271103, + "learning_rate": 9.751346871114796e-07, + "loss": 0.4666, + "step": 9713 + }, + { + "epoch": 0.8051388313302943, + "grad_norm": 0.26860862970352173, + "learning_rate": 9.747202652300043e-07, + "loss": 0.5241, + "step": 9714 + }, + { + "epoch": 0.8052217157065893, + "grad_norm": 0.24420854449272156, + "learning_rate": 9.74305843348529e-07, + "loss": 0.5201, + "step": 9715 + }, + { + "epoch": 0.8053046000828844, + "grad_norm": 0.2499796599149704, + "learning_rate": 9.738914214670536e-07, + "loss": 0.4964, + "step": 9716 + }, + { + "epoch": 0.8053874844591794, + "grad_norm": 0.24691668152809143, + "learning_rate": 9.734769995855782e-07, + "loss": 0.488, + "step": 9717 + }, + { + "epoch": 0.8054703688354745, + "grad_norm": 0.23665708303451538, + "learning_rate": 9.730625777041028e-07, + "loss": 0.5023, + "step": 9718 + }, + { + "epoch": 0.8055532532117696, + "grad_norm": 0.23981694877147675, + "learning_rate": 9.726481558226275e-07, + "loss": 0.4856, + "step": 9719 + }, + { + "epoch": 0.8056361375880646, + "grad_norm": 0.2763442099094391, + "learning_rate": 9.722337339411521e-07, + "loss": 0.5328, + "step": 9720 + }, + { + "epoch": 0.8057190219643597, + "grad_norm": 0.25191646814346313, + "learning_rate": 9.718193120596768e-07, + "loss": 0.4958, + "step": 9721 + }, + { + "epoch": 0.8058019063406547, + "grad_norm": 0.22742076218128204, + "learning_rate": 9.714048901782016e-07, + "loss": 0.5089, + "step": 9722 + }, + { + "epoch": 0.8058847907169499, + "grad_norm": 0.23919452726840973, + "learning_rate": 9.70990468296726e-07, + "loss": 0.4551, + "step": 9723 + }, + { + "epoch": 0.805967675093245, + "grad_norm": 0.23108597099781036, + "learning_rate": 9.70576046415251e-07, + "loss": 0.4695, + "step": 9724 + }, + { + "epoch": 0.80605055946954, + "grad_norm": 0.2469334900379181, + "learning_rate": 9.701616245337753e-07, + "loss": 0.5328, + "step": 9725 + }, + { + "epoch": 0.8061334438458351, + "grad_norm": 0.2456199824810028, + "learning_rate": 9.697472026523002e-07, + "loss": 0.4956, + "step": 9726 + }, + { + "epoch": 0.8062163282221301, + "grad_norm": 0.23640549182891846, + "learning_rate": 9.693327807708246e-07, + "loss": 0.4986, + "step": 9727 + }, + { + "epoch": 0.8062992125984252, + "grad_norm": 0.24166174232959747, + "learning_rate": 9.689183588893495e-07, + "loss": 0.5042, + "step": 9728 + }, + { + "epoch": 0.8063820969747203, + "grad_norm": 0.23861593008041382, + "learning_rate": 9.685039370078741e-07, + "loss": 0.4906, + "step": 9729 + }, + { + "epoch": 0.8064649813510153, + "grad_norm": 0.24875923991203308, + "learning_rate": 9.680895151263988e-07, + "loss": 0.5273, + "step": 9730 + }, + { + "epoch": 0.8065478657273104, + "grad_norm": 0.2497028410434723, + "learning_rate": 9.676750932449234e-07, + "loss": 0.5078, + "step": 9731 + }, + { + "epoch": 0.8066307501036055, + "grad_norm": 0.2593577206134796, + "learning_rate": 9.67260671363448e-07, + "loss": 0.5427, + "step": 9732 + }, + { + "epoch": 0.8067136344799005, + "grad_norm": 0.2518407702445984, + "learning_rate": 9.668462494819727e-07, + "loss": 0.5202, + "step": 9733 + }, + { + "epoch": 0.8067965188561956, + "grad_norm": 0.2431517243385315, + "learning_rate": 9.664318276004974e-07, + "loss": 0.483, + "step": 9734 + }, + { + "epoch": 0.8068794032324906, + "grad_norm": 0.2353769689798355, + "learning_rate": 9.66017405719022e-07, + "loss": 0.4779, + "step": 9735 + }, + { + "epoch": 0.8069622876087857, + "grad_norm": 0.23758532106876373, + "learning_rate": 9.656029838375466e-07, + "loss": 0.4421, + "step": 9736 + }, + { + "epoch": 0.8070451719850809, + "grad_norm": 0.23285192251205444, + "learning_rate": 9.651885619560713e-07, + "loss": 0.4824, + "step": 9737 + }, + { + "epoch": 0.8071280563613759, + "grad_norm": 0.22896552085876465, + "learning_rate": 9.647741400745961e-07, + "loss": 0.485, + "step": 9738 + }, + { + "epoch": 0.807210940737671, + "grad_norm": 0.2378280758857727, + "learning_rate": 9.643597181931206e-07, + "loss": 0.5176, + "step": 9739 + }, + { + "epoch": 0.807293825113966, + "grad_norm": 0.22920571267604828, + "learning_rate": 9.639452963116454e-07, + "loss": 0.4643, + "step": 9740 + }, + { + "epoch": 0.8073767094902611, + "grad_norm": 0.27138015627861023, + "learning_rate": 9.635308744301699e-07, + "loss": 0.4625, + "step": 9741 + }, + { + "epoch": 0.8074595938665562, + "grad_norm": 0.2520560622215271, + "learning_rate": 9.631164525486947e-07, + "loss": 0.5184, + "step": 9742 + }, + { + "epoch": 0.8075424782428512, + "grad_norm": 0.25634661316871643, + "learning_rate": 9.627020306672194e-07, + "loss": 0.4792, + "step": 9743 + }, + { + "epoch": 0.8076253626191463, + "grad_norm": 0.2401900440454483, + "learning_rate": 9.62287608785744e-07, + "loss": 0.4781, + "step": 9744 + }, + { + "epoch": 0.8077082469954414, + "grad_norm": 0.24647016823291779, + "learning_rate": 9.618731869042686e-07, + "loss": 0.5123, + "step": 9745 + }, + { + "epoch": 0.8077911313717364, + "grad_norm": 0.25749826431274414, + "learning_rate": 9.614587650227933e-07, + "loss": 0.5452, + "step": 9746 + }, + { + "epoch": 0.8078740157480315, + "grad_norm": 0.2386871576309204, + "learning_rate": 9.61044343141318e-07, + "loss": 0.4607, + "step": 9747 + }, + { + "epoch": 0.8079569001243265, + "grad_norm": 0.24201956391334534, + "learning_rate": 9.606299212598426e-07, + "loss": 0.5054, + "step": 9748 + }, + { + "epoch": 0.8080397845006216, + "grad_norm": 0.2298676073551178, + "learning_rate": 9.602154993783672e-07, + "loss": 0.4509, + "step": 9749 + }, + { + "epoch": 0.8081226688769167, + "grad_norm": 0.2312229424715042, + "learning_rate": 9.598010774968919e-07, + "loss": 0.468, + "step": 9750 + }, + { + "epoch": 0.8082055532532118, + "grad_norm": 0.23667356371879578, + "learning_rate": 9.593866556154165e-07, + "loss": 0.4838, + "step": 9751 + }, + { + "epoch": 0.8082884376295069, + "grad_norm": 0.24855121970176697, + "learning_rate": 9.589722337339411e-07, + "loss": 0.5311, + "step": 9752 + }, + { + "epoch": 0.8083713220058019, + "grad_norm": 0.3076818585395813, + "learning_rate": 9.58557811852466e-07, + "loss": 0.5161, + "step": 9753 + }, + { + "epoch": 0.808454206382097, + "grad_norm": 0.24067382514476776, + "learning_rate": 9.581433899709904e-07, + "loss": 0.4888, + "step": 9754 + }, + { + "epoch": 0.8085370907583921, + "grad_norm": 0.23676489293575287, + "learning_rate": 9.577289680895153e-07, + "loss": 0.4987, + "step": 9755 + }, + { + "epoch": 0.8086199751346871, + "grad_norm": 0.22398808598518372, + "learning_rate": 9.5731454620804e-07, + "loss": 0.4795, + "step": 9756 + }, + { + "epoch": 0.8087028595109822, + "grad_norm": 0.2648778557777405, + "learning_rate": 9.569001243265646e-07, + "loss": 0.554, + "step": 9757 + }, + { + "epoch": 0.8087857438872772, + "grad_norm": 0.2565188407897949, + "learning_rate": 9.564857024450892e-07, + "loss": 0.5078, + "step": 9758 + }, + { + "epoch": 0.8088686282635723, + "grad_norm": 0.24113653600215912, + "learning_rate": 9.560712805636139e-07, + "loss": 0.4721, + "step": 9759 + }, + { + "epoch": 0.8089515126398674, + "grad_norm": 0.25198253989219666, + "learning_rate": 9.556568586821385e-07, + "loss": 0.5167, + "step": 9760 + }, + { + "epoch": 0.8090343970161624, + "grad_norm": 0.26177191734313965, + "learning_rate": 9.552424368006632e-07, + "loss": 0.5028, + "step": 9761 + }, + { + "epoch": 0.8091172813924575, + "grad_norm": 0.22872452437877655, + "learning_rate": 9.548280149191878e-07, + "loss": 0.4647, + "step": 9762 + }, + { + "epoch": 0.8092001657687526, + "grad_norm": 0.2349000871181488, + "learning_rate": 9.544135930377124e-07, + "loss": 0.5023, + "step": 9763 + }, + { + "epoch": 0.8092830501450476, + "grad_norm": 0.25382840633392334, + "learning_rate": 9.53999171156237e-07, + "loss": 0.5013, + "step": 9764 + }, + { + "epoch": 0.8093659345213428, + "grad_norm": 0.2590975761413574, + "learning_rate": 9.535847492747617e-07, + "loss": 0.5095, + "step": 9765 + }, + { + "epoch": 0.8094488188976378, + "grad_norm": 0.25359734892845154, + "learning_rate": 9.531703273932865e-07, + "loss": 0.5002, + "step": 9766 + }, + { + "epoch": 0.8095317032739329, + "grad_norm": 0.24864192306995392, + "learning_rate": 9.527559055118111e-07, + "loss": 0.5487, + "step": 9767 + }, + { + "epoch": 0.809614587650228, + "grad_norm": 0.24689893424510956, + "learning_rate": 9.523414836303358e-07, + "loss": 0.5201, + "step": 9768 + }, + { + "epoch": 0.809697472026523, + "grad_norm": 0.23221725225448608, + "learning_rate": 9.519270617488604e-07, + "loss": 0.5151, + "step": 9769 + }, + { + "epoch": 0.8097803564028181, + "grad_norm": 0.255401611328125, + "learning_rate": 9.51512639867385e-07, + "loss": 0.5253, + "step": 9770 + }, + { + "epoch": 0.8098632407791131, + "grad_norm": 0.27492761611938477, + "learning_rate": 9.510982179859098e-07, + "loss": 0.4982, + "step": 9771 + }, + { + "epoch": 0.8099461251554082, + "grad_norm": 0.22783048450946808, + "learning_rate": 9.506837961044343e-07, + "loss": 0.4681, + "step": 9772 + }, + { + "epoch": 0.8100290095317033, + "grad_norm": 0.24867813289165497, + "learning_rate": 9.502693742229591e-07, + "loss": 0.4868, + "step": 9773 + }, + { + "epoch": 0.8101118939079983, + "grad_norm": 0.2436957061290741, + "learning_rate": 9.498549523414836e-07, + "loss": 0.5145, + "step": 9774 + }, + { + "epoch": 0.8101947782842934, + "grad_norm": 0.24859970808029175, + "learning_rate": 9.494405304600084e-07, + "loss": 0.5317, + "step": 9775 + }, + { + "epoch": 0.8102776626605884, + "grad_norm": 0.2614707052707672, + "learning_rate": 9.490261085785331e-07, + "loss": 0.5109, + "step": 9776 + }, + { + "epoch": 0.8103605470368835, + "grad_norm": 0.24609072506427765, + "learning_rate": 9.486116866970577e-07, + "loss": 0.5144, + "step": 9777 + }, + { + "epoch": 0.8104434314131786, + "grad_norm": 0.2370235025882721, + "learning_rate": 9.481972648155824e-07, + "loss": 0.4694, + "step": 9778 + }, + { + "epoch": 0.8105263157894737, + "grad_norm": 0.24376636743545532, + "learning_rate": 9.47782842934107e-07, + "loss": 0.4949, + "step": 9779 + }, + { + "epoch": 0.8106092001657688, + "grad_norm": 0.24963639676570892, + "learning_rate": 9.473684210526317e-07, + "loss": 0.5359, + "step": 9780 + }, + { + "epoch": 0.8106920845420639, + "grad_norm": 0.21756446361541748, + "learning_rate": 9.469539991711562e-07, + "loss": 0.4552, + "step": 9781 + }, + { + "epoch": 0.8107749689183589, + "grad_norm": 0.24737116694450378, + "learning_rate": 9.46539577289681e-07, + "loss": 0.4894, + "step": 9782 + }, + { + "epoch": 0.810857853294654, + "grad_norm": 0.24245326220989227, + "learning_rate": 9.461251554082055e-07, + "loss": 0.4613, + "step": 9783 + }, + { + "epoch": 0.810940737670949, + "grad_norm": 0.24612176418304443, + "learning_rate": 9.457107335267303e-07, + "loss": 0.502, + "step": 9784 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 0.25450369715690613, + "learning_rate": 9.45296311645255e-07, + "loss": 0.5008, + "step": 9785 + }, + { + "epoch": 0.8111065064235392, + "grad_norm": 0.23810330033302307, + "learning_rate": 9.448818897637796e-07, + "loss": 0.4993, + "step": 9786 + }, + { + "epoch": 0.8111893907998342, + "grad_norm": 0.23540052771568298, + "learning_rate": 9.444674678823043e-07, + "loss": 0.4803, + "step": 9787 + }, + { + "epoch": 0.8112722751761293, + "grad_norm": 0.28143590688705444, + "learning_rate": 9.440530460008288e-07, + "loss": 0.4966, + "step": 9788 + }, + { + "epoch": 0.8113551595524243, + "grad_norm": 0.24976450204849243, + "learning_rate": 9.436386241193536e-07, + "loss": 0.4934, + "step": 9789 + }, + { + "epoch": 0.8114380439287194, + "grad_norm": 0.23398953676223755, + "learning_rate": 9.432242022378782e-07, + "loss": 0.469, + "step": 9790 + }, + { + "epoch": 0.8115209283050145, + "grad_norm": 0.255262166261673, + "learning_rate": 9.428097803564029e-07, + "loss": 0.4803, + "step": 9791 + }, + { + "epoch": 0.8116038126813095, + "grad_norm": 0.24841533601284027, + "learning_rate": 9.423953584749275e-07, + "loss": 0.4974, + "step": 9792 + }, + { + "epoch": 0.8116866970576047, + "grad_norm": 0.226023331284523, + "learning_rate": 9.419809365934522e-07, + "loss": 0.446, + "step": 9793 + }, + { + "epoch": 0.8117695814338997, + "grad_norm": 0.2339816391468048, + "learning_rate": 9.415665147119769e-07, + "loss": 0.5085, + "step": 9794 + }, + { + "epoch": 0.8118524658101948, + "grad_norm": 0.239628866314888, + "learning_rate": 9.411520928305016e-07, + "loss": 0.4734, + "step": 9795 + }, + { + "epoch": 0.8119353501864899, + "grad_norm": 0.23718571662902832, + "learning_rate": 9.407376709490262e-07, + "loss": 0.4713, + "step": 9796 + }, + { + "epoch": 0.8120182345627849, + "grad_norm": 0.22454044222831726, + "learning_rate": 9.403232490675508e-07, + "loss": 0.4621, + "step": 9797 + }, + { + "epoch": 0.81210111893908, + "grad_norm": 0.22715407609939575, + "learning_rate": 9.399088271860755e-07, + "loss": 0.4958, + "step": 9798 + }, + { + "epoch": 0.8121840033153751, + "grad_norm": 0.23666779696941376, + "learning_rate": 9.394944053046001e-07, + "loss": 0.4678, + "step": 9799 + }, + { + "epoch": 0.8122668876916701, + "grad_norm": 0.2325560599565506, + "learning_rate": 9.390799834231249e-07, + "loss": 0.4395, + "step": 9800 + }, + { + "epoch": 0.8123497720679652, + "grad_norm": 0.24637115001678467, + "learning_rate": 9.386655615416494e-07, + "loss": 0.4803, + "step": 9801 + }, + { + "epoch": 0.8124326564442602, + "grad_norm": 0.24222879111766815, + "learning_rate": 9.382511396601742e-07, + "loss": 0.5026, + "step": 9802 + }, + { + "epoch": 0.8125155408205553, + "grad_norm": 0.23308725655078888, + "learning_rate": 9.378367177786987e-07, + "loss": 0.5153, + "step": 9803 + }, + { + "epoch": 0.8125984251968504, + "grad_norm": 0.23619213700294495, + "learning_rate": 9.374222958972235e-07, + "loss": 0.4573, + "step": 9804 + }, + { + "epoch": 0.8126813095731454, + "grad_norm": 0.2351343035697937, + "learning_rate": 9.370078740157482e-07, + "loss": 0.4896, + "step": 9805 + }, + { + "epoch": 0.8127641939494405, + "grad_norm": 0.23225870728492737, + "learning_rate": 9.365934521342727e-07, + "loss": 0.4881, + "step": 9806 + }, + { + "epoch": 0.8128470783257356, + "grad_norm": 0.24678654968738556, + "learning_rate": 9.361790302527975e-07, + "loss": 0.4498, + "step": 9807 + }, + { + "epoch": 0.8129299627020307, + "grad_norm": 0.23857372999191284, + "learning_rate": 9.35764608371322e-07, + "loss": 0.5137, + "step": 9808 + }, + { + "epoch": 0.8130128470783258, + "grad_norm": 0.2695607542991638, + "learning_rate": 9.353501864898468e-07, + "loss": 0.5279, + "step": 9809 + }, + { + "epoch": 0.8130957314546208, + "grad_norm": 0.23342856764793396, + "learning_rate": 9.349357646083713e-07, + "loss": 0.4736, + "step": 9810 + }, + { + "epoch": 0.8131786158309159, + "grad_norm": 0.23133207857608795, + "learning_rate": 9.345213427268961e-07, + "loss": 0.4553, + "step": 9811 + }, + { + "epoch": 0.8132615002072109, + "grad_norm": 0.25838544964790344, + "learning_rate": 9.341069208454206e-07, + "loss": 0.5308, + "step": 9812 + }, + { + "epoch": 0.813344384583506, + "grad_norm": 0.22402246296405792, + "learning_rate": 9.336924989639454e-07, + "loss": 0.4594, + "step": 9813 + }, + { + "epoch": 0.8134272689598011, + "grad_norm": 0.24499014019966125, + "learning_rate": 9.332780770824701e-07, + "loss": 0.5331, + "step": 9814 + }, + { + "epoch": 0.8135101533360961, + "grad_norm": 0.24721474945545197, + "learning_rate": 9.328636552009946e-07, + "loss": 0.481, + "step": 9815 + }, + { + "epoch": 0.8135930377123912, + "grad_norm": 0.25553810596466064, + "learning_rate": 9.324492333195194e-07, + "loss": 0.4974, + "step": 9816 + }, + { + "epoch": 0.8136759220886863, + "grad_norm": 0.23129479587078094, + "learning_rate": 9.320348114380439e-07, + "loss": 0.4839, + "step": 9817 + }, + { + "epoch": 0.8137588064649813, + "grad_norm": 0.23772019147872925, + "learning_rate": 9.316203895565687e-07, + "loss": 0.5186, + "step": 9818 + }, + { + "epoch": 0.8138416908412764, + "grad_norm": 0.25991588830947876, + "learning_rate": 9.312059676750933e-07, + "loss": 0.52, + "step": 9819 + }, + { + "epoch": 0.8139245752175714, + "grad_norm": 0.2358582466840744, + "learning_rate": 9.30791545793618e-07, + "loss": 0.4778, + "step": 9820 + }, + { + "epoch": 0.8140074595938666, + "grad_norm": 0.2404380887746811, + "learning_rate": 9.303771239121426e-07, + "loss": 0.4942, + "step": 9821 + }, + { + "epoch": 0.8140903439701617, + "grad_norm": 0.265122652053833, + "learning_rate": 9.299627020306673e-07, + "loss": 0.5276, + "step": 9822 + }, + { + "epoch": 0.8141732283464567, + "grad_norm": 0.24760957062244415, + "learning_rate": 9.29548280149192e-07, + "loss": 0.4857, + "step": 9823 + }, + { + "epoch": 0.8142561127227518, + "grad_norm": 0.23587417602539062, + "learning_rate": 9.291338582677166e-07, + "loss": 0.5033, + "step": 9824 + }, + { + "epoch": 0.8143389970990468, + "grad_norm": 0.34862032532691956, + "learning_rate": 9.287194363862413e-07, + "loss": 0.4801, + "step": 9825 + }, + { + "epoch": 0.8144218814753419, + "grad_norm": 0.23797102272510529, + "learning_rate": 9.283050145047659e-07, + "loss": 0.5151, + "step": 9826 + }, + { + "epoch": 0.814504765851637, + "grad_norm": 0.2501679062843323, + "learning_rate": 9.278905926232906e-07, + "loss": 0.4575, + "step": 9827 + }, + { + "epoch": 0.814587650227932, + "grad_norm": 0.2430824637413025, + "learning_rate": 9.274761707418152e-07, + "loss": 0.4842, + "step": 9828 + }, + { + "epoch": 0.8146705346042271, + "grad_norm": 0.23864732682704926, + "learning_rate": 9.2706174886034e-07, + "loss": 0.4857, + "step": 9829 + }, + { + "epoch": 0.8147534189805222, + "grad_norm": 0.2502564489841461, + "learning_rate": 9.266473269788645e-07, + "loss": 0.5425, + "step": 9830 + }, + { + "epoch": 0.8148363033568172, + "grad_norm": 0.2306893765926361, + "learning_rate": 9.262329050973893e-07, + "loss": 0.5017, + "step": 9831 + }, + { + "epoch": 0.8149191877331123, + "grad_norm": 0.2212308943271637, + "learning_rate": 9.258184832159139e-07, + "loss": 0.4827, + "step": 9832 + }, + { + "epoch": 0.8150020721094073, + "grad_norm": 0.23482365906238556, + "learning_rate": 9.254040613344385e-07, + "loss": 0.4758, + "step": 9833 + }, + { + "epoch": 0.8150849564857024, + "grad_norm": 0.2422025203704834, + "learning_rate": 9.249896394529633e-07, + "loss": 0.5102, + "step": 9834 + }, + { + "epoch": 0.8151678408619976, + "grad_norm": 0.25206252932548523, + "learning_rate": 9.245752175714878e-07, + "loss": 0.4969, + "step": 9835 + }, + { + "epoch": 0.8152507252382926, + "grad_norm": 0.24675221741199493, + "learning_rate": 9.241607956900126e-07, + "loss": 0.4737, + "step": 9836 + }, + { + "epoch": 0.8153336096145877, + "grad_norm": 0.23617562651634216, + "learning_rate": 9.237463738085371e-07, + "loss": 0.4578, + "step": 9837 + }, + { + "epoch": 0.8154164939908827, + "grad_norm": 0.24961379170417786, + "learning_rate": 9.233319519270619e-07, + "loss": 0.5166, + "step": 9838 + }, + { + "epoch": 0.8154993783671778, + "grad_norm": 0.251310259103775, + "learning_rate": 9.229175300455864e-07, + "loss": 0.4919, + "step": 9839 + }, + { + "epoch": 0.8155822627434729, + "grad_norm": 0.2561209797859192, + "learning_rate": 9.225031081641112e-07, + "loss": 0.5489, + "step": 9840 + }, + { + "epoch": 0.8156651471197679, + "grad_norm": 0.24602286517620087, + "learning_rate": 9.220886862826357e-07, + "loss": 0.5029, + "step": 9841 + }, + { + "epoch": 0.815748031496063, + "grad_norm": 0.2396889477968216, + "learning_rate": 9.216742644011604e-07, + "loss": 0.4542, + "step": 9842 + }, + { + "epoch": 0.815830915872358, + "grad_norm": 0.2659671902656555, + "learning_rate": 9.212598425196852e-07, + "loss": 0.5197, + "step": 9843 + }, + { + "epoch": 0.8159138002486531, + "grad_norm": 0.24347516894340515, + "learning_rate": 9.208454206382097e-07, + "loss": 0.4851, + "step": 9844 + }, + { + "epoch": 0.8159966846249482, + "grad_norm": 0.2328321635723114, + "learning_rate": 9.204309987567345e-07, + "loss": 0.4616, + "step": 9845 + }, + { + "epoch": 0.8160795690012432, + "grad_norm": 0.24348361790180206, + "learning_rate": 9.20016576875259e-07, + "loss": 0.5068, + "step": 9846 + }, + { + "epoch": 0.8161624533775383, + "grad_norm": 0.22897978127002716, + "learning_rate": 9.196021549937838e-07, + "loss": 0.4938, + "step": 9847 + }, + { + "epoch": 0.8162453377538335, + "grad_norm": 0.2559409737586975, + "learning_rate": 9.191877331123083e-07, + "loss": 0.5069, + "step": 9848 + }, + { + "epoch": 0.8163282221301285, + "grad_norm": 0.22506742179393768, + "learning_rate": 9.187733112308331e-07, + "loss": 0.5203, + "step": 9849 + }, + { + "epoch": 0.8164111065064236, + "grad_norm": 0.24295979738235474, + "learning_rate": 9.183588893493577e-07, + "loss": 0.5019, + "step": 9850 + }, + { + "epoch": 0.8164939908827186, + "grad_norm": 0.2402399331331253, + "learning_rate": 9.179444674678823e-07, + "loss": 0.467, + "step": 9851 + }, + { + "epoch": 0.8165768752590137, + "grad_norm": 0.23872220516204834, + "learning_rate": 9.175300455864071e-07, + "loss": 0.517, + "step": 9852 + }, + { + "epoch": 0.8166597596353088, + "grad_norm": 0.2538854479789734, + "learning_rate": 9.171156237049316e-07, + "loss": 0.515, + "step": 9853 + }, + { + "epoch": 0.8167426440116038, + "grad_norm": 0.23573583364486694, + "learning_rate": 9.167012018234564e-07, + "loss": 0.4718, + "step": 9854 + }, + { + "epoch": 0.8168255283878989, + "grad_norm": 0.22748109698295593, + "learning_rate": 9.16286779941981e-07, + "loss": 0.486, + "step": 9855 + }, + { + "epoch": 0.8169084127641939, + "grad_norm": 0.25185221433639526, + "learning_rate": 9.158723580605057e-07, + "loss": 0.5189, + "step": 9856 + }, + { + "epoch": 0.816991297140489, + "grad_norm": 0.25337889790534973, + "learning_rate": 9.154579361790303e-07, + "loss": 0.561, + "step": 9857 + }, + { + "epoch": 0.8170741815167841, + "grad_norm": 0.25118666887283325, + "learning_rate": 9.15043514297555e-07, + "loss": 0.5087, + "step": 9858 + }, + { + "epoch": 0.8171570658930791, + "grad_norm": 0.24642157554626465, + "learning_rate": 9.146290924160796e-07, + "loss": 0.4746, + "step": 9859 + }, + { + "epoch": 0.8172399502693742, + "grad_norm": 0.2344837486743927, + "learning_rate": 9.142146705346043e-07, + "loss": 0.4972, + "step": 9860 + }, + { + "epoch": 0.8173228346456692, + "grad_norm": 0.2437809407711029, + "learning_rate": 9.13800248653129e-07, + "loss": 0.506, + "step": 9861 + }, + { + "epoch": 0.8174057190219643, + "grad_norm": 0.22572484612464905, + "learning_rate": 9.133858267716536e-07, + "loss": 0.4551, + "step": 9862 + }, + { + "epoch": 0.8174886033982595, + "grad_norm": 0.24301820993423462, + "learning_rate": 9.129714048901783e-07, + "loss": 0.5087, + "step": 9863 + }, + { + "epoch": 0.8175714877745545, + "grad_norm": 0.2653495967388153, + "learning_rate": 9.125569830087029e-07, + "loss": 0.5003, + "step": 9864 + }, + { + "epoch": 0.8176543721508496, + "grad_norm": 0.23124350607395172, + "learning_rate": 9.121425611272277e-07, + "loss": 0.4538, + "step": 9865 + }, + { + "epoch": 0.8177372565271447, + "grad_norm": 0.2286759614944458, + "learning_rate": 9.117281392457522e-07, + "loss": 0.4406, + "step": 9866 + }, + { + "epoch": 0.8178201409034397, + "grad_norm": 0.2541085481643677, + "learning_rate": 9.11313717364277e-07, + "loss": 0.5189, + "step": 9867 + }, + { + "epoch": 0.8179030252797348, + "grad_norm": 0.2258886992931366, + "learning_rate": 9.108992954828015e-07, + "loss": 0.4809, + "step": 9868 + }, + { + "epoch": 0.8179859096560298, + "grad_norm": 0.23116883635520935, + "learning_rate": 9.104848736013262e-07, + "loss": 0.4637, + "step": 9869 + }, + { + "epoch": 0.8180687940323249, + "grad_norm": 0.229501873254776, + "learning_rate": 9.10070451719851e-07, + "loss": 0.4885, + "step": 9870 + }, + { + "epoch": 0.81815167840862, + "grad_norm": 0.22219321131706238, + "learning_rate": 9.096560298383755e-07, + "loss": 0.4714, + "step": 9871 + }, + { + "epoch": 0.818234562784915, + "grad_norm": 0.22711151838302612, + "learning_rate": 9.092416079569003e-07, + "loss": 0.4613, + "step": 9872 + }, + { + "epoch": 0.8183174471612101, + "grad_norm": 0.24113386869430542, + "learning_rate": 9.088271860754248e-07, + "loss": 0.4808, + "step": 9873 + }, + { + "epoch": 0.8184003315375051, + "grad_norm": 0.2574576437473297, + "learning_rate": 9.084127641939496e-07, + "loss": 0.5362, + "step": 9874 + }, + { + "epoch": 0.8184832159138002, + "grad_norm": 0.22775717079639435, + "learning_rate": 9.079983423124741e-07, + "loss": 0.4474, + "step": 9875 + }, + { + "epoch": 0.8185661002900954, + "grad_norm": 0.2447051852941513, + "learning_rate": 9.075839204309989e-07, + "loss": 0.4466, + "step": 9876 + }, + { + "epoch": 0.8186489846663904, + "grad_norm": 0.24979224801063538, + "learning_rate": 9.071694985495234e-07, + "loss": 0.5281, + "step": 9877 + }, + { + "epoch": 0.8187318690426855, + "grad_norm": 0.2761164605617523, + "learning_rate": 9.067550766680481e-07, + "loss": 0.5286, + "step": 9878 + }, + { + "epoch": 0.8188147534189805, + "grad_norm": 0.24827733635902405, + "learning_rate": 9.063406547865728e-07, + "loss": 0.502, + "step": 9879 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 0.23283758759498596, + "learning_rate": 9.059262329050974e-07, + "loss": 0.5021, + "step": 9880 + }, + { + "epoch": 0.8189805221715707, + "grad_norm": 0.240292489528656, + "learning_rate": 9.055118110236222e-07, + "loss": 0.4948, + "step": 9881 + }, + { + "epoch": 0.8190634065478657, + "grad_norm": 0.2490752786397934, + "learning_rate": 9.050973891421467e-07, + "loss": 0.4893, + "step": 9882 + }, + { + "epoch": 0.8191462909241608, + "grad_norm": 0.2527432143688202, + "learning_rate": 9.046829672606715e-07, + "loss": 0.5355, + "step": 9883 + }, + { + "epoch": 0.8192291753004559, + "grad_norm": 0.25967615842819214, + "learning_rate": 9.042685453791961e-07, + "loss": 0.5029, + "step": 9884 + }, + { + "epoch": 0.8193120596767509, + "grad_norm": 0.23246918618679047, + "learning_rate": 9.038541234977208e-07, + "loss": 0.4793, + "step": 9885 + }, + { + "epoch": 0.819394944053046, + "grad_norm": 0.22177287936210632, + "learning_rate": 9.034397016162454e-07, + "loss": 0.4246, + "step": 9886 + }, + { + "epoch": 0.819477828429341, + "grad_norm": 0.21613679826259613, + "learning_rate": 9.0302527973477e-07, + "loss": 0.4466, + "step": 9887 + }, + { + "epoch": 0.8195607128056361, + "grad_norm": 0.24387206137180328, + "learning_rate": 9.026108578532947e-07, + "loss": 0.4783, + "step": 9888 + }, + { + "epoch": 0.8196435971819312, + "grad_norm": 0.23413144052028656, + "learning_rate": 9.021964359718194e-07, + "loss": 0.4915, + "step": 9889 + }, + { + "epoch": 0.8197264815582262, + "grad_norm": 0.24385516345500946, + "learning_rate": 9.017820140903441e-07, + "loss": 0.4767, + "step": 9890 + }, + { + "epoch": 0.8198093659345214, + "grad_norm": 0.27351492643356323, + "learning_rate": 9.013675922088687e-07, + "loss": 0.5458, + "step": 9891 + }, + { + "epoch": 0.8198922503108164, + "grad_norm": 0.2560613453388214, + "learning_rate": 9.009531703273934e-07, + "loss": 0.4883, + "step": 9892 + }, + { + "epoch": 0.8199751346871115, + "grad_norm": 0.2403527945280075, + "learning_rate": 9.00538748445918e-07, + "loss": 0.4655, + "step": 9893 + }, + { + "epoch": 0.8200580190634066, + "grad_norm": 0.25028523802757263, + "learning_rate": 9.001243265644428e-07, + "loss": 0.5105, + "step": 9894 + }, + { + "epoch": 0.8201409034397016, + "grad_norm": 0.23240794241428375, + "learning_rate": 8.997099046829673e-07, + "loss": 0.4853, + "step": 9895 + }, + { + "epoch": 0.8202237878159967, + "grad_norm": 0.24330028891563416, + "learning_rate": 8.99295482801492e-07, + "loss": 0.5026, + "step": 9896 + }, + { + "epoch": 0.8203066721922918, + "grad_norm": 0.22840146720409393, + "learning_rate": 8.988810609200166e-07, + "loss": 0.4836, + "step": 9897 + }, + { + "epoch": 0.8203895565685868, + "grad_norm": 0.24885053932666779, + "learning_rate": 8.984666390385413e-07, + "loss": 0.4987, + "step": 9898 + }, + { + "epoch": 0.8204724409448819, + "grad_norm": 0.24369047582149506, + "learning_rate": 8.980522171570661e-07, + "loss": 0.5062, + "step": 9899 + }, + { + "epoch": 0.8205553253211769, + "grad_norm": 0.24177931249141693, + "learning_rate": 8.976377952755906e-07, + "loss": 0.5079, + "step": 9900 + }, + { + "epoch": 0.820638209697472, + "grad_norm": 0.2487955540418625, + "learning_rate": 8.972233733941154e-07, + "loss": 0.532, + "step": 9901 + }, + { + "epoch": 0.8207210940737671, + "grad_norm": 0.23817600309848785, + "learning_rate": 8.968089515126399e-07, + "loss": 0.4678, + "step": 9902 + }, + { + "epoch": 0.8208039784500621, + "grad_norm": 0.22956407070159912, + "learning_rate": 8.963945296311647e-07, + "loss": 0.483, + "step": 9903 + }, + { + "epoch": 0.8208868628263573, + "grad_norm": 0.252402126789093, + "learning_rate": 8.959801077496892e-07, + "loss": 0.4698, + "step": 9904 + }, + { + "epoch": 0.8209697472026523, + "grad_norm": 0.24298320710659027, + "learning_rate": 8.955656858682139e-07, + "loss": 0.4983, + "step": 9905 + }, + { + "epoch": 0.8210526315789474, + "grad_norm": 0.24302884936332703, + "learning_rate": 8.951512639867385e-07, + "loss": 0.5319, + "step": 9906 + }, + { + "epoch": 0.8211355159552425, + "grad_norm": 0.23893876373767853, + "learning_rate": 8.947368421052632e-07, + "loss": 0.5013, + "step": 9907 + }, + { + "epoch": 0.8212184003315375, + "grad_norm": 0.25012749433517456, + "learning_rate": 8.94322420223788e-07, + "loss": 0.4922, + "step": 9908 + }, + { + "epoch": 0.8213012847078326, + "grad_norm": 0.22894440591335297, + "learning_rate": 8.939079983423125e-07, + "loss": 0.477, + "step": 9909 + }, + { + "epoch": 0.8213841690841276, + "grad_norm": 0.2703799605369568, + "learning_rate": 8.934935764608373e-07, + "loss": 0.5318, + "step": 9910 + }, + { + "epoch": 0.8214670534604227, + "grad_norm": 0.22368276119232178, + "learning_rate": 8.930791545793618e-07, + "loss": 0.4531, + "step": 9911 + }, + { + "epoch": 0.8215499378367178, + "grad_norm": 0.2482854127883911, + "learning_rate": 8.926647326978866e-07, + "loss": 0.494, + "step": 9912 + }, + { + "epoch": 0.8216328222130128, + "grad_norm": 0.25626158714294434, + "learning_rate": 8.922503108164111e-07, + "loss": 0.4397, + "step": 9913 + }, + { + "epoch": 0.8217157065893079, + "grad_norm": 0.2422264814376831, + "learning_rate": 8.918358889349358e-07, + "loss": 0.5016, + "step": 9914 + }, + { + "epoch": 0.821798590965603, + "grad_norm": 0.22546489536762238, + "learning_rate": 8.914214670534605e-07, + "loss": 0.4755, + "step": 9915 + }, + { + "epoch": 0.821881475341898, + "grad_norm": 0.2561276853084564, + "learning_rate": 8.910070451719851e-07, + "loss": 0.5244, + "step": 9916 + }, + { + "epoch": 0.8219643597181931, + "grad_norm": 0.23983876407146454, + "learning_rate": 8.905926232905098e-07, + "loss": 0.5016, + "step": 9917 + }, + { + "epoch": 0.8220472440944881, + "grad_norm": 0.23900650441646576, + "learning_rate": 8.901782014090344e-07, + "loss": 0.5242, + "step": 9918 + }, + { + "epoch": 0.8221301284707833, + "grad_norm": 0.2402048408985138, + "learning_rate": 8.897637795275592e-07, + "loss": 0.4996, + "step": 9919 + }, + { + "epoch": 0.8222130128470784, + "grad_norm": 0.25700998306274414, + "learning_rate": 8.893493576460838e-07, + "loss": 0.5248, + "step": 9920 + }, + { + "epoch": 0.8222958972233734, + "grad_norm": 0.2415207028388977, + "learning_rate": 8.889349357646085e-07, + "loss": 0.496, + "step": 9921 + }, + { + "epoch": 0.8223787815996685, + "grad_norm": 0.2320256531238556, + "learning_rate": 8.885205138831331e-07, + "loss": 0.4751, + "step": 9922 + }, + { + "epoch": 0.8224616659759635, + "grad_norm": 0.253726065158844, + "learning_rate": 8.881060920016577e-07, + "loss": 0.5463, + "step": 9923 + }, + { + "epoch": 0.8225445503522586, + "grad_norm": 0.24048392474651337, + "learning_rate": 8.876916701201824e-07, + "loss": 0.518, + "step": 9924 + }, + { + "epoch": 0.8226274347285537, + "grad_norm": 0.2296401560306549, + "learning_rate": 8.872772482387071e-07, + "loss": 0.4549, + "step": 9925 + }, + { + "epoch": 0.8227103191048487, + "grad_norm": 0.24883905053138733, + "learning_rate": 8.868628263572317e-07, + "loss": 0.5096, + "step": 9926 + }, + { + "epoch": 0.8227932034811438, + "grad_norm": 0.24362295866012573, + "learning_rate": 8.864484044757564e-07, + "loss": 0.5103, + "step": 9927 + }, + { + "epoch": 0.8228760878574388, + "grad_norm": 0.23714140057563782, + "learning_rate": 8.860339825942811e-07, + "loss": 0.5262, + "step": 9928 + }, + { + "epoch": 0.8229589722337339, + "grad_norm": 0.24650827050209045, + "learning_rate": 8.856195607128057e-07, + "loss": 0.5082, + "step": 9929 + }, + { + "epoch": 0.823041856610029, + "grad_norm": 0.24677348136901855, + "learning_rate": 8.852051388313305e-07, + "loss": 0.5287, + "step": 9930 + }, + { + "epoch": 0.823124740986324, + "grad_norm": 0.2393384575843811, + "learning_rate": 8.84790716949855e-07, + "loss": 0.5329, + "step": 9931 + }, + { + "epoch": 0.8232076253626192, + "grad_norm": 0.25296422839164734, + "learning_rate": 8.843762950683797e-07, + "loss": 0.535, + "step": 9932 + }, + { + "epoch": 0.8232905097389143, + "grad_norm": 0.2995743155479431, + "learning_rate": 8.839618731869043e-07, + "loss": 0.525, + "step": 9933 + }, + { + "epoch": 0.8233733941152093, + "grad_norm": 0.25555384159088135, + "learning_rate": 8.83547451305429e-07, + "loss": 0.546, + "step": 9934 + }, + { + "epoch": 0.8234562784915044, + "grad_norm": 0.2503458857536316, + "learning_rate": 8.831330294239536e-07, + "loss": 0.5089, + "step": 9935 + }, + { + "epoch": 0.8235391628677994, + "grad_norm": 0.23157715797424316, + "learning_rate": 8.827186075424783e-07, + "loss": 0.4741, + "step": 9936 + }, + { + "epoch": 0.8236220472440945, + "grad_norm": 0.22015385329723358, + "learning_rate": 8.823041856610031e-07, + "loss": 0.4626, + "step": 9937 + }, + { + "epoch": 0.8237049316203896, + "grad_norm": 0.2478223294019699, + "learning_rate": 8.818897637795276e-07, + "loss": 0.4761, + "step": 9938 + }, + { + "epoch": 0.8237878159966846, + "grad_norm": 0.2509728968143463, + "learning_rate": 8.814753418980524e-07, + "loss": 0.5054, + "step": 9939 + }, + { + "epoch": 0.8238707003729797, + "grad_norm": 0.24563109874725342, + "learning_rate": 8.810609200165769e-07, + "loss": 0.4698, + "step": 9940 + }, + { + "epoch": 0.8239535847492747, + "grad_norm": 0.24942804872989655, + "learning_rate": 8.806464981351016e-07, + "loss": 0.5097, + "step": 9941 + }, + { + "epoch": 0.8240364691255698, + "grad_norm": 0.24876244366168976, + "learning_rate": 8.802320762536262e-07, + "loss": 0.5016, + "step": 9942 + }, + { + "epoch": 0.8241193535018649, + "grad_norm": 0.2349521666765213, + "learning_rate": 8.798176543721509e-07, + "loss": 0.5174, + "step": 9943 + }, + { + "epoch": 0.8242022378781599, + "grad_norm": 0.25612425804138184, + "learning_rate": 8.794032324906756e-07, + "loss": 0.5399, + "step": 9944 + }, + { + "epoch": 0.824285122254455, + "grad_norm": 0.23037123680114746, + "learning_rate": 8.789888106092002e-07, + "loss": 0.5002, + "step": 9945 + }, + { + "epoch": 0.82436800663075, + "grad_norm": 0.24060919880867004, + "learning_rate": 8.78574388727725e-07, + "loss": 0.4932, + "step": 9946 + }, + { + "epoch": 0.8244508910070452, + "grad_norm": 0.22246460616588593, + "learning_rate": 8.781599668462495e-07, + "loss": 0.454, + "step": 9947 + }, + { + "epoch": 0.8245337753833403, + "grad_norm": 0.2464425265789032, + "learning_rate": 8.777455449647742e-07, + "loss": 0.5084, + "step": 9948 + }, + { + "epoch": 0.8246166597596353, + "grad_norm": 0.2544879913330078, + "learning_rate": 8.773311230832989e-07, + "loss": 0.5255, + "step": 9949 + }, + { + "epoch": 0.8246995441359304, + "grad_norm": 0.22848616540431976, + "learning_rate": 8.769167012018235e-07, + "loss": 0.4999, + "step": 9950 + }, + { + "epoch": 0.8247824285122255, + "grad_norm": 0.24528570473194122, + "learning_rate": 8.765022793203482e-07, + "loss": 0.4846, + "step": 9951 + }, + { + "epoch": 0.8248653128885205, + "grad_norm": 0.2400393933057785, + "learning_rate": 8.760878574388728e-07, + "loss": 0.4724, + "step": 9952 + }, + { + "epoch": 0.8249481972648156, + "grad_norm": 0.22867369651794434, + "learning_rate": 8.756734355573975e-07, + "loss": 0.484, + "step": 9953 + }, + { + "epoch": 0.8250310816411106, + "grad_norm": 0.23885202407836914, + "learning_rate": 8.752590136759222e-07, + "loss": 0.4914, + "step": 9954 + }, + { + "epoch": 0.8251139660174057, + "grad_norm": 0.23325105011463165, + "learning_rate": 8.748445917944468e-07, + "loss": 0.4593, + "step": 9955 + }, + { + "epoch": 0.8251968503937008, + "grad_norm": 1.9288480281829834, + "learning_rate": 8.744301699129715e-07, + "loss": 0.5005, + "step": 9956 + }, + { + "epoch": 0.8252797347699958, + "grad_norm": 0.2604454755783081, + "learning_rate": 8.740157480314961e-07, + "loss": 0.4592, + "step": 9957 + }, + { + "epoch": 0.8253626191462909, + "grad_norm": 0.2509896159172058, + "learning_rate": 8.736013261500208e-07, + "loss": 0.4776, + "step": 9958 + }, + { + "epoch": 0.8254455035225859, + "grad_norm": 0.2389155477285385, + "learning_rate": 8.731869042685455e-07, + "loss": 0.495, + "step": 9959 + }, + { + "epoch": 0.825528387898881, + "grad_norm": 0.2456994354724884, + "learning_rate": 8.727724823870701e-07, + "loss": 0.5101, + "step": 9960 + }, + { + "epoch": 0.8256112722751762, + "grad_norm": 0.2559318542480469, + "learning_rate": 8.723580605055948e-07, + "loss": 0.527, + "step": 9961 + }, + { + "epoch": 0.8256941566514712, + "grad_norm": 0.21561601758003235, + "learning_rate": 8.719436386241194e-07, + "loss": 0.4557, + "step": 9962 + }, + { + "epoch": 0.8257770410277663, + "grad_norm": 0.25343629717826843, + "learning_rate": 8.715292167426441e-07, + "loss": 0.533, + "step": 9963 + }, + { + "epoch": 0.8258599254040613, + "grad_norm": 0.22899749875068665, + "learning_rate": 8.711147948611687e-07, + "loss": 0.4846, + "step": 9964 + }, + { + "epoch": 0.8259428097803564, + "grad_norm": 0.24385316669940948, + "learning_rate": 8.707003729796934e-07, + "loss": 0.4659, + "step": 9965 + }, + { + "epoch": 0.8260256941566515, + "grad_norm": 0.24540655314922333, + "learning_rate": 8.702859510982182e-07, + "loss": 0.49, + "step": 9966 + }, + { + "epoch": 0.8261085785329465, + "grad_norm": 0.23359344899654388, + "learning_rate": 8.698715292167427e-07, + "loss": 0.4654, + "step": 9967 + }, + { + "epoch": 0.8261914629092416, + "grad_norm": 0.2463885247707367, + "learning_rate": 8.694571073352674e-07, + "loss": 0.4733, + "step": 9968 + }, + { + "epoch": 0.8262743472855367, + "grad_norm": 0.23997296392917633, + "learning_rate": 8.69042685453792e-07, + "loss": 0.4801, + "step": 9969 + }, + { + "epoch": 0.8263572316618317, + "grad_norm": 0.25957489013671875, + "learning_rate": 8.686282635723167e-07, + "loss": 0.5005, + "step": 9970 + }, + { + "epoch": 0.8264401160381268, + "grad_norm": 0.24692967534065247, + "learning_rate": 8.682138416908413e-07, + "loss": 0.5104, + "step": 9971 + }, + { + "epoch": 0.8265230004144218, + "grad_norm": 0.24130430817604065, + "learning_rate": 8.67799419809366e-07, + "loss": 0.4716, + "step": 9972 + }, + { + "epoch": 0.826605884790717, + "grad_norm": 0.22703434526920319, + "learning_rate": 8.673849979278907e-07, + "loss": 0.4683, + "step": 9973 + }, + { + "epoch": 0.8266887691670121, + "grad_norm": 0.23234739899635315, + "learning_rate": 8.669705760464153e-07, + "loss": 0.476, + "step": 9974 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 0.24907751381397247, + "learning_rate": 8.6655615416494e-07, + "loss": 0.5171, + "step": 9975 + }, + { + "epoch": 0.8268545379196022, + "grad_norm": 0.2489747256040573, + "learning_rate": 8.661417322834646e-07, + "loss": 0.4879, + "step": 9976 + }, + { + "epoch": 0.8269374222958972, + "grad_norm": 0.22923964262008667, + "learning_rate": 8.657273104019893e-07, + "loss": 0.4581, + "step": 9977 + }, + { + "epoch": 0.8270203066721923, + "grad_norm": 0.244643896818161, + "learning_rate": 8.65312888520514e-07, + "loss": 0.4771, + "step": 9978 + }, + { + "epoch": 0.8271031910484874, + "grad_norm": 0.24583986401557922, + "learning_rate": 8.648984666390386e-07, + "loss": 0.5017, + "step": 9979 + }, + { + "epoch": 0.8271860754247824, + "grad_norm": 0.2500803768634796, + "learning_rate": 8.644840447575633e-07, + "loss": 0.4931, + "step": 9980 + }, + { + "epoch": 0.8272689598010775, + "grad_norm": 0.238322913646698, + "learning_rate": 8.640696228760879e-07, + "loss": 0.483, + "step": 9981 + }, + { + "epoch": 0.8273518441773726, + "grad_norm": 0.24224679172039032, + "learning_rate": 8.636552009946126e-07, + "loss": 0.5078, + "step": 9982 + }, + { + "epoch": 0.8274347285536676, + "grad_norm": 0.2333906590938568, + "learning_rate": 8.632407791131372e-07, + "loss": 0.4717, + "step": 9983 + }, + { + "epoch": 0.8275176129299627, + "grad_norm": 0.2463604062795639, + "learning_rate": 8.62826357231662e-07, + "loss": 0.4787, + "step": 9984 + }, + { + "epoch": 0.8276004973062577, + "grad_norm": 0.24454137682914734, + "learning_rate": 8.624119353501866e-07, + "loss": 0.5105, + "step": 9985 + }, + { + "epoch": 0.8276833816825528, + "grad_norm": 0.23638348281383514, + "learning_rate": 8.619975134687112e-07, + "loss": 0.4401, + "step": 9986 + }, + { + "epoch": 0.827766266058848, + "grad_norm": 0.23480123281478882, + "learning_rate": 8.615830915872359e-07, + "loss": 0.4752, + "step": 9987 + }, + { + "epoch": 0.827849150435143, + "grad_norm": 0.24006669223308563, + "learning_rate": 8.611686697057605e-07, + "loss": 0.4709, + "step": 9988 + }, + { + "epoch": 0.8279320348114381, + "grad_norm": 0.25444698333740234, + "learning_rate": 8.607542478242852e-07, + "loss": 0.4772, + "step": 9989 + }, + { + "epoch": 0.8280149191877331, + "grad_norm": 0.2490633726119995, + "learning_rate": 8.603398259428099e-07, + "loss": 0.5156, + "step": 9990 + }, + { + "epoch": 0.8280978035640282, + "grad_norm": 0.2611975073814392, + "learning_rate": 8.599254040613345e-07, + "loss": 0.4699, + "step": 9991 + }, + { + "epoch": 0.8281806879403233, + "grad_norm": 0.24396789073944092, + "learning_rate": 8.595109821798592e-07, + "loss": 0.4756, + "step": 9992 + }, + { + "epoch": 0.8282635723166183, + "grad_norm": 0.22217494249343872, + "learning_rate": 8.590965602983837e-07, + "loss": 0.4383, + "step": 9993 + }, + { + "epoch": 0.8283464566929134, + "grad_norm": 0.24467700719833374, + "learning_rate": 8.586821384169085e-07, + "loss": 0.4955, + "step": 9994 + }, + { + "epoch": 0.8284293410692084, + "grad_norm": 0.25394171476364136, + "learning_rate": 8.582677165354332e-07, + "loss": 0.481, + "step": 9995 + }, + { + "epoch": 0.8285122254455035, + "grad_norm": 0.24946308135986328, + "learning_rate": 8.578532946539578e-07, + "loss": 0.5414, + "step": 9996 + }, + { + "epoch": 0.8285951098217986, + "grad_norm": 0.26028600335121155, + "learning_rate": 8.574388727724825e-07, + "loss": 0.4776, + "step": 9997 + }, + { + "epoch": 0.8286779941980936, + "grad_norm": 0.24516598880290985, + "learning_rate": 8.570244508910071e-07, + "loss": 0.494, + "step": 9998 + }, + { + "epoch": 0.8287608785743887, + "grad_norm": 0.29818570613861084, + "learning_rate": 8.566100290095318e-07, + "loss": 0.5234, + "step": 9999 + }, + { + "epoch": 0.8288437629506838, + "grad_norm": 0.2579766809940338, + "learning_rate": 8.561956071280563e-07, + "loss": 0.4924, + "step": 10000 + }, + { + "epoch": 0.8289266473269788, + "grad_norm": 0.23476462066173553, + "learning_rate": 8.557811852465811e-07, + "loss": 0.4899, + "step": 10001 + }, + { + "epoch": 0.829009531703274, + "grad_norm": 0.24617992341518402, + "learning_rate": 8.553667633651056e-07, + "loss": 0.5355, + "step": 10002 + }, + { + "epoch": 0.829092416079569, + "grad_norm": 0.23550431430339813, + "learning_rate": 8.549523414836304e-07, + "loss": 0.4924, + "step": 10003 + }, + { + "epoch": 0.8291753004558641, + "grad_norm": 0.2362932711839676, + "learning_rate": 8.545379196021551e-07, + "loss": 0.4648, + "step": 10004 + }, + { + "epoch": 0.8292581848321592, + "grad_norm": 0.24741438031196594, + "learning_rate": 8.541234977206797e-07, + "loss": 0.4984, + "step": 10005 + }, + { + "epoch": 0.8293410692084542, + "grad_norm": 0.23676697909832, + "learning_rate": 8.537090758392044e-07, + "loss": 0.4761, + "step": 10006 + }, + { + "epoch": 0.8294239535847493, + "grad_norm": 0.2454746812582016, + "learning_rate": 8.53294653957729e-07, + "loss": 0.4994, + "step": 10007 + }, + { + "epoch": 0.8295068379610443, + "grad_norm": 0.24545104801654816, + "learning_rate": 8.528802320762537e-07, + "loss": 0.5651, + "step": 10008 + }, + { + "epoch": 0.8295897223373394, + "grad_norm": 0.23125353455543518, + "learning_rate": 8.524658101947784e-07, + "loss": 0.4849, + "step": 10009 + }, + { + "epoch": 0.8296726067136345, + "grad_norm": 0.2496875822544098, + "learning_rate": 8.52051388313303e-07, + "loss": 0.4762, + "step": 10010 + }, + { + "epoch": 0.8297554910899295, + "grad_norm": 0.23437295854091644, + "learning_rate": 8.516369664318276e-07, + "loss": 0.4674, + "step": 10011 + }, + { + "epoch": 0.8298383754662246, + "grad_norm": 0.23719444870948792, + "learning_rate": 8.512225445503523e-07, + "loss": 0.4925, + "step": 10012 + }, + { + "epoch": 0.8299212598425196, + "grad_norm": 0.248506560921669, + "learning_rate": 8.50808122668877e-07, + "loss": 0.538, + "step": 10013 + }, + { + "epoch": 0.8300041442188147, + "grad_norm": 0.2648535668849945, + "learning_rate": 8.503937007874017e-07, + "loss": 0.4926, + "step": 10014 + }, + { + "epoch": 0.8300870285951099, + "grad_norm": 0.23467402160167694, + "learning_rate": 8.499792789059263e-07, + "loss": 0.5052, + "step": 10015 + }, + { + "epoch": 0.8301699129714049, + "grad_norm": 0.27154165506362915, + "learning_rate": 8.49564857024451e-07, + "loss": 0.5027, + "step": 10016 + }, + { + "epoch": 0.8302527973477, + "grad_norm": 0.2674507796764374, + "learning_rate": 8.491504351429756e-07, + "loss": 0.4911, + "step": 10017 + }, + { + "epoch": 0.8303356817239951, + "grad_norm": 0.2376575917005539, + "learning_rate": 8.487360132615003e-07, + "loss": 0.4836, + "step": 10018 + }, + { + "epoch": 0.8304185661002901, + "grad_norm": 0.22856204211711884, + "learning_rate": 8.48321591380025e-07, + "loss": 0.4709, + "step": 10019 + }, + { + "epoch": 0.8305014504765852, + "grad_norm": 0.24044018983840942, + "learning_rate": 8.479071694985495e-07, + "loss": 0.4876, + "step": 10020 + }, + { + "epoch": 0.8305843348528802, + "grad_norm": 0.23930101096630096, + "learning_rate": 8.474927476170743e-07, + "loss": 0.5143, + "step": 10021 + }, + { + "epoch": 0.8306672192291753, + "grad_norm": 0.2578476369380951, + "learning_rate": 8.470783257355989e-07, + "loss": 0.5116, + "step": 10022 + }, + { + "epoch": 0.8307501036054704, + "grad_norm": 0.2633090913295746, + "learning_rate": 8.466639038541236e-07, + "loss": 0.5185, + "step": 10023 + }, + { + "epoch": 0.8308329879817654, + "grad_norm": 0.23009124398231506, + "learning_rate": 8.462494819726483e-07, + "loss": 0.4708, + "step": 10024 + }, + { + "epoch": 0.8309158723580605, + "grad_norm": 0.23178869485855103, + "learning_rate": 8.458350600911729e-07, + "loss": 0.4609, + "step": 10025 + }, + { + "epoch": 0.8309987567343555, + "grad_norm": 0.22887535393238068, + "learning_rate": 8.454206382096976e-07, + "loss": 0.4699, + "step": 10026 + }, + { + "epoch": 0.8310816411106506, + "grad_norm": 0.2877136766910553, + "learning_rate": 8.450062163282221e-07, + "loss": 0.4958, + "step": 10027 + }, + { + "epoch": 0.8311645254869457, + "grad_norm": 0.22720719873905182, + "learning_rate": 8.445917944467469e-07, + "loss": 0.4503, + "step": 10028 + }, + { + "epoch": 0.8312474098632407, + "grad_norm": 0.24144597351551056, + "learning_rate": 8.441773725652714e-07, + "loss": 0.5063, + "step": 10029 + }, + { + "epoch": 0.8313302942395359, + "grad_norm": 0.2505830228328705, + "learning_rate": 8.437629506837962e-07, + "loss": 0.5041, + "step": 10030 + }, + { + "epoch": 0.8314131786158309, + "grad_norm": 0.2531561255455017, + "learning_rate": 8.433485288023207e-07, + "loss": 0.5034, + "step": 10031 + }, + { + "epoch": 0.831496062992126, + "grad_norm": 0.2341858595609665, + "learning_rate": 8.429341069208455e-07, + "loss": 0.4677, + "step": 10032 + }, + { + "epoch": 0.8315789473684211, + "grad_norm": 0.2314787060022354, + "learning_rate": 8.425196850393702e-07, + "loss": 0.4704, + "step": 10033 + }, + { + "epoch": 0.8316618317447161, + "grad_norm": 0.2543647587299347, + "learning_rate": 8.421052631578948e-07, + "loss": 0.5359, + "step": 10034 + }, + { + "epoch": 0.8317447161210112, + "grad_norm": 0.22283078730106354, + "learning_rate": 8.416908412764195e-07, + "loss": 0.487, + "step": 10035 + }, + { + "epoch": 0.8318276004973063, + "grad_norm": 0.24474400281906128, + "learning_rate": 8.41276419394944e-07, + "loss": 0.4632, + "step": 10036 + }, + { + "epoch": 0.8319104848736013, + "grad_norm": 0.2663940489292145, + "learning_rate": 8.408619975134688e-07, + "loss": 0.5321, + "step": 10037 + }, + { + "epoch": 0.8319933692498964, + "grad_norm": 0.24957750737667084, + "learning_rate": 8.404475756319934e-07, + "loss": 0.4981, + "step": 10038 + }, + { + "epoch": 0.8320762536261914, + "grad_norm": 0.2811509370803833, + "learning_rate": 8.400331537505181e-07, + "loss": 0.5914, + "step": 10039 + }, + { + "epoch": 0.8321591380024865, + "grad_norm": 0.22278086841106415, + "learning_rate": 8.396187318690427e-07, + "loss": 0.486, + "step": 10040 + }, + { + "epoch": 0.8322420223787816, + "grad_norm": 0.2469116896390915, + "learning_rate": 8.392043099875674e-07, + "loss": 0.4958, + "step": 10041 + }, + { + "epoch": 0.8323249067550766, + "grad_norm": 0.2409919798374176, + "learning_rate": 8.387898881060921e-07, + "loss": 0.5118, + "step": 10042 + }, + { + "epoch": 0.8324077911313718, + "grad_norm": 0.2470969557762146, + "learning_rate": 8.383754662246168e-07, + "loss": 0.5157, + "step": 10043 + }, + { + "epoch": 0.8324906755076668, + "grad_norm": 0.2455275058746338, + "learning_rate": 8.379610443431414e-07, + "loss": 0.4893, + "step": 10044 + }, + { + "epoch": 0.8325735598839619, + "grad_norm": 0.25625893473625183, + "learning_rate": 8.37546622461666e-07, + "loss": 0.5015, + "step": 10045 + }, + { + "epoch": 0.832656444260257, + "grad_norm": 0.24286773800849915, + "learning_rate": 8.371322005801907e-07, + "loss": 0.5049, + "step": 10046 + }, + { + "epoch": 0.832739328636552, + "grad_norm": 0.23499144613742828, + "learning_rate": 8.367177786987153e-07, + "loss": 0.4521, + "step": 10047 + }, + { + "epoch": 0.8328222130128471, + "grad_norm": 0.23404765129089355, + "learning_rate": 8.363033568172401e-07, + "loss": 0.46, + "step": 10048 + }, + { + "epoch": 0.8329050973891422, + "grad_norm": 0.22517666220664978, + "learning_rate": 8.358889349357646e-07, + "loss": 0.4773, + "step": 10049 + }, + { + "epoch": 0.8329879817654372, + "grad_norm": 0.2430327832698822, + "learning_rate": 8.354745130542894e-07, + "loss": 0.4889, + "step": 10050 + }, + { + "epoch": 0.8330708661417323, + "grad_norm": 0.23013073205947876, + "learning_rate": 8.35060091172814e-07, + "loss": 0.4306, + "step": 10051 + }, + { + "epoch": 0.8331537505180273, + "grad_norm": 0.22763602435588837, + "learning_rate": 8.346456692913387e-07, + "loss": 0.4564, + "step": 10052 + }, + { + "epoch": 0.8332366348943224, + "grad_norm": 0.23527641594409943, + "learning_rate": 8.342312474098634e-07, + "loss": 0.5105, + "step": 10053 + }, + { + "epoch": 0.8333195192706175, + "grad_norm": 0.27043259143829346, + "learning_rate": 8.33816825528388e-07, + "loss": 0.4832, + "step": 10054 + }, + { + "epoch": 0.8334024036469125, + "grad_norm": 0.23337335884571075, + "learning_rate": 8.334024036469127e-07, + "loss": 0.494, + "step": 10055 + }, + { + "epoch": 0.8334852880232076, + "grad_norm": 0.2352861613035202, + "learning_rate": 8.329879817654372e-07, + "loss": 0.4924, + "step": 10056 + }, + { + "epoch": 0.8335681723995026, + "grad_norm": 0.25841137766838074, + "learning_rate": 8.32573559883962e-07, + "loss": 0.5093, + "step": 10057 + }, + { + "epoch": 0.8336510567757978, + "grad_norm": 0.23645806312561035, + "learning_rate": 8.321591380024865e-07, + "loss": 0.4784, + "step": 10058 + }, + { + "epoch": 0.8337339411520929, + "grad_norm": 0.24664905667304993, + "learning_rate": 8.317447161210113e-07, + "loss": 0.4827, + "step": 10059 + }, + { + "epoch": 0.8338168255283879, + "grad_norm": 0.2345493882894516, + "learning_rate": 8.31330294239536e-07, + "loss": 0.5144, + "step": 10060 + }, + { + "epoch": 0.833899709904683, + "grad_norm": 0.2537612020969391, + "learning_rate": 8.309158723580606e-07, + "loss": 0.5219, + "step": 10061 + }, + { + "epoch": 0.833982594280978, + "grad_norm": 0.2459963858127594, + "learning_rate": 8.305014504765853e-07, + "loss": 0.4846, + "step": 10062 + }, + { + "epoch": 0.8340654786572731, + "grad_norm": 0.27045249938964844, + "learning_rate": 8.300870285951098e-07, + "loss": 0.4901, + "step": 10063 + }, + { + "epoch": 0.8341483630335682, + "grad_norm": 0.2450253963470459, + "learning_rate": 8.296726067136346e-07, + "loss": 0.4583, + "step": 10064 + }, + { + "epoch": 0.8342312474098632, + "grad_norm": 0.23722976446151733, + "learning_rate": 8.292581848321591e-07, + "loss": 0.4864, + "step": 10065 + }, + { + "epoch": 0.8343141317861583, + "grad_norm": 0.2450307458639145, + "learning_rate": 8.288437629506839e-07, + "loss": 0.4842, + "step": 10066 + }, + { + "epoch": 0.8343970161624534, + "grad_norm": 0.25491341948509216, + "learning_rate": 8.284293410692084e-07, + "loss": 0.5154, + "step": 10067 + }, + { + "epoch": 0.8344799005387484, + "grad_norm": 0.23091356456279755, + "learning_rate": 8.280149191877332e-07, + "loss": 0.4498, + "step": 10068 + }, + { + "epoch": 0.8345627849150435, + "grad_norm": 0.24259191751480103, + "learning_rate": 8.276004973062578e-07, + "loss": 0.4977, + "step": 10069 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 0.24161656200885773, + "learning_rate": 8.271860754247825e-07, + "loss": 0.5044, + "step": 10070 + }, + { + "epoch": 0.8347285536676337, + "grad_norm": 0.23383526504039764, + "learning_rate": 8.267716535433072e-07, + "loss": 0.4931, + "step": 10071 + }, + { + "epoch": 0.8348114380439288, + "grad_norm": 0.234603151679039, + "learning_rate": 8.263572316618317e-07, + "loss": 0.4669, + "step": 10072 + }, + { + "epoch": 0.8348943224202238, + "grad_norm": 0.23448331654071808, + "learning_rate": 8.259428097803565e-07, + "loss": 0.452, + "step": 10073 + }, + { + "epoch": 0.8349772067965189, + "grad_norm": 0.24363656342029572, + "learning_rate": 8.255283878988811e-07, + "loss": 0.5297, + "step": 10074 + }, + { + "epoch": 0.8350600911728139, + "grad_norm": 0.26202213764190674, + "learning_rate": 8.251139660174058e-07, + "loss": 0.5144, + "step": 10075 + }, + { + "epoch": 0.835142975549109, + "grad_norm": 0.24428711831569672, + "learning_rate": 8.246995441359304e-07, + "loss": 0.4992, + "step": 10076 + }, + { + "epoch": 0.8352258599254041, + "grad_norm": 0.242131769657135, + "learning_rate": 8.242851222544551e-07, + "loss": 0.4526, + "step": 10077 + }, + { + "epoch": 0.8353087443016991, + "grad_norm": 0.24355696141719818, + "learning_rate": 8.238707003729797e-07, + "loss": 0.5259, + "step": 10078 + }, + { + "epoch": 0.8353916286779942, + "grad_norm": 0.24709680676460266, + "learning_rate": 8.234562784915045e-07, + "loss": 0.5084, + "step": 10079 + }, + { + "epoch": 0.8354745130542892, + "grad_norm": 0.26125505566596985, + "learning_rate": 8.230418566100291e-07, + "loss": 0.4899, + "step": 10080 + }, + { + "epoch": 0.8355573974305843, + "grad_norm": 0.23250509798526764, + "learning_rate": 8.226274347285537e-07, + "loss": 0.5163, + "step": 10081 + }, + { + "epoch": 0.8356402818068794, + "grad_norm": 0.2513051927089691, + "learning_rate": 8.222130128470784e-07, + "loss": 0.5205, + "step": 10082 + }, + { + "epoch": 0.8357231661831744, + "grad_norm": 0.2671016752719879, + "learning_rate": 8.21798590965603e-07, + "loss": 0.5061, + "step": 10083 + }, + { + "epoch": 0.8358060505594695, + "grad_norm": 0.24556753039360046, + "learning_rate": 8.213841690841278e-07, + "loss": 0.4815, + "step": 10084 + }, + { + "epoch": 0.8358889349357647, + "grad_norm": 0.22824516892433167, + "learning_rate": 8.209697472026523e-07, + "loss": 0.4659, + "step": 10085 + }, + { + "epoch": 0.8359718193120597, + "grad_norm": 0.22653315961360931, + "learning_rate": 8.205553253211771e-07, + "loss": 0.4729, + "step": 10086 + }, + { + "epoch": 0.8360547036883548, + "grad_norm": 0.23377732932567596, + "learning_rate": 8.201409034397016e-07, + "loss": 0.5267, + "step": 10087 + }, + { + "epoch": 0.8361375880646498, + "grad_norm": 0.2320941984653473, + "learning_rate": 8.197264815582264e-07, + "loss": 0.4877, + "step": 10088 + }, + { + "epoch": 0.8362204724409449, + "grad_norm": 0.23853157460689545, + "learning_rate": 8.193120596767511e-07, + "loss": 0.4889, + "step": 10089 + }, + { + "epoch": 0.83630335681724, + "grad_norm": 0.22935272753238678, + "learning_rate": 8.188976377952756e-07, + "loss": 0.4868, + "step": 10090 + }, + { + "epoch": 0.836386241193535, + "grad_norm": 0.24446089565753937, + "learning_rate": 8.184832159138004e-07, + "loss": 0.4983, + "step": 10091 + }, + { + "epoch": 0.8364691255698301, + "grad_norm": 0.25647053122520447, + "learning_rate": 8.180687940323249e-07, + "loss": 0.5381, + "step": 10092 + }, + { + "epoch": 0.8365520099461251, + "grad_norm": 0.2470097839832306, + "learning_rate": 8.176543721508497e-07, + "loss": 0.5402, + "step": 10093 + }, + { + "epoch": 0.8366348943224202, + "grad_norm": 0.23388756811618805, + "learning_rate": 8.172399502693742e-07, + "loss": 0.4847, + "step": 10094 + }, + { + "epoch": 0.8367177786987153, + "grad_norm": 0.2343408763408661, + "learning_rate": 8.16825528387899e-07, + "loss": 0.5131, + "step": 10095 + }, + { + "epoch": 0.8368006630750103, + "grad_norm": 0.2615944445133209, + "learning_rate": 8.164111065064235e-07, + "loss": 0.4503, + "step": 10096 + }, + { + "epoch": 0.8368835474513054, + "grad_norm": 0.2417733520269394, + "learning_rate": 8.159966846249483e-07, + "loss": 0.4781, + "step": 10097 + }, + { + "epoch": 0.8369664318276004, + "grad_norm": 0.2559216320514679, + "learning_rate": 8.15582262743473e-07, + "loss": 0.5031, + "step": 10098 + }, + { + "epoch": 0.8370493162038956, + "grad_norm": 0.2470269650220871, + "learning_rate": 8.151678408619975e-07, + "loss": 0.4952, + "step": 10099 + }, + { + "epoch": 0.8371322005801907, + "grad_norm": 0.2469528466463089, + "learning_rate": 8.147534189805223e-07, + "loss": 0.5261, + "step": 10100 + }, + { + "epoch": 0.8372150849564857, + "grad_norm": 0.23986677825450897, + "learning_rate": 8.143389970990468e-07, + "loss": 0.5417, + "step": 10101 + }, + { + "epoch": 0.8372979693327808, + "grad_norm": 0.234523743391037, + "learning_rate": 8.139245752175716e-07, + "loss": 0.5016, + "step": 10102 + }, + { + "epoch": 0.8373808537090759, + "grad_norm": 0.2211342602968216, + "learning_rate": 8.135101533360962e-07, + "loss": 0.4942, + "step": 10103 + }, + { + "epoch": 0.8374637380853709, + "grad_norm": 0.24028488993644714, + "learning_rate": 8.130957314546209e-07, + "loss": 0.5051, + "step": 10104 + }, + { + "epoch": 0.837546622461666, + "grad_norm": 0.22416077554225922, + "learning_rate": 8.126813095731455e-07, + "loss": 0.4689, + "step": 10105 + }, + { + "epoch": 0.837629506837961, + "grad_norm": 0.23212140798568726, + "learning_rate": 8.122668876916702e-07, + "loss": 0.5004, + "step": 10106 + }, + { + "epoch": 0.8377123912142561, + "grad_norm": 0.2558324933052063, + "learning_rate": 8.118524658101948e-07, + "loss": 0.535, + "step": 10107 + }, + { + "epoch": 0.8377952755905512, + "grad_norm": 0.23076669871807098, + "learning_rate": 8.114380439287195e-07, + "loss": 0.4883, + "step": 10108 + }, + { + "epoch": 0.8378781599668462, + "grad_norm": 0.24213409423828125, + "learning_rate": 8.110236220472442e-07, + "loss": 0.529, + "step": 10109 + }, + { + "epoch": 0.8379610443431413, + "grad_norm": 0.23881028592586517, + "learning_rate": 8.106092001657688e-07, + "loss": 0.5095, + "step": 10110 + }, + { + "epoch": 0.8380439287194363, + "grad_norm": 0.23412708938121796, + "learning_rate": 8.101947782842935e-07, + "loss": 0.4777, + "step": 10111 + }, + { + "epoch": 0.8381268130957314, + "grad_norm": 0.23676276206970215, + "learning_rate": 8.097803564028181e-07, + "loss": 0.4759, + "step": 10112 + }, + { + "epoch": 0.8382096974720266, + "grad_norm": 0.22943055629730225, + "learning_rate": 8.093659345213429e-07, + "loss": 0.4988, + "step": 10113 + }, + { + "epoch": 0.8382925818483216, + "grad_norm": 0.2313789576292038, + "learning_rate": 8.089515126398674e-07, + "loss": 0.4933, + "step": 10114 + }, + { + "epoch": 0.8383754662246167, + "grad_norm": 0.2585154175758362, + "learning_rate": 8.085370907583922e-07, + "loss": 0.5173, + "step": 10115 + }, + { + "epoch": 0.8384583506009117, + "grad_norm": 0.2424432635307312, + "learning_rate": 8.081226688769167e-07, + "loss": 0.5349, + "step": 10116 + }, + { + "epoch": 0.8385412349772068, + "grad_norm": 0.2372218817472458, + "learning_rate": 8.077082469954414e-07, + "loss": 0.4963, + "step": 10117 + }, + { + "epoch": 0.8386241193535019, + "grad_norm": 0.2439260631799698, + "learning_rate": 8.072938251139662e-07, + "loss": 0.4897, + "step": 10118 + }, + { + "epoch": 0.8387070037297969, + "grad_norm": 0.24415986239910126, + "learning_rate": 8.068794032324907e-07, + "loss": 0.5277, + "step": 10119 + }, + { + "epoch": 0.838789888106092, + "grad_norm": 0.242290660738945, + "learning_rate": 8.064649813510155e-07, + "loss": 0.4759, + "step": 10120 + }, + { + "epoch": 0.8388727724823871, + "grad_norm": 0.230577751994133, + "learning_rate": 8.0605055946954e-07, + "loss": 0.4453, + "step": 10121 + }, + { + "epoch": 0.8389556568586821, + "grad_norm": 0.25632116198539734, + "learning_rate": 8.056361375880648e-07, + "loss": 0.5127, + "step": 10122 + }, + { + "epoch": 0.8390385412349772, + "grad_norm": 0.2262369692325592, + "learning_rate": 8.052217157065893e-07, + "loss": 0.4782, + "step": 10123 + }, + { + "epoch": 0.8391214256112722, + "grad_norm": 0.2593905031681061, + "learning_rate": 8.048072938251141e-07, + "loss": 0.5329, + "step": 10124 + }, + { + "epoch": 0.8392043099875673, + "grad_norm": 0.2620788514614105, + "learning_rate": 8.043928719436386e-07, + "loss": 0.5175, + "step": 10125 + }, + { + "epoch": 0.8392871943638625, + "grad_norm": 0.24085195362567902, + "learning_rate": 8.039784500621633e-07, + "loss": 0.5101, + "step": 10126 + }, + { + "epoch": 0.8393700787401575, + "grad_norm": 0.24159058928489685, + "learning_rate": 8.035640281806881e-07, + "loss": 0.511, + "step": 10127 + }, + { + "epoch": 0.8394529631164526, + "grad_norm": 0.2492523193359375, + "learning_rate": 8.031496062992126e-07, + "loss": 0.5267, + "step": 10128 + }, + { + "epoch": 0.8395358474927476, + "grad_norm": 0.24204450845718384, + "learning_rate": 8.027351844177374e-07, + "loss": 0.5201, + "step": 10129 + }, + { + "epoch": 0.8396187318690427, + "grad_norm": 0.24423366785049438, + "learning_rate": 8.023207625362619e-07, + "loss": 0.5128, + "step": 10130 + }, + { + "epoch": 0.8397016162453378, + "grad_norm": 0.23374147713184357, + "learning_rate": 8.019063406547867e-07, + "loss": 0.5267, + "step": 10131 + }, + { + "epoch": 0.8397845006216328, + "grad_norm": 0.2372070848941803, + "learning_rate": 8.014919187733112e-07, + "loss": 0.5416, + "step": 10132 + }, + { + "epoch": 0.8398673849979279, + "grad_norm": 0.2500975728034973, + "learning_rate": 8.01077496891836e-07, + "loss": 0.5599, + "step": 10133 + }, + { + "epoch": 0.839950269374223, + "grad_norm": 0.2627750337123871, + "learning_rate": 8.006630750103606e-07, + "loss": 0.5378, + "step": 10134 + }, + { + "epoch": 0.840033153750518, + "grad_norm": 0.25872737169265747, + "learning_rate": 8.002486531288852e-07, + "loss": 0.4968, + "step": 10135 + }, + { + "epoch": 0.8401160381268131, + "grad_norm": 0.2314119189977646, + "learning_rate": 7.9983423124741e-07, + "loss": 0.4894, + "step": 10136 + }, + { + "epoch": 0.8401989225031081, + "grad_norm": 0.2621421813964844, + "learning_rate": 7.994198093659345e-07, + "loss": 0.5598, + "step": 10137 + }, + { + "epoch": 0.8402818068794032, + "grad_norm": 0.2847527265548706, + "learning_rate": 7.990053874844593e-07, + "loss": 0.505, + "step": 10138 + }, + { + "epoch": 0.8403646912556983, + "grad_norm": 0.23815031349658966, + "learning_rate": 7.985909656029839e-07, + "loss": 0.4941, + "step": 10139 + }, + { + "epoch": 0.8404475756319933, + "grad_norm": 0.2343861311674118, + "learning_rate": 7.981765437215086e-07, + "loss": 0.4846, + "step": 10140 + }, + { + "epoch": 0.8405304600082885, + "grad_norm": 0.2515202760696411, + "learning_rate": 7.977621218400332e-07, + "loss": 0.5415, + "step": 10141 + }, + { + "epoch": 0.8406133443845835, + "grad_norm": 0.23115664720535278, + "learning_rate": 7.973476999585579e-07, + "loss": 0.495, + "step": 10142 + }, + { + "epoch": 0.8406962287608786, + "grad_norm": 0.24662677943706512, + "learning_rate": 7.969332780770825e-07, + "loss": 0.519, + "step": 10143 + }, + { + "epoch": 0.8407791131371737, + "grad_norm": 0.23910662531852722, + "learning_rate": 7.965188561956072e-07, + "loss": 0.494, + "step": 10144 + }, + { + "epoch": 0.8408619975134687, + "grad_norm": 0.22588324546813965, + "learning_rate": 7.961044343141318e-07, + "loss": 0.444, + "step": 10145 + }, + { + "epoch": 0.8409448818897638, + "grad_norm": 0.2563161849975586, + "learning_rate": 7.956900124326565e-07, + "loss": 0.4801, + "step": 10146 + }, + { + "epoch": 0.8410277662660588, + "grad_norm": 0.221974715590477, + "learning_rate": 7.952755905511812e-07, + "loss": 0.5064, + "step": 10147 + }, + { + "epoch": 0.8411106506423539, + "grad_norm": 0.23390412330627441, + "learning_rate": 7.948611686697058e-07, + "loss": 0.4553, + "step": 10148 + }, + { + "epoch": 0.841193535018649, + "grad_norm": 0.2294778823852539, + "learning_rate": 7.944467467882306e-07, + "loss": 0.4817, + "step": 10149 + }, + { + "epoch": 0.841276419394944, + "grad_norm": 0.2671568691730499, + "learning_rate": 7.940323249067551e-07, + "loss": 0.5181, + "step": 10150 + }, + { + "epoch": 0.8413593037712391, + "grad_norm": 0.24713373184204102, + "learning_rate": 7.936179030252799e-07, + "loss": 0.4691, + "step": 10151 + }, + { + "epoch": 0.8414421881475342, + "grad_norm": 0.24485941231250763, + "learning_rate": 7.932034811438044e-07, + "loss": 0.4601, + "step": 10152 + }, + { + "epoch": 0.8415250725238292, + "grad_norm": 0.2356433868408203, + "learning_rate": 7.927890592623291e-07, + "loss": 0.4487, + "step": 10153 + }, + { + "epoch": 0.8416079569001244, + "grad_norm": 0.26460474729537964, + "learning_rate": 7.923746373808537e-07, + "loss": 0.5591, + "step": 10154 + }, + { + "epoch": 0.8416908412764194, + "grad_norm": 0.2491329461336136, + "learning_rate": 7.919602154993784e-07, + "loss": 0.5396, + "step": 10155 + }, + { + "epoch": 0.8417737256527145, + "grad_norm": 0.24197269976139069, + "learning_rate": 7.915457936179032e-07, + "loss": 0.4897, + "step": 10156 + }, + { + "epoch": 0.8418566100290096, + "grad_norm": 0.2524791359901428, + "learning_rate": 7.911313717364277e-07, + "loss": 0.4757, + "step": 10157 + }, + { + "epoch": 0.8419394944053046, + "grad_norm": 0.25136780738830566, + "learning_rate": 7.907169498549525e-07, + "loss": 0.503, + "step": 10158 + }, + { + "epoch": 0.8420223787815997, + "grad_norm": 0.25175532698631287, + "learning_rate": 7.90302527973477e-07, + "loss": 0.493, + "step": 10159 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.2625806927680969, + "learning_rate": 7.898881060920018e-07, + "loss": 0.495, + "step": 10160 + }, + { + "epoch": 0.8421881475341898, + "grad_norm": 0.22156263887882233, + "learning_rate": 7.894736842105263e-07, + "loss": 0.4818, + "step": 10161 + }, + { + "epoch": 0.8422710319104849, + "grad_norm": 0.23555529117584229, + "learning_rate": 7.89059262329051e-07, + "loss": 0.4454, + "step": 10162 + }, + { + "epoch": 0.8423539162867799, + "grad_norm": 0.2506594657897949, + "learning_rate": 7.886448404475757e-07, + "loss": 0.4982, + "step": 10163 + }, + { + "epoch": 0.842436800663075, + "grad_norm": 0.22705425322055817, + "learning_rate": 7.882304185661003e-07, + "loss": 0.4488, + "step": 10164 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 0.24070727825164795, + "learning_rate": 7.878159966846251e-07, + "loss": 0.5264, + "step": 10165 + }, + { + "epoch": 0.8426025694156651, + "grad_norm": 0.22292786836624146, + "learning_rate": 7.874015748031496e-07, + "loss": 0.4506, + "step": 10166 + }, + { + "epoch": 0.8426854537919602, + "grad_norm": 0.2503713369369507, + "learning_rate": 7.869871529216744e-07, + "loss": 0.4962, + "step": 10167 + }, + { + "epoch": 0.8427683381682552, + "grad_norm": 0.23515407741069794, + "learning_rate": 7.86572731040199e-07, + "loss": 0.4984, + "step": 10168 + }, + { + "epoch": 0.8428512225445504, + "grad_norm": 0.24776782095432281, + "learning_rate": 7.861583091587237e-07, + "loss": 0.5282, + "step": 10169 + }, + { + "epoch": 0.8429341069208455, + "grad_norm": 0.25283724069595337, + "learning_rate": 7.857438872772483e-07, + "loss": 0.4672, + "step": 10170 + }, + { + "epoch": 0.8430169912971405, + "grad_norm": 0.23361128568649292, + "learning_rate": 7.853294653957729e-07, + "loss": 0.4783, + "step": 10171 + }, + { + "epoch": 0.8430998756734356, + "grad_norm": 0.24068455398082733, + "learning_rate": 7.849150435142976e-07, + "loss": 0.5085, + "step": 10172 + }, + { + "epoch": 0.8431827600497306, + "grad_norm": 0.24443770945072174, + "learning_rate": 7.845006216328223e-07, + "loss": 0.4766, + "step": 10173 + }, + { + "epoch": 0.8432656444260257, + "grad_norm": 0.2866753041744232, + "learning_rate": 7.840861997513469e-07, + "loss": 0.4706, + "step": 10174 + }, + { + "epoch": 0.8433485288023208, + "grad_norm": 0.259777307510376, + "learning_rate": 7.836717778698716e-07, + "loss": 0.5386, + "step": 10175 + }, + { + "epoch": 0.8434314131786158, + "grad_norm": 0.22292403876781464, + "learning_rate": 7.832573559883963e-07, + "loss": 0.4769, + "step": 10176 + }, + { + "epoch": 0.8435142975549109, + "grad_norm": 0.2235766053199768, + "learning_rate": 7.828429341069209e-07, + "loss": 0.4697, + "step": 10177 + }, + { + "epoch": 0.8435971819312059, + "grad_norm": 0.24934591352939606, + "learning_rate": 7.824285122254457e-07, + "loss": 0.4744, + "step": 10178 + }, + { + "epoch": 0.843680066307501, + "grad_norm": 0.28421175479888916, + "learning_rate": 7.820140903439702e-07, + "loss": 0.4921, + "step": 10179 + }, + { + "epoch": 0.8437629506837961, + "grad_norm": 0.2604561448097229, + "learning_rate": 7.815996684624949e-07, + "loss": 0.527, + "step": 10180 + }, + { + "epoch": 0.8438458350600911, + "grad_norm": 0.2743438184261322, + "learning_rate": 7.811852465810195e-07, + "loss": 0.5123, + "step": 10181 + }, + { + "epoch": 0.8439287194363863, + "grad_norm": 0.24493473768234253, + "learning_rate": 7.807708246995442e-07, + "loss": 0.495, + "step": 10182 + }, + { + "epoch": 0.8440116038126813, + "grad_norm": 0.24135015904903412, + "learning_rate": 7.803564028180688e-07, + "loss": 0.4895, + "step": 10183 + }, + { + "epoch": 0.8440944881889764, + "grad_norm": 0.24622102081775665, + "learning_rate": 7.799419809365935e-07, + "loss": 0.477, + "step": 10184 + }, + { + "epoch": 0.8441773725652715, + "grad_norm": 0.26968783140182495, + "learning_rate": 7.795275590551183e-07, + "loss": 0.5572, + "step": 10185 + }, + { + "epoch": 0.8442602569415665, + "grad_norm": 0.24333328008651733, + "learning_rate": 7.791131371736428e-07, + "loss": 0.4871, + "step": 10186 + }, + { + "epoch": 0.8443431413178616, + "grad_norm": 0.22169114649295807, + "learning_rate": 7.786987152921676e-07, + "loss": 0.517, + "step": 10187 + }, + { + "epoch": 0.8444260256941567, + "grad_norm": 0.23010756075382233, + "learning_rate": 7.782842934106921e-07, + "loss": 0.4714, + "step": 10188 + }, + { + "epoch": 0.8445089100704517, + "grad_norm": 0.2453446388244629, + "learning_rate": 7.778698715292168e-07, + "loss": 0.4926, + "step": 10189 + }, + { + "epoch": 0.8445917944467468, + "grad_norm": 0.23541627824306488, + "learning_rate": 7.774554496477414e-07, + "loss": 0.4594, + "step": 10190 + }, + { + "epoch": 0.8446746788230418, + "grad_norm": 0.25566592812538147, + "learning_rate": 7.770410277662661e-07, + "loss": 0.5599, + "step": 10191 + }, + { + "epoch": 0.8447575631993369, + "grad_norm": 0.2428893893957138, + "learning_rate": 7.766266058847908e-07, + "loss": 0.4724, + "step": 10192 + }, + { + "epoch": 0.844840447575632, + "grad_norm": 0.24328738451004028, + "learning_rate": 7.762121840033154e-07, + "loss": 0.5148, + "step": 10193 + }, + { + "epoch": 0.844923331951927, + "grad_norm": 0.2388807237148285, + "learning_rate": 7.757977621218402e-07, + "loss": 0.4788, + "step": 10194 + }, + { + "epoch": 0.8450062163282221, + "grad_norm": 0.24679581820964813, + "learning_rate": 7.753833402403647e-07, + "loss": 0.504, + "step": 10195 + }, + { + "epoch": 0.8450891007045171, + "grad_norm": 0.23309776186943054, + "learning_rate": 7.749689183588895e-07, + "loss": 0.4517, + "step": 10196 + }, + { + "epoch": 0.8451719850808123, + "grad_norm": 0.25306060910224915, + "learning_rate": 7.745544964774141e-07, + "loss": 0.5204, + "step": 10197 + }, + { + "epoch": 0.8452548694571074, + "grad_norm": 0.2389984130859375, + "learning_rate": 7.741400745959387e-07, + "loss": 0.4768, + "step": 10198 + }, + { + "epoch": 0.8453377538334024, + "grad_norm": 0.22327910363674164, + "learning_rate": 7.737256527144634e-07, + "loss": 0.5046, + "step": 10199 + }, + { + "epoch": 0.8454206382096975, + "grad_norm": 0.22932758927345276, + "learning_rate": 7.73311230832988e-07, + "loss": 0.4691, + "step": 10200 + }, + { + "epoch": 0.8455035225859926, + "grad_norm": 0.24751418828964233, + "learning_rate": 7.728968089515127e-07, + "loss": 0.485, + "step": 10201 + }, + { + "epoch": 0.8455864069622876, + "grad_norm": 0.22842128574848175, + "learning_rate": 7.724823870700373e-07, + "loss": 0.4736, + "step": 10202 + }, + { + "epoch": 0.8456692913385827, + "grad_norm": 0.23811180889606476, + "learning_rate": 7.720679651885621e-07, + "loss": 0.506, + "step": 10203 + }, + { + "epoch": 0.8457521757148777, + "grad_norm": 0.23406483232975006, + "learning_rate": 7.716535433070867e-07, + "loss": 0.4812, + "step": 10204 + }, + { + "epoch": 0.8458350600911728, + "grad_norm": 0.2342226356267929, + "learning_rate": 7.712391214256113e-07, + "loss": 0.4875, + "step": 10205 + }, + { + "epoch": 0.8459179444674679, + "grad_norm": 0.2362060695886612, + "learning_rate": 7.70824699544136e-07, + "loss": 0.4795, + "step": 10206 + }, + { + "epoch": 0.8460008288437629, + "grad_norm": 0.2238895297050476, + "learning_rate": 7.704102776626606e-07, + "loss": 0.464, + "step": 10207 + }, + { + "epoch": 0.846083713220058, + "grad_norm": 0.23910756409168243, + "learning_rate": 7.699958557811853e-07, + "loss": 0.4813, + "step": 10208 + }, + { + "epoch": 0.846166597596353, + "grad_norm": 0.25002291798591614, + "learning_rate": 7.6958143389971e-07, + "loss": 0.476, + "step": 10209 + }, + { + "epoch": 0.8462494819726482, + "grad_norm": 0.2399374097585678, + "learning_rate": 7.691670120182346e-07, + "loss": 0.5255, + "step": 10210 + }, + { + "epoch": 0.8463323663489433, + "grad_norm": 0.21112670004367828, + "learning_rate": 7.687525901367593e-07, + "loss": 0.4514, + "step": 10211 + }, + { + "epoch": 0.8464152507252383, + "grad_norm": 0.266826331615448, + "learning_rate": 7.683381682552839e-07, + "loss": 0.5137, + "step": 10212 + }, + { + "epoch": 0.8464981351015334, + "grad_norm": 0.24070951342582703, + "learning_rate": 7.679237463738086e-07, + "loss": 0.4917, + "step": 10213 + }, + { + "epoch": 0.8465810194778284, + "grad_norm": 0.237960085272789, + "learning_rate": 7.675093244923334e-07, + "loss": 0.4781, + "step": 10214 + }, + { + "epoch": 0.8466639038541235, + "grad_norm": 0.24628598988056183, + "learning_rate": 7.670949026108579e-07, + "loss": 0.4817, + "step": 10215 + }, + { + "epoch": 0.8467467882304186, + "grad_norm": 0.2843930721282959, + "learning_rate": 7.666804807293826e-07, + "loss": 0.5184, + "step": 10216 + }, + { + "epoch": 0.8468296726067136, + "grad_norm": 0.24460619688034058, + "learning_rate": 7.662660588479072e-07, + "loss": 0.5101, + "step": 10217 + }, + { + "epoch": 0.8469125569830087, + "grad_norm": 0.2363709956407547, + "learning_rate": 7.658516369664319e-07, + "loss": 0.4782, + "step": 10218 + }, + { + "epoch": 0.8469954413593038, + "grad_norm": 0.24478328227996826, + "learning_rate": 7.654372150849565e-07, + "loss": 0.5005, + "step": 10219 + }, + { + "epoch": 0.8470783257355988, + "grad_norm": 0.24337463080883026, + "learning_rate": 7.650227932034812e-07, + "loss": 0.5032, + "step": 10220 + }, + { + "epoch": 0.8471612101118939, + "grad_norm": 0.23531699180603027, + "learning_rate": 7.646083713220058e-07, + "loss": 0.4749, + "step": 10221 + }, + { + "epoch": 0.8472440944881889, + "grad_norm": 0.2210611253976822, + "learning_rate": 7.641939494405305e-07, + "loss": 0.4709, + "step": 10222 + }, + { + "epoch": 0.847326978864484, + "grad_norm": 0.25200968980789185, + "learning_rate": 7.637795275590553e-07, + "loss": 0.5336, + "step": 10223 + }, + { + "epoch": 0.8474098632407792, + "grad_norm": 0.23492465913295746, + "learning_rate": 7.633651056775798e-07, + "loss": 0.4677, + "step": 10224 + }, + { + "epoch": 0.8474927476170742, + "grad_norm": 0.24301587045192719, + "learning_rate": 7.629506837961045e-07, + "loss": 0.5045, + "step": 10225 + }, + { + "epoch": 0.8475756319933693, + "grad_norm": 0.2397782951593399, + "learning_rate": 7.625362619146291e-07, + "loss": 0.4842, + "step": 10226 + }, + { + "epoch": 0.8476585163696643, + "grad_norm": 0.2351786196231842, + "learning_rate": 7.621218400331538e-07, + "loss": 0.4928, + "step": 10227 + }, + { + "epoch": 0.8477414007459594, + "grad_norm": 0.25371989607810974, + "learning_rate": 7.617074181516785e-07, + "loss": 0.4824, + "step": 10228 + }, + { + "epoch": 0.8478242851222545, + "grad_norm": 0.24256379902362823, + "learning_rate": 7.612929962702031e-07, + "loss": 0.4987, + "step": 10229 + }, + { + "epoch": 0.8479071694985495, + "grad_norm": 0.23335842788219452, + "learning_rate": 7.608785743887278e-07, + "loss": 0.4945, + "step": 10230 + }, + { + "epoch": 0.8479900538748446, + "grad_norm": 0.2375490516424179, + "learning_rate": 7.604641525072524e-07, + "loss": 0.5069, + "step": 10231 + }, + { + "epoch": 0.8480729382511396, + "grad_norm": 0.2385551631450653, + "learning_rate": 7.600497306257771e-07, + "loss": 0.5228, + "step": 10232 + }, + { + "epoch": 0.8481558226274347, + "grad_norm": 0.22750362753868103, + "learning_rate": 7.596353087443018e-07, + "loss": 0.4612, + "step": 10233 + }, + { + "epoch": 0.8482387070037298, + "grad_norm": 0.23502418398857117, + "learning_rate": 7.592208868628264e-07, + "loss": 0.4508, + "step": 10234 + }, + { + "epoch": 0.8483215913800248, + "grad_norm": 0.24024175107479095, + "learning_rate": 7.588064649813511e-07, + "loss": 0.4659, + "step": 10235 + }, + { + "epoch": 0.8484044757563199, + "grad_norm": 0.2265358865261078, + "learning_rate": 7.583920430998757e-07, + "loss": 0.4901, + "step": 10236 + }, + { + "epoch": 0.848487360132615, + "grad_norm": 0.2514173686504364, + "learning_rate": 7.579776212184004e-07, + "loss": 0.5054, + "step": 10237 + }, + { + "epoch": 0.84857024450891, + "grad_norm": 0.24235643446445465, + "learning_rate": 7.575631993369251e-07, + "loss": 0.4788, + "step": 10238 + }, + { + "epoch": 0.8486531288852052, + "grad_norm": 0.2494291067123413, + "learning_rate": 7.571487774554497e-07, + "loss": 0.501, + "step": 10239 + }, + { + "epoch": 0.8487360132615002, + "grad_norm": 0.2534453272819519, + "learning_rate": 7.567343555739744e-07, + "loss": 0.5012, + "step": 10240 + }, + { + "epoch": 0.8488188976377953, + "grad_norm": 0.2492733746767044, + "learning_rate": 7.56319933692499e-07, + "loss": 0.5355, + "step": 10241 + }, + { + "epoch": 0.8489017820140904, + "grad_norm": 0.22774836421012878, + "learning_rate": 7.559055118110237e-07, + "loss": 0.4948, + "step": 10242 + }, + { + "epoch": 0.8489846663903854, + "grad_norm": 0.28252774477005005, + "learning_rate": 7.554910899295484e-07, + "loss": 0.5565, + "step": 10243 + }, + { + "epoch": 0.8490675507666805, + "grad_norm": 0.23447680473327637, + "learning_rate": 7.55076668048073e-07, + "loss": 0.4816, + "step": 10244 + }, + { + "epoch": 0.8491504351429755, + "grad_norm": 0.24591998755931854, + "learning_rate": 7.546622461665977e-07, + "loss": 0.5004, + "step": 10245 + }, + { + "epoch": 0.8492333195192706, + "grad_norm": 0.22875338792800903, + "learning_rate": 7.542478242851223e-07, + "loss": 0.472, + "step": 10246 + }, + { + "epoch": 0.8493162038955657, + "grad_norm": 0.2394888997077942, + "learning_rate": 7.53833402403647e-07, + "loss": 0.5126, + "step": 10247 + }, + { + "epoch": 0.8493990882718607, + "grad_norm": 0.24785587191581726, + "learning_rate": 7.534189805221716e-07, + "loss": 0.4951, + "step": 10248 + }, + { + "epoch": 0.8494819726481558, + "grad_norm": 0.2450931817293167, + "learning_rate": 7.530045586406963e-07, + "loss": 0.4828, + "step": 10249 + }, + { + "epoch": 0.8495648570244508, + "grad_norm": 0.24782444536685944, + "learning_rate": 7.525901367592208e-07, + "loss": 0.5049, + "step": 10250 + }, + { + "epoch": 0.849647741400746, + "grad_norm": 0.24820558726787567, + "learning_rate": 7.521757148777456e-07, + "loss": 0.4992, + "step": 10251 + }, + { + "epoch": 0.8497306257770411, + "grad_norm": 0.22807583212852478, + "learning_rate": 7.517612929962703e-07, + "loss": 0.4766, + "step": 10252 + }, + { + "epoch": 0.8498135101533361, + "grad_norm": 0.22400085628032684, + "learning_rate": 7.513468711147949e-07, + "loss": 0.4789, + "step": 10253 + }, + { + "epoch": 0.8498963945296312, + "grad_norm": 0.23235014081001282, + "learning_rate": 7.509324492333196e-07, + "loss": 0.4994, + "step": 10254 + }, + { + "epoch": 0.8499792789059263, + "grad_norm": 0.2462155669927597, + "learning_rate": 7.505180273518442e-07, + "loss": 0.5058, + "step": 10255 + }, + { + "epoch": 0.8500621632822213, + "grad_norm": 0.2391049712896347, + "learning_rate": 7.501036054703689e-07, + "loss": 0.5468, + "step": 10256 + }, + { + "epoch": 0.8501450476585164, + "grad_norm": 0.23903873562812805, + "learning_rate": 7.496891835888936e-07, + "loss": 0.4841, + "step": 10257 + }, + { + "epoch": 0.8502279320348114, + "grad_norm": 0.23884131014347076, + "learning_rate": 7.492747617074182e-07, + "loss": 0.4876, + "step": 10258 + }, + { + "epoch": 0.8503108164111065, + "grad_norm": 0.24862734973430634, + "learning_rate": 7.488603398259428e-07, + "loss": 0.5333, + "step": 10259 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 0.2161952406167984, + "learning_rate": 7.484459179444675e-07, + "loss": 0.4611, + "step": 10260 + }, + { + "epoch": 0.8504765851636966, + "grad_norm": 0.2502710521221161, + "learning_rate": 7.480314960629922e-07, + "loss": 0.5306, + "step": 10261 + }, + { + "epoch": 0.8505594695399917, + "grad_norm": 0.24468404054641724, + "learning_rate": 7.476170741815169e-07, + "loss": 0.5074, + "step": 10262 + }, + { + "epoch": 0.8506423539162867, + "grad_norm": 0.24619796872138977, + "learning_rate": 7.472026523000415e-07, + "loss": 0.5387, + "step": 10263 + }, + { + "epoch": 0.8507252382925818, + "grad_norm": 0.22379039227962494, + "learning_rate": 7.467882304185662e-07, + "loss": 0.487, + "step": 10264 + }, + { + "epoch": 0.850808122668877, + "grad_norm": 0.24092887341976166, + "learning_rate": 7.463738085370908e-07, + "loss": 0.4539, + "step": 10265 + }, + { + "epoch": 0.850891007045172, + "grad_norm": 0.24239270389080048, + "learning_rate": 7.459593866556155e-07, + "loss": 0.4875, + "step": 10266 + }, + { + "epoch": 0.8509738914214671, + "grad_norm": 0.23193904757499695, + "learning_rate": 7.455449647741402e-07, + "loss": 0.4796, + "step": 10267 + }, + { + "epoch": 0.8510567757977622, + "grad_norm": 0.2605139911174774, + "learning_rate": 7.451305428926647e-07, + "loss": 0.5261, + "step": 10268 + }, + { + "epoch": 0.8511396601740572, + "grad_norm": 0.23296606540679932, + "learning_rate": 7.447161210111895e-07, + "loss": 0.4879, + "step": 10269 + }, + { + "epoch": 0.8512225445503523, + "grad_norm": 0.26499849557876587, + "learning_rate": 7.443016991297141e-07, + "loss": 0.5073, + "step": 10270 + }, + { + "epoch": 0.8513054289266473, + "grad_norm": 0.24981312453746796, + "learning_rate": 7.438872772482388e-07, + "loss": 0.4882, + "step": 10271 + }, + { + "epoch": 0.8513883133029424, + "grad_norm": 0.23697280883789062, + "learning_rate": 7.434728553667635e-07, + "loss": 0.4617, + "step": 10272 + }, + { + "epoch": 0.8514711976792375, + "grad_norm": 0.21207234263420105, + "learning_rate": 7.430584334852881e-07, + "loss": 0.4292, + "step": 10273 + }, + { + "epoch": 0.8515540820555325, + "grad_norm": 0.2366618812084198, + "learning_rate": 7.426440116038128e-07, + "loss": 0.4505, + "step": 10274 + }, + { + "epoch": 0.8516369664318276, + "grad_norm": 0.26723235845565796, + "learning_rate": 7.422295897223374e-07, + "loss": 0.5699, + "step": 10275 + }, + { + "epoch": 0.8517198508081226, + "grad_norm": 0.249552920460701, + "learning_rate": 7.418151678408621e-07, + "loss": 0.4911, + "step": 10276 + }, + { + "epoch": 0.8518027351844177, + "grad_norm": 0.2542005777359009, + "learning_rate": 7.414007459593866e-07, + "loss": 0.4945, + "step": 10277 + }, + { + "epoch": 0.8518856195607128, + "grad_norm": 0.2286694496870041, + "learning_rate": 7.409863240779114e-07, + "loss": 0.4656, + "step": 10278 + }, + { + "epoch": 0.8519685039370078, + "grad_norm": 0.25849151611328125, + "learning_rate": 7.405719021964361e-07, + "loss": 0.5158, + "step": 10279 + }, + { + "epoch": 0.852051388313303, + "grad_norm": 0.22923961281776428, + "learning_rate": 7.401574803149607e-07, + "loss": 0.4764, + "step": 10280 + }, + { + "epoch": 0.852134272689598, + "grad_norm": 0.24103963375091553, + "learning_rate": 7.397430584334854e-07, + "loss": 0.5023, + "step": 10281 + }, + { + "epoch": 0.8522171570658931, + "grad_norm": 0.25627610087394714, + "learning_rate": 7.3932863655201e-07, + "loss": 0.5245, + "step": 10282 + }, + { + "epoch": 0.8523000414421882, + "grad_norm": 0.2324981987476349, + "learning_rate": 7.389142146705347e-07, + "loss": 0.4712, + "step": 10283 + }, + { + "epoch": 0.8523829258184832, + "grad_norm": 0.2457270622253418, + "learning_rate": 7.384997927890592e-07, + "loss": 0.511, + "step": 10284 + }, + { + "epoch": 0.8524658101947783, + "grad_norm": 0.23013751208782196, + "learning_rate": 7.38085370907584e-07, + "loss": 0.4939, + "step": 10285 + }, + { + "epoch": 0.8525486945710734, + "grad_norm": 0.2314053773880005, + "learning_rate": 7.376709490261085e-07, + "loss": 0.4989, + "step": 10286 + }, + { + "epoch": 0.8526315789473684, + "grad_norm": 0.27006953954696655, + "learning_rate": 7.372565271446333e-07, + "loss": 0.5377, + "step": 10287 + }, + { + "epoch": 0.8527144633236635, + "grad_norm": 0.23315472900867462, + "learning_rate": 7.368421052631579e-07, + "loss": 0.5193, + "step": 10288 + }, + { + "epoch": 0.8527973476999585, + "grad_norm": 0.24342957139015198, + "learning_rate": 7.364276833816826e-07, + "loss": 0.4855, + "step": 10289 + }, + { + "epoch": 0.8528802320762536, + "grad_norm": 0.25753676891326904, + "learning_rate": 7.360132615002073e-07, + "loss": 0.5265, + "step": 10290 + }, + { + "epoch": 0.8529631164525487, + "grad_norm": 0.2313823401927948, + "learning_rate": 7.355988396187319e-07, + "loss": 0.4965, + "step": 10291 + }, + { + "epoch": 0.8530460008288437, + "grad_norm": 0.25406044721603394, + "learning_rate": 7.351844177372566e-07, + "loss": 0.5222, + "step": 10292 + }, + { + "epoch": 0.8531288852051389, + "grad_norm": 0.23764486610889435, + "learning_rate": 7.347699958557813e-07, + "loss": 0.4936, + "step": 10293 + }, + { + "epoch": 0.8532117695814339, + "grad_norm": 0.24005569517612457, + "learning_rate": 7.343555739743059e-07, + "loss": 0.4953, + "step": 10294 + }, + { + "epoch": 0.853294653957729, + "grad_norm": 0.23824508488178253, + "learning_rate": 7.339411520928305e-07, + "loss": 0.5046, + "step": 10295 + }, + { + "epoch": 0.8533775383340241, + "grad_norm": 0.23661229014396667, + "learning_rate": 7.335267302113552e-07, + "loss": 0.4714, + "step": 10296 + }, + { + "epoch": 0.8534604227103191, + "grad_norm": 0.23133811354637146, + "learning_rate": 7.331123083298798e-07, + "loss": 0.4848, + "step": 10297 + }, + { + "epoch": 0.8535433070866142, + "grad_norm": 0.23323944211006165, + "learning_rate": 7.326978864484046e-07, + "loss": 0.4584, + "step": 10298 + }, + { + "epoch": 0.8536261914629092, + "grad_norm": 0.231731116771698, + "learning_rate": 7.322834645669292e-07, + "loss": 0.4528, + "step": 10299 + }, + { + "epoch": 0.8537090758392043, + "grad_norm": 0.22438903152942657, + "learning_rate": 7.318690426854539e-07, + "loss": 0.4829, + "step": 10300 + }, + { + "epoch": 0.8537919602154994, + "grad_norm": 0.2393823117017746, + "learning_rate": 7.314546208039785e-07, + "loss": 0.5468, + "step": 10301 + }, + { + "epoch": 0.8538748445917944, + "grad_norm": 0.23480971157550812, + "learning_rate": 7.310401989225031e-07, + "loss": 0.4919, + "step": 10302 + }, + { + "epoch": 0.8539577289680895, + "grad_norm": 0.2438078671693802, + "learning_rate": 7.306257770410279e-07, + "loss": 0.4935, + "step": 10303 + }, + { + "epoch": 0.8540406133443846, + "grad_norm": 0.24427977204322815, + "learning_rate": 7.302113551595524e-07, + "loss": 0.5359, + "step": 10304 + }, + { + "epoch": 0.8541234977206796, + "grad_norm": 0.267578125, + "learning_rate": 7.297969332780772e-07, + "loss": 0.4977, + "step": 10305 + }, + { + "epoch": 0.8542063820969747, + "grad_norm": 0.22946268320083618, + "learning_rate": 7.293825113966017e-07, + "loss": 0.4544, + "step": 10306 + }, + { + "epoch": 0.8542892664732697, + "grad_norm": 0.2354222685098648, + "learning_rate": 7.289680895151265e-07, + "loss": 0.5042, + "step": 10307 + }, + { + "epoch": 0.8543721508495649, + "grad_norm": 0.2489498406648636, + "learning_rate": 7.285536676336512e-07, + "loss": 0.5067, + "step": 10308 + }, + { + "epoch": 0.85445503522586, + "grad_norm": 0.2371445596218109, + "learning_rate": 7.281392457521758e-07, + "loss": 0.4611, + "step": 10309 + }, + { + "epoch": 0.854537919602155, + "grad_norm": 0.2634475827217102, + "learning_rate": 7.277248238707005e-07, + "loss": 0.5504, + "step": 10310 + }, + { + "epoch": 0.8546208039784501, + "grad_norm": 0.2441987544298172, + "learning_rate": 7.27310401989225e-07, + "loss": 0.4808, + "step": 10311 + }, + { + "epoch": 0.8547036883547451, + "grad_norm": 0.23447704315185547, + "learning_rate": 7.268959801077498e-07, + "loss": 0.4729, + "step": 10312 + }, + { + "epoch": 0.8547865727310402, + "grad_norm": 0.24465283751487732, + "learning_rate": 7.264815582262743e-07, + "loss": 0.4528, + "step": 10313 + }, + { + "epoch": 0.8548694571073353, + "grad_norm": 0.225692480802536, + "learning_rate": 7.260671363447991e-07, + "loss": 0.445, + "step": 10314 + }, + { + "epoch": 0.8549523414836303, + "grad_norm": 0.23202499747276306, + "learning_rate": 7.256527144633236e-07, + "loss": 0.4998, + "step": 10315 + }, + { + "epoch": 0.8550352258599254, + "grad_norm": 0.254281222820282, + "learning_rate": 7.252382925818484e-07, + "loss": 0.5444, + "step": 10316 + }, + { + "epoch": 0.8551181102362204, + "grad_norm": 0.24952062964439392, + "learning_rate": 7.248238707003731e-07, + "loss": 0.4904, + "step": 10317 + }, + { + "epoch": 0.8552009946125155, + "grad_norm": 0.23256927728652954, + "learning_rate": 7.244094488188977e-07, + "loss": 0.5009, + "step": 10318 + }, + { + "epoch": 0.8552838789888106, + "grad_norm": 0.24827435612678528, + "learning_rate": 7.239950269374224e-07, + "loss": 0.5154, + "step": 10319 + }, + { + "epoch": 0.8553667633651056, + "grad_norm": 0.21789886057376862, + "learning_rate": 7.235806050559469e-07, + "loss": 0.4532, + "step": 10320 + }, + { + "epoch": 0.8554496477414008, + "grad_norm": 0.2551954984664917, + "learning_rate": 7.231661831744717e-07, + "loss": 0.5198, + "step": 10321 + }, + { + "epoch": 0.8555325321176959, + "grad_norm": 0.24333997070789337, + "learning_rate": 7.227517612929963e-07, + "loss": 0.4733, + "step": 10322 + }, + { + "epoch": 0.8556154164939909, + "grad_norm": 0.22016507387161255, + "learning_rate": 7.22337339411521e-07, + "loss": 0.465, + "step": 10323 + }, + { + "epoch": 0.855698300870286, + "grad_norm": 0.2439410537481308, + "learning_rate": 7.219229175300456e-07, + "loss": 0.5042, + "step": 10324 + }, + { + "epoch": 0.855781185246581, + "grad_norm": 0.2310163378715515, + "learning_rate": 7.215084956485703e-07, + "loss": 0.503, + "step": 10325 + }, + { + "epoch": 0.8558640696228761, + "grad_norm": 0.25032171607017517, + "learning_rate": 7.210940737670949e-07, + "loss": 0.5029, + "step": 10326 + }, + { + "epoch": 0.8559469539991712, + "grad_norm": 0.22677713632583618, + "learning_rate": 7.206796518856197e-07, + "loss": 0.471, + "step": 10327 + }, + { + "epoch": 0.8560298383754662, + "grad_norm": 0.23667623102664948, + "learning_rate": 7.202652300041443e-07, + "loss": 0.4291, + "step": 10328 + }, + { + "epoch": 0.8561127227517613, + "grad_norm": 0.23550140857696533, + "learning_rate": 7.19850808122669e-07, + "loss": 0.5002, + "step": 10329 + }, + { + "epoch": 0.8561956071280563, + "grad_norm": 0.2509869337081909, + "learning_rate": 7.194363862411936e-07, + "loss": 0.521, + "step": 10330 + }, + { + "epoch": 0.8562784915043514, + "grad_norm": 0.24491722881793976, + "learning_rate": 7.190219643597182e-07, + "loss": 0.5035, + "step": 10331 + }, + { + "epoch": 0.8563613758806465, + "grad_norm": 0.26505595445632935, + "learning_rate": 7.18607542478243e-07, + "loss": 0.5079, + "step": 10332 + }, + { + "epoch": 0.8564442602569415, + "grad_norm": 0.22993023693561554, + "learning_rate": 7.181931205967675e-07, + "loss": 0.4731, + "step": 10333 + }, + { + "epoch": 0.8565271446332366, + "grad_norm": 0.23025564849376678, + "learning_rate": 7.177786987152923e-07, + "loss": 0.4731, + "step": 10334 + }, + { + "epoch": 0.8566100290095316, + "grad_norm": 0.26918190717697144, + "learning_rate": 7.173642768338168e-07, + "loss": 0.4578, + "step": 10335 + }, + { + "epoch": 0.8566929133858268, + "grad_norm": 0.22983355820178986, + "learning_rate": 7.169498549523416e-07, + "loss": 0.464, + "step": 10336 + }, + { + "epoch": 0.8567757977621219, + "grad_norm": 0.23102450370788574, + "learning_rate": 7.165354330708663e-07, + "loss": 0.5042, + "step": 10337 + }, + { + "epoch": 0.8568586821384169, + "grad_norm": 0.21612678468227386, + "learning_rate": 7.161210111893908e-07, + "loss": 0.442, + "step": 10338 + }, + { + "epoch": 0.856941566514712, + "grad_norm": 0.22449417412281036, + "learning_rate": 7.157065893079156e-07, + "loss": 0.4873, + "step": 10339 + }, + { + "epoch": 0.8570244508910071, + "grad_norm": 0.2370496392250061, + "learning_rate": 7.152921674264401e-07, + "loss": 0.4784, + "step": 10340 + }, + { + "epoch": 0.8571073352673021, + "grad_norm": 0.24634592235088348, + "learning_rate": 7.148777455449649e-07, + "loss": 0.4847, + "step": 10341 + }, + { + "epoch": 0.8571902196435972, + "grad_norm": 0.24353289604187012, + "learning_rate": 7.144633236634894e-07, + "loss": 0.5012, + "step": 10342 + }, + { + "epoch": 0.8572731040198922, + "grad_norm": 0.24136339128017426, + "learning_rate": 7.140489017820142e-07, + "loss": 0.5, + "step": 10343 + }, + { + "epoch": 0.8573559883961873, + "grad_norm": 0.2271755039691925, + "learning_rate": 7.136344799005387e-07, + "loss": 0.4826, + "step": 10344 + }, + { + "epoch": 0.8574388727724824, + "grad_norm": 0.22803209722042084, + "learning_rate": 7.132200580190635e-07, + "loss": 0.4696, + "step": 10345 + }, + { + "epoch": 0.8575217571487774, + "grad_norm": 0.22872911393642426, + "learning_rate": 7.128056361375882e-07, + "loss": 0.4494, + "step": 10346 + }, + { + "epoch": 0.8576046415250725, + "grad_norm": 0.24746474623680115, + "learning_rate": 7.123912142561127e-07, + "loss": 0.5176, + "step": 10347 + }, + { + "epoch": 0.8576875259013675, + "grad_norm": 0.25377973914146423, + "learning_rate": 7.119767923746375e-07, + "loss": 0.5138, + "step": 10348 + }, + { + "epoch": 0.8577704102776627, + "grad_norm": 0.2288646101951599, + "learning_rate": 7.11562370493162e-07, + "loss": 0.4758, + "step": 10349 + }, + { + "epoch": 0.8578532946539578, + "grad_norm": 0.24531289935112, + "learning_rate": 7.111479486116868e-07, + "loss": 0.4924, + "step": 10350 + }, + { + "epoch": 0.8579361790302528, + "grad_norm": 0.23416048288345337, + "learning_rate": 7.107335267302113e-07, + "loss": 0.459, + "step": 10351 + }, + { + "epoch": 0.8580190634065479, + "grad_norm": 0.24261118471622467, + "learning_rate": 7.103191048487361e-07, + "loss": 0.523, + "step": 10352 + }, + { + "epoch": 0.858101947782843, + "grad_norm": 0.2542513906955719, + "learning_rate": 7.099046829672607e-07, + "loss": 0.5281, + "step": 10353 + }, + { + "epoch": 0.858184832159138, + "grad_norm": 0.23680295050144196, + "learning_rate": 7.094902610857854e-07, + "loss": 0.4946, + "step": 10354 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 0.2580215334892273, + "learning_rate": 7.090758392043101e-07, + "loss": 0.4668, + "step": 10355 + }, + { + "epoch": 0.8583506009117281, + "grad_norm": 0.24382171034812927, + "learning_rate": 7.086614173228346e-07, + "loss": 0.5299, + "step": 10356 + }, + { + "epoch": 0.8584334852880232, + "grad_norm": 0.23492635786533356, + "learning_rate": 7.082469954413594e-07, + "loss": 0.4965, + "step": 10357 + }, + { + "epoch": 0.8585163696643183, + "grad_norm": 0.25048553943634033, + "learning_rate": 7.07832573559884e-07, + "loss": 0.515, + "step": 10358 + }, + { + "epoch": 0.8585992540406133, + "grad_norm": 0.23849879205226898, + "learning_rate": 7.074181516784087e-07, + "loss": 0.509, + "step": 10359 + }, + { + "epoch": 0.8586821384169084, + "grad_norm": 0.23073303699493408, + "learning_rate": 7.070037297969333e-07, + "loss": 0.4695, + "step": 10360 + }, + { + "epoch": 0.8587650227932034, + "grad_norm": 0.23332346975803375, + "learning_rate": 7.06589307915458e-07, + "loss": 0.4623, + "step": 10361 + }, + { + "epoch": 0.8588479071694985, + "grad_norm": 0.23851469159126282, + "learning_rate": 7.061748860339826e-07, + "loss": 0.5027, + "step": 10362 + }, + { + "epoch": 0.8589307915457937, + "grad_norm": 0.22192692756652832, + "learning_rate": 7.057604641525074e-07, + "loss": 0.4797, + "step": 10363 + }, + { + "epoch": 0.8590136759220887, + "grad_norm": 0.24306362867355347, + "learning_rate": 7.053460422710319e-07, + "loss": 0.4867, + "step": 10364 + }, + { + "epoch": 0.8590965602983838, + "grad_norm": 0.23576945066452026, + "learning_rate": 7.049316203895566e-07, + "loss": 0.4848, + "step": 10365 + }, + { + "epoch": 0.8591794446746788, + "grad_norm": 0.237528994679451, + "learning_rate": 7.045171985080813e-07, + "loss": 0.5115, + "step": 10366 + }, + { + "epoch": 0.8592623290509739, + "grad_norm": 0.2530480921268463, + "learning_rate": 7.041027766266059e-07, + "loss": 0.5057, + "step": 10367 + }, + { + "epoch": 0.859345213427269, + "grad_norm": 0.2640637159347534, + "learning_rate": 7.036883547451307e-07, + "loss": 0.5105, + "step": 10368 + }, + { + "epoch": 0.859428097803564, + "grad_norm": 0.2343357503414154, + "learning_rate": 7.032739328636552e-07, + "loss": 0.4818, + "step": 10369 + }, + { + "epoch": 0.8595109821798591, + "grad_norm": 0.24232929944992065, + "learning_rate": 7.0285951098218e-07, + "loss": 0.4845, + "step": 10370 + }, + { + "epoch": 0.8595938665561542, + "grad_norm": 0.22817008197307587, + "learning_rate": 7.024450891007045e-07, + "loss": 0.4837, + "step": 10371 + }, + { + "epoch": 0.8596767509324492, + "grad_norm": 0.23796483874320984, + "learning_rate": 7.020306672192293e-07, + "loss": 0.502, + "step": 10372 + }, + { + "epoch": 0.8597596353087443, + "grad_norm": 0.23888789117336273, + "learning_rate": 7.016162453377538e-07, + "loss": 0.451, + "step": 10373 + }, + { + "epoch": 0.8598425196850393, + "grad_norm": 0.2502925395965576, + "learning_rate": 7.012018234562785e-07, + "loss": 0.5055, + "step": 10374 + }, + { + "epoch": 0.8599254040613344, + "grad_norm": 0.24075886607170105, + "learning_rate": 7.007874015748033e-07, + "loss": 0.4977, + "step": 10375 + }, + { + "epoch": 0.8600082884376296, + "grad_norm": 0.2617565095424652, + "learning_rate": 7.003729796933278e-07, + "loss": 0.5253, + "step": 10376 + }, + { + "epoch": 0.8600911728139246, + "grad_norm": 0.2341056913137436, + "learning_rate": 6.999585578118526e-07, + "loss": 0.4894, + "step": 10377 + }, + { + "epoch": 0.8601740571902197, + "grad_norm": 0.25291991233825684, + "learning_rate": 6.995441359303771e-07, + "loss": 0.5288, + "step": 10378 + }, + { + "epoch": 0.8602569415665147, + "grad_norm": 0.24152733385562897, + "learning_rate": 6.991297140489019e-07, + "loss": 0.4894, + "step": 10379 + }, + { + "epoch": 0.8603398259428098, + "grad_norm": 0.23378592729568481, + "learning_rate": 6.987152921674264e-07, + "loss": 0.4383, + "step": 10380 + }, + { + "epoch": 0.8604227103191049, + "grad_norm": 0.23871521651744843, + "learning_rate": 6.983008702859512e-07, + "loss": 0.4959, + "step": 10381 + }, + { + "epoch": 0.8605055946953999, + "grad_norm": 0.22963115572929382, + "learning_rate": 6.978864484044758e-07, + "loss": 0.4827, + "step": 10382 + }, + { + "epoch": 0.860588479071695, + "grad_norm": 0.22990164160728455, + "learning_rate": 6.974720265230004e-07, + "loss": 0.4872, + "step": 10383 + }, + { + "epoch": 0.86067136344799, + "grad_norm": 0.24107606709003448, + "learning_rate": 6.970576046415252e-07, + "loss": 0.4969, + "step": 10384 + }, + { + "epoch": 0.8607542478242851, + "grad_norm": 0.22772182524204254, + "learning_rate": 6.966431827600497e-07, + "loss": 0.4613, + "step": 10385 + }, + { + "epoch": 0.8608371322005802, + "grad_norm": 0.22678476572036743, + "learning_rate": 6.962287608785745e-07, + "loss": 0.476, + "step": 10386 + }, + { + "epoch": 0.8609200165768752, + "grad_norm": 0.24610254168510437, + "learning_rate": 6.958143389970991e-07, + "loss": 0.5103, + "step": 10387 + }, + { + "epoch": 0.8610029009531703, + "grad_norm": 0.2335968315601349, + "learning_rate": 6.953999171156238e-07, + "loss": 0.4645, + "step": 10388 + }, + { + "epoch": 0.8610857853294654, + "grad_norm": 0.2321230173110962, + "learning_rate": 6.949854952341484e-07, + "loss": 0.465, + "step": 10389 + }, + { + "epoch": 0.8611686697057604, + "grad_norm": 0.23537367582321167, + "learning_rate": 6.945710733526731e-07, + "loss": 0.5184, + "step": 10390 + }, + { + "epoch": 0.8612515540820556, + "grad_norm": 0.22801150381565094, + "learning_rate": 6.941566514711977e-07, + "loss": 0.4616, + "step": 10391 + }, + { + "epoch": 0.8613344384583506, + "grad_norm": 0.25867757201194763, + "learning_rate": 6.937422295897224e-07, + "loss": 0.5328, + "step": 10392 + }, + { + "epoch": 0.8614173228346457, + "grad_norm": 0.23789997398853302, + "learning_rate": 6.933278077082471e-07, + "loss": 0.474, + "step": 10393 + }, + { + "epoch": 0.8615002072109408, + "grad_norm": 0.24744157493114471, + "learning_rate": 6.929133858267717e-07, + "loss": 0.4971, + "step": 10394 + }, + { + "epoch": 0.8615830915872358, + "grad_norm": 0.22648538649082184, + "learning_rate": 6.924989639452964e-07, + "loss": 0.4856, + "step": 10395 + }, + { + "epoch": 0.8616659759635309, + "grad_norm": 0.24273936450481415, + "learning_rate": 6.92084542063821e-07, + "loss": 0.466, + "step": 10396 + }, + { + "epoch": 0.8617488603398259, + "grad_norm": 0.2271689474582672, + "learning_rate": 6.916701201823458e-07, + "loss": 0.478, + "step": 10397 + }, + { + "epoch": 0.861831744716121, + "grad_norm": 0.23523102700710297, + "learning_rate": 6.912556983008703e-07, + "loss": 0.489, + "step": 10398 + }, + { + "epoch": 0.8619146290924161, + "grad_norm": 0.2557067275047302, + "learning_rate": 6.908412764193951e-07, + "loss": 0.5002, + "step": 10399 + }, + { + "epoch": 0.8619975134687111, + "grad_norm": 0.2373007833957672, + "learning_rate": 6.904268545379196e-07, + "loss": 0.4804, + "step": 10400 + }, + { + "epoch": 0.8620803978450062, + "grad_norm": 0.2559155523777008, + "learning_rate": 6.900124326564443e-07, + "loss": 0.5363, + "step": 10401 + }, + { + "epoch": 0.8621632822213012, + "grad_norm": 0.21870605647563934, + "learning_rate": 6.895980107749689e-07, + "loss": 0.4904, + "step": 10402 + }, + { + "epoch": 0.8622461665975963, + "grad_norm": 0.2441318929195404, + "learning_rate": 6.891835888934936e-07, + "loss": 0.4891, + "step": 10403 + }, + { + "epoch": 0.8623290509738915, + "grad_norm": 0.2548455595970154, + "learning_rate": 6.887691670120184e-07, + "loss": 0.4706, + "step": 10404 + }, + { + "epoch": 0.8624119353501865, + "grad_norm": 0.22640292346477509, + "learning_rate": 6.883547451305429e-07, + "loss": 0.4458, + "step": 10405 + }, + { + "epoch": 0.8624948197264816, + "grad_norm": 0.22574564814567566, + "learning_rate": 6.879403232490677e-07, + "loss": 0.4877, + "step": 10406 + }, + { + "epoch": 0.8625777041027767, + "grad_norm": 0.24554680287837982, + "learning_rate": 6.875259013675922e-07, + "loss": 0.5095, + "step": 10407 + }, + { + "epoch": 0.8626605884790717, + "grad_norm": 0.21897383034229279, + "learning_rate": 6.87111479486117e-07, + "loss": 0.4823, + "step": 10408 + }, + { + "epoch": 0.8627434728553668, + "grad_norm": 0.24826480448246002, + "learning_rate": 6.866970576046415e-07, + "loss": 0.5131, + "step": 10409 + }, + { + "epoch": 0.8628263572316618, + "grad_norm": 0.2470664232969284, + "learning_rate": 6.862826357231662e-07, + "loss": 0.5159, + "step": 10410 + }, + { + "epoch": 0.8629092416079569, + "grad_norm": 0.2501594126224518, + "learning_rate": 6.858682138416909e-07, + "loss": 0.5221, + "step": 10411 + }, + { + "epoch": 0.862992125984252, + "grad_norm": 0.23408162593841553, + "learning_rate": 6.854537919602155e-07, + "loss": 0.4863, + "step": 10412 + }, + { + "epoch": 0.863075010360547, + "grad_norm": 0.23542188107967377, + "learning_rate": 6.850393700787403e-07, + "loss": 0.4829, + "step": 10413 + }, + { + "epoch": 0.8631578947368421, + "grad_norm": 0.23727452754974365, + "learning_rate": 6.846249481972648e-07, + "loss": 0.5089, + "step": 10414 + }, + { + "epoch": 0.8632407791131371, + "grad_norm": 0.25566110014915466, + "learning_rate": 6.842105263157896e-07, + "loss": 0.5419, + "step": 10415 + }, + { + "epoch": 0.8633236634894322, + "grad_norm": 0.22418782114982605, + "learning_rate": 6.837961044343142e-07, + "loss": 0.4614, + "step": 10416 + }, + { + "epoch": 0.8634065478657273, + "grad_norm": 0.23182792961597443, + "learning_rate": 6.833816825528389e-07, + "loss": 0.5073, + "step": 10417 + }, + { + "epoch": 0.8634894322420223, + "grad_norm": 0.2347622960805893, + "learning_rate": 6.829672606713635e-07, + "loss": 0.504, + "step": 10418 + }, + { + "epoch": 0.8635723166183175, + "grad_norm": 0.24801212549209595, + "learning_rate": 6.825528387898881e-07, + "loss": 0.5044, + "step": 10419 + }, + { + "epoch": 0.8636552009946126, + "grad_norm": 0.2306470423936844, + "learning_rate": 6.821384169084128e-07, + "loss": 0.4586, + "step": 10420 + }, + { + "epoch": 0.8637380853709076, + "grad_norm": 0.24718262255191803, + "learning_rate": 6.817239950269374e-07, + "loss": 0.4908, + "step": 10421 + }, + { + "epoch": 0.8638209697472027, + "grad_norm": 0.24341876804828644, + "learning_rate": 6.813095731454622e-07, + "loss": 0.4733, + "step": 10422 + }, + { + "epoch": 0.8639038541234977, + "grad_norm": 0.22687986493110657, + "learning_rate": 6.808951512639868e-07, + "loss": 0.459, + "step": 10423 + }, + { + "epoch": 0.8639867384997928, + "grad_norm": 0.23734596371650696, + "learning_rate": 6.804807293825115e-07, + "loss": 0.5018, + "step": 10424 + }, + { + "epoch": 0.8640696228760879, + "grad_norm": 0.2531887888908386, + "learning_rate": 6.800663075010361e-07, + "loss": 0.5125, + "step": 10425 + }, + { + "epoch": 0.8641525072523829, + "grad_norm": 0.24079376459121704, + "learning_rate": 6.796518856195608e-07, + "loss": 0.532, + "step": 10426 + }, + { + "epoch": 0.864235391628678, + "grad_norm": 0.2560756504535675, + "learning_rate": 6.792374637380854e-07, + "loss": 0.528, + "step": 10427 + }, + { + "epoch": 0.864318276004973, + "grad_norm": 0.24267137050628662, + "learning_rate": 6.788230418566101e-07, + "loss": 0.4809, + "step": 10428 + }, + { + "epoch": 0.8644011603812681, + "grad_norm": 0.24853646755218506, + "learning_rate": 6.784086199751347e-07, + "loss": 0.4913, + "step": 10429 + }, + { + "epoch": 0.8644840447575632, + "grad_norm": 0.22673338651657104, + "learning_rate": 6.779941980936594e-07, + "loss": 0.4729, + "step": 10430 + }, + { + "epoch": 0.8645669291338582, + "grad_norm": 0.22156505286693573, + "learning_rate": 6.775797762121841e-07, + "loss": 0.4626, + "step": 10431 + }, + { + "epoch": 0.8646498135101534, + "grad_norm": 0.2630302906036377, + "learning_rate": 6.771653543307087e-07, + "loss": 0.502, + "step": 10432 + }, + { + "epoch": 0.8647326978864484, + "grad_norm": 0.2405131459236145, + "learning_rate": 6.767509324492335e-07, + "loss": 0.5059, + "step": 10433 + }, + { + "epoch": 0.8648155822627435, + "grad_norm": 0.23788610100746155, + "learning_rate": 6.76336510567758e-07, + "loss": 0.5097, + "step": 10434 + }, + { + "epoch": 0.8648984666390386, + "grad_norm": 0.24350892007350922, + "learning_rate": 6.759220886862828e-07, + "loss": 0.4691, + "step": 10435 + }, + { + "epoch": 0.8649813510153336, + "grad_norm": 0.234684556722641, + "learning_rate": 6.755076668048073e-07, + "loss": 0.4762, + "step": 10436 + }, + { + "epoch": 0.8650642353916287, + "grad_norm": 0.2676331400871277, + "learning_rate": 6.75093244923332e-07, + "loss": 0.511, + "step": 10437 + }, + { + "epoch": 0.8651471197679238, + "grad_norm": 0.23197218775749207, + "learning_rate": 6.746788230418566e-07, + "loss": 0.4643, + "step": 10438 + }, + { + "epoch": 0.8652300041442188, + "grad_norm": 0.25092610716819763, + "learning_rate": 6.742644011603813e-07, + "loss": 0.5014, + "step": 10439 + }, + { + "epoch": 0.8653128885205139, + "grad_norm": 0.2332131415605545, + "learning_rate": 6.738499792789059e-07, + "loss": 0.4911, + "step": 10440 + }, + { + "epoch": 0.8653957728968089, + "grad_norm": 0.22043362259864807, + "learning_rate": 6.734355573974306e-07, + "loss": 0.4901, + "step": 10441 + }, + { + "epoch": 0.865478657273104, + "grad_norm": 0.24154694378376007, + "learning_rate": 6.730211355159554e-07, + "loss": 0.4698, + "step": 10442 + }, + { + "epoch": 0.8655615416493991, + "grad_norm": 0.2275063693523407, + "learning_rate": 6.726067136344799e-07, + "loss": 0.489, + "step": 10443 + }, + { + "epoch": 0.8656444260256941, + "grad_norm": 0.24147139489650726, + "learning_rate": 6.721922917530047e-07, + "loss": 0.4853, + "step": 10444 + }, + { + "epoch": 0.8657273104019892, + "grad_norm": 0.24849610030651093, + "learning_rate": 6.717778698715292e-07, + "loss": 0.4695, + "step": 10445 + }, + { + "epoch": 0.8658101947782842, + "grad_norm": 0.2240774929523468, + "learning_rate": 6.713634479900539e-07, + "loss": 0.4839, + "step": 10446 + }, + { + "epoch": 0.8658930791545794, + "grad_norm": 0.24636028707027435, + "learning_rate": 6.709490261085786e-07, + "loss": 0.5023, + "step": 10447 + }, + { + "epoch": 0.8659759635308745, + "grad_norm": 0.24324464797973633, + "learning_rate": 6.705346042271032e-07, + "loss": 0.4543, + "step": 10448 + }, + { + "epoch": 0.8660588479071695, + "grad_norm": 0.2414129376411438, + "learning_rate": 6.701201823456279e-07, + "loss": 0.4863, + "step": 10449 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 0.21970386803150177, + "learning_rate": 6.697057604641525e-07, + "loss": 0.4619, + "step": 10450 + }, + { + "epoch": 0.8662246166597596, + "grad_norm": 0.26780542731285095, + "learning_rate": 6.692913385826773e-07, + "loss": 0.5017, + "step": 10451 + }, + { + "epoch": 0.8663075010360547, + "grad_norm": 0.2349386215209961, + "learning_rate": 6.688769167012019e-07, + "loss": 0.5269, + "step": 10452 + }, + { + "epoch": 0.8663903854123498, + "grad_norm": 0.24408020079135895, + "learning_rate": 6.684624948197265e-07, + "loss": 0.4782, + "step": 10453 + }, + { + "epoch": 0.8664732697886448, + "grad_norm": 0.2714211940765381, + "learning_rate": 6.680480729382512e-07, + "loss": 0.4749, + "step": 10454 + }, + { + "epoch": 0.8665561541649399, + "grad_norm": 0.2409812957048416, + "learning_rate": 6.676336510567758e-07, + "loss": 0.52, + "step": 10455 + }, + { + "epoch": 0.866639038541235, + "grad_norm": 0.24099195003509521, + "learning_rate": 6.672192291753005e-07, + "loss": 0.4919, + "step": 10456 + }, + { + "epoch": 0.86672192291753, + "grad_norm": 0.2697199583053589, + "learning_rate": 6.668048072938252e-07, + "loss": 0.473, + "step": 10457 + }, + { + "epoch": 0.8668048072938251, + "grad_norm": 0.22637178003787994, + "learning_rate": 6.663903854123498e-07, + "loss": 0.4603, + "step": 10458 + }, + { + "epoch": 0.8668876916701201, + "grad_norm": 0.23107092082500458, + "learning_rate": 6.659759635308745e-07, + "loss": 0.4876, + "step": 10459 + }, + { + "epoch": 0.8669705760464153, + "grad_norm": 0.24836061894893646, + "learning_rate": 6.655615416493992e-07, + "loss": 0.4845, + "step": 10460 + }, + { + "epoch": 0.8670534604227104, + "grad_norm": 0.24095042049884796, + "learning_rate": 6.651471197679238e-07, + "loss": 0.4936, + "step": 10461 + }, + { + "epoch": 0.8671363447990054, + "grad_norm": 0.24323716759681702, + "learning_rate": 6.647326978864486e-07, + "loss": 0.5236, + "step": 10462 + }, + { + "epoch": 0.8672192291753005, + "grad_norm": 0.24480850994586945, + "learning_rate": 6.643182760049731e-07, + "loss": 0.4797, + "step": 10463 + }, + { + "epoch": 0.8673021135515955, + "grad_norm": 0.22926463186740875, + "learning_rate": 6.639038541234978e-07, + "loss": 0.4706, + "step": 10464 + }, + { + "epoch": 0.8673849979278906, + "grad_norm": 0.21717849373817444, + "learning_rate": 6.634894322420224e-07, + "loss": 0.4124, + "step": 10465 + }, + { + "epoch": 0.8674678823041857, + "grad_norm": 0.24205565452575684, + "learning_rate": 6.630750103605471e-07, + "loss": 0.48, + "step": 10466 + }, + { + "epoch": 0.8675507666804807, + "grad_norm": 0.24205125868320465, + "learning_rate": 6.626605884790717e-07, + "loss": 0.5018, + "step": 10467 + }, + { + "epoch": 0.8676336510567758, + "grad_norm": 0.248072549700737, + "learning_rate": 6.622461665975964e-07, + "loss": 0.5045, + "step": 10468 + }, + { + "epoch": 0.8677165354330708, + "grad_norm": 0.23933669924736023, + "learning_rate": 6.618317447161212e-07, + "loss": 0.5069, + "step": 10469 + }, + { + "epoch": 0.8677994198093659, + "grad_norm": 0.2457849085330963, + "learning_rate": 6.614173228346457e-07, + "loss": 0.5038, + "step": 10470 + }, + { + "epoch": 0.867882304185661, + "grad_norm": 0.23469290137290955, + "learning_rate": 6.610029009531705e-07, + "loss": 0.5164, + "step": 10471 + }, + { + "epoch": 0.867965188561956, + "grad_norm": 0.25297197699546814, + "learning_rate": 6.60588479071695e-07, + "loss": 0.5192, + "step": 10472 + }, + { + "epoch": 0.8680480729382511, + "grad_norm": 0.2624039053916931, + "learning_rate": 6.601740571902197e-07, + "loss": 0.502, + "step": 10473 + }, + { + "epoch": 0.8681309573145463, + "grad_norm": 0.22233620285987854, + "learning_rate": 6.597596353087443e-07, + "loss": 0.4521, + "step": 10474 + }, + { + "epoch": 0.8682138416908413, + "grad_norm": 0.24949030578136444, + "learning_rate": 6.59345213427269e-07, + "loss": 0.4991, + "step": 10475 + }, + { + "epoch": 0.8682967260671364, + "grad_norm": 0.23062443733215332, + "learning_rate": 6.589307915457937e-07, + "loss": 0.4716, + "step": 10476 + }, + { + "epoch": 0.8683796104434314, + "grad_norm": 0.2523786127567291, + "learning_rate": 6.585163696643183e-07, + "loss": 0.4838, + "step": 10477 + }, + { + "epoch": 0.8684624948197265, + "grad_norm": 0.24099092185497284, + "learning_rate": 6.58101947782843e-07, + "loss": 0.5303, + "step": 10478 + }, + { + "epoch": 0.8685453791960216, + "grad_norm": 0.23888930678367615, + "learning_rate": 6.576875259013676e-07, + "loss": 0.4638, + "step": 10479 + }, + { + "epoch": 0.8686282635723166, + "grad_norm": 0.2385653555393219, + "learning_rate": 6.572731040198923e-07, + "loss": 0.4595, + "step": 10480 + }, + { + "epoch": 0.8687111479486117, + "grad_norm": 0.25700628757476807, + "learning_rate": 6.56858682138417e-07, + "loss": 0.5514, + "step": 10481 + }, + { + "epoch": 0.8687940323249067, + "grad_norm": 0.23981933295726776, + "learning_rate": 6.564442602569416e-07, + "loss": 0.5159, + "step": 10482 + }, + { + "epoch": 0.8688769167012018, + "grad_norm": 0.24990224838256836, + "learning_rate": 6.560298383754663e-07, + "loss": 0.5032, + "step": 10483 + }, + { + "epoch": 0.8689598010774969, + "grad_norm": 0.23191678524017334, + "learning_rate": 6.556154164939909e-07, + "loss": 0.4665, + "step": 10484 + }, + { + "epoch": 0.8690426854537919, + "grad_norm": 0.23803876340389252, + "learning_rate": 6.552009946125156e-07, + "loss": 0.4688, + "step": 10485 + }, + { + "epoch": 0.869125569830087, + "grad_norm": 0.25340545177459717, + "learning_rate": 6.547865727310403e-07, + "loss": 0.5233, + "step": 10486 + }, + { + "epoch": 0.869208454206382, + "grad_norm": 0.2557438611984253, + "learning_rate": 6.543721508495649e-07, + "loss": 0.5231, + "step": 10487 + }, + { + "epoch": 0.8692913385826772, + "grad_norm": 0.24921031296253204, + "learning_rate": 6.539577289680896e-07, + "loss": 0.5302, + "step": 10488 + }, + { + "epoch": 0.8693742229589723, + "grad_norm": 0.23744003474712372, + "learning_rate": 6.535433070866142e-07, + "loss": 0.4982, + "step": 10489 + }, + { + "epoch": 0.8694571073352673, + "grad_norm": 0.24132639169692993, + "learning_rate": 6.531288852051389e-07, + "loss": 0.5002, + "step": 10490 + }, + { + "epoch": 0.8695399917115624, + "grad_norm": 0.21418680250644684, + "learning_rate": 6.527144633236636e-07, + "loss": 0.4401, + "step": 10491 + }, + { + "epoch": 0.8696228760878575, + "grad_norm": 0.23018322885036469, + "learning_rate": 6.523000414421882e-07, + "loss": 0.497, + "step": 10492 + }, + { + "epoch": 0.8697057604641525, + "grad_norm": 0.2461976259946823, + "learning_rate": 6.518856195607129e-07, + "loss": 0.532, + "step": 10493 + }, + { + "epoch": 0.8697886448404476, + "grad_norm": 0.23889011144638062, + "learning_rate": 6.514711976792375e-07, + "loss": 0.4758, + "step": 10494 + }, + { + "epoch": 0.8698715292167426, + "grad_norm": 0.2225179374217987, + "learning_rate": 6.510567757977622e-07, + "loss": 0.449, + "step": 10495 + }, + { + "epoch": 0.8699544135930377, + "grad_norm": 0.2365172952413559, + "learning_rate": 6.506423539162868e-07, + "loss": 0.4565, + "step": 10496 + }, + { + "epoch": 0.8700372979693328, + "grad_norm": 0.23833070695400238, + "learning_rate": 6.502279320348115e-07, + "loss": 0.4683, + "step": 10497 + }, + { + "epoch": 0.8701201823456278, + "grad_norm": 0.217156320810318, + "learning_rate": 6.498135101533363e-07, + "loss": 0.4376, + "step": 10498 + }, + { + "epoch": 0.8702030667219229, + "grad_norm": 0.23175308108329773, + "learning_rate": 6.493990882718608e-07, + "loss": 0.5276, + "step": 10499 + }, + { + "epoch": 0.8702859510982179, + "grad_norm": 0.2654508948326111, + "learning_rate": 6.489846663903855e-07, + "loss": 0.5148, + "step": 10500 + }, + { + "epoch": 0.870368835474513, + "grad_norm": 0.23307643830776215, + "learning_rate": 6.485702445089101e-07, + "loss": 0.4586, + "step": 10501 + }, + { + "epoch": 0.8704517198508082, + "grad_norm": 0.24874544143676758, + "learning_rate": 6.481558226274348e-07, + "loss": 0.5287, + "step": 10502 + }, + { + "epoch": 0.8705346042271032, + "grad_norm": 0.235933318734169, + "learning_rate": 6.477414007459594e-07, + "loss": 0.4918, + "step": 10503 + }, + { + "epoch": 0.8706174886033983, + "grad_norm": 0.23185276985168457, + "learning_rate": 6.473269788644841e-07, + "loss": 0.4663, + "step": 10504 + }, + { + "epoch": 0.8707003729796934, + "grad_norm": 0.2342299520969391, + "learning_rate": 6.469125569830086e-07, + "loss": 0.4989, + "step": 10505 + }, + { + "epoch": 0.8707832573559884, + "grad_norm": 0.3425648510456085, + "learning_rate": 6.464981351015334e-07, + "loss": 0.5179, + "step": 10506 + }, + { + "epoch": 0.8708661417322835, + "grad_norm": 0.2490633726119995, + "learning_rate": 6.460837132200581e-07, + "loss": 0.5302, + "step": 10507 + }, + { + "epoch": 0.8709490261085785, + "grad_norm": 0.25358283519744873, + "learning_rate": 6.456692913385827e-07, + "loss": 0.4971, + "step": 10508 + }, + { + "epoch": 0.8710319104848736, + "grad_norm": 0.23610176146030426, + "learning_rate": 6.452548694571074e-07, + "loss": 0.4875, + "step": 10509 + }, + { + "epoch": 0.8711147948611687, + "grad_norm": 0.2379601150751114, + "learning_rate": 6.44840447575632e-07, + "loss": 0.4967, + "step": 10510 + }, + { + "epoch": 0.8711976792374637, + "grad_norm": 0.23384535312652588, + "learning_rate": 6.444260256941567e-07, + "loss": 0.4679, + "step": 10511 + }, + { + "epoch": 0.8712805636137588, + "grad_norm": 0.24586232006549835, + "learning_rate": 6.440116038126814e-07, + "loss": 0.4952, + "step": 10512 + }, + { + "epoch": 0.8713634479900538, + "grad_norm": 0.2507035434246063, + "learning_rate": 6.43597181931206e-07, + "loss": 0.5309, + "step": 10513 + }, + { + "epoch": 0.8714463323663489, + "grad_norm": 0.2305823117494583, + "learning_rate": 6.431827600497307e-07, + "loss": 0.5489, + "step": 10514 + }, + { + "epoch": 0.871529216742644, + "grad_norm": 0.23354998230934143, + "learning_rate": 6.427683381682553e-07, + "loss": 0.4781, + "step": 10515 + }, + { + "epoch": 0.871612101118939, + "grad_norm": 0.2497081160545349, + "learning_rate": 6.423539162867799e-07, + "loss": 0.4978, + "step": 10516 + }, + { + "epoch": 0.8716949854952342, + "grad_norm": 0.2374807596206665, + "learning_rate": 6.419394944053047e-07, + "loss": 0.4741, + "step": 10517 + }, + { + "epoch": 0.8717778698715292, + "grad_norm": 0.23748111724853516, + "learning_rate": 6.415250725238293e-07, + "loss": 0.4797, + "step": 10518 + }, + { + "epoch": 0.8718607542478243, + "grad_norm": 0.23820841312408447, + "learning_rate": 6.41110650642354e-07, + "loss": 0.4908, + "step": 10519 + }, + { + "epoch": 0.8719436386241194, + "grad_norm": 0.23988866806030273, + "learning_rate": 6.406962287608786e-07, + "loss": 0.4934, + "step": 10520 + }, + { + "epoch": 0.8720265230004144, + "grad_norm": 0.2461511492729187, + "learning_rate": 6.402818068794033e-07, + "loss": 0.4923, + "step": 10521 + }, + { + "epoch": 0.8721094073767095, + "grad_norm": 0.24514631927013397, + "learning_rate": 6.39867384997928e-07, + "loss": 0.5025, + "step": 10522 + }, + { + "epoch": 0.8721922917530046, + "grad_norm": 0.23723910748958588, + "learning_rate": 6.394529631164526e-07, + "loss": 0.4647, + "step": 10523 + }, + { + "epoch": 0.8722751761292996, + "grad_norm": 0.24523504078388214, + "learning_rate": 6.390385412349773e-07, + "loss": 0.5078, + "step": 10524 + }, + { + "epoch": 0.8723580605055947, + "grad_norm": 0.23270638287067413, + "learning_rate": 6.386241193535018e-07, + "loss": 0.5164, + "step": 10525 + }, + { + "epoch": 0.8724409448818897, + "grad_norm": 0.2561385929584503, + "learning_rate": 6.382096974720266e-07, + "loss": 0.5703, + "step": 10526 + }, + { + "epoch": 0.8725238292581848, + "grad_norm": 0.24209950864315033, + "learning_rate": 6.377952755905513e-07, + "loss": 0.5057, + "step": 10527 + }, + { + "epoch": 0.87260671363448, + "grad_norm": 0.2415260672569275, + "learning_rate": 6.373808537090759e-07, + "loss": 0.4847, + "step": 10528 + }, + { + "epoch": 0.872689598010775, + "grad_norm": 0.26207756996154785, + "learning_rate": 6.369664318276006e-07, + "loss": 0.5148, + "step": 10529 + }, + { + "epoch": 0.8727724823870701, + "grad_norm": 0.2177925407886505, + "learning_rate": 6.365520099461252e-07, + "loss": 0.4066, + "step": 10530 + }, + { + "epoch": 0.8728553667633651, + "grad_norm": 0.23358161747455597, + "learning_rate": 6.361375880646499e-07, + "loss": 0.4393, + "step": 10531 + }, + { + "epoch": 0.8729382511396602, + "grad_norm": 0.2372596263885498, + "learning_rate": 6.357231661831744e-07, + "loss": 0.4919, + "step": 10532 + }, + { + "epoch": 0.8730211355159553, + "grad_norm": 0.23004713654518127, + "learning_rate": 6.353087443016992e-07, + "loss": 0.4922, + "step": 10533 + }, + { + "epoch": 0.8731040198922503, + "grad_norm": 0.24972255527973175, + "learning_rate": 6.348943224202237e-07, + "loss": 0.4923, + "step": 10534 + }, + { + "epoch": 0.8731869042685454, + "grad_norm": 0.25839030742645264, + "learning_rate": 6.344799005387485e-07, + "loss": 0.5023, + "step": 10535 + }, + { + "epoch": 0.8732697886448404, + "grad_norm": 0.2593533992767334, + "learning_rate": 6.340654786572732e-07, + "loss": 0.5229, + "step": 10536 + }, + { + "epoch": 0.8733526730211355, + "grad_norm": 0.22724947333335876, + "learning_rate": 6.336510567757978e-07, + "loss": 0.4615, + "step": 10537 + }, + { + "epoch": 0.8734355573974306, + "grad_norm": 0.23050376772880554, + "learning_rate": 6.332366348943225e-07, + "loss": 0.4992, + "step": 10538 + }, + { + "epoch": 0.8735184417737256, + "grad_norm": 0.24865610897541046, + "learning_rate": 6.328222130128471e-07, + "loss": 0.4755, + "step": 10539 + }, + { + "epoch": 0.8736013261500207, + "grad_norm": 0.23608516156673431, + "learning_rate": 6.324077911313718e-07, + "loss": 0.486, + "step": 10540 + }, + { + "epoch": 0.8736842105263158, + "grad_norm": 0.235658660531044, + "learning_rate": 6.319933692498965e-07, + "loss": 0.5016, + "step": 10541 + }, + { + "epoch": 0.8737670949026108, + "grad_norm": 0.24481327831745148, + "learning_rate": 6.315789473684211e-07, + "loss": 0.4958, + "step": 10542 + }, + { + "epoch": 0.873849979278906, + "grad_norm": 0.23944178223609924, + "learning_rate": 6.311645254869457e-07, + "loss": 0.5034, + "step": 10543 + }, + { + "epoch": 0.873932863655201, + "grad_norm": 0.235836461186409, + "learning_rate": 6.307501036054704e-07, + "loss": 0.5271, + "step": 10544 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 0.21936780214309692, + "learning_rate": 6.303356817239951e-07, + "loss": 0.4402, + "step": 10545 + }, + { + "epoch": 0.8740986324077912, + "grad_norm": 0.23109248280525208, + "learning_rate": 6.299212598425198e-07, + "loss": 0.4724, + "step": 10546 + }, + { + "epoch": 0.8741815167840862, + "grad_norm": 0.2507898211479187, + "learning_rate": 6.295068379610444e-07, + "loss": 0.4901, + "step": 10547 + }, + { + "epoch": 0.8742644011603813, + "grad_norm": 0.266422301530838, + "learning_rate": 6.290924160795691e-07, + "loss": 0.5272, + "step": 10548 + }, + { + "epoch": 0.8743472855366763, + "grad_norm": 0.2422172725200653, + "learning_rate": 6.286779941980937e-07, + "loss": 0.4805, + "step": 10549 + }, + { + "epoch": 0.8744301699129714, + "grad_norm": 0.23662853240966797, + "learning_rate": 6.282635723166184e-07, + "loss": 0.4625, + "step": 10550 + }, + { + "epoch": 0.8745130542892665, + "grad_norm": 0.24254420399665833, + "learning_rate": 6.278491504351431e-07, + "loss": 0.496, + "step": 10551 + }, + { + "epoch": 0.8745959386655615, + "grad_norm": 0.25531691312789917, + "learning_rate": 6.274347285536676e-07, + "loss": 0.5046, + "step": 10552 + }, + { + "epoch": 0.8746788230418566, + "grad_norm": 0.2366497963666916, + "learning_rate": 6.270203066721924e-07, + "loss": 0.4888, + "step": 10553 + }, + { + "epoch": 0.8747617074181516, + "grad_norm": 0.23355457186698914, + "learning_rate": 6.266058847907169e-07, + "loss": 0.4655, + "step": 10554 + }, + { + "epoch": 0.8748445917944467, + "grad_norm": 0.23415374755859375, + "learning_rate": 6.261914629092417e-07, + "loss": 0.4413, + "step": 10555 + }, + { + "epoch": 0.8749274761707418, + "grad_norm": 0.23752576112747192, + "learning_rate": 6.257770410277664e-07, + "loss": 0.4977, + "step": 10556 + }, + { + "epoch": 0.8750103605470368, + "grad_norm": 0.22895501554012299, + "learning_rate": 6.25362619146291e-07, + "loss": 0.4479, + "step": 10557 + }, + { + "epoch": 0.875093244923332, + "grad_norm": 0.23855642974376678, + "learning_rate": 6.249481972648156e-07, + "loss": 0.4935, + "step": 10558 + }, + { + "epoch": 0.8751761292996271, + "grad_norm": 0.2311975508928299, + "learning_rate": 6.245337753833402e-07, + "loss": 0.4688, + "step": 10559 + }, + { + "epoch": 0.8752590136759221, + "grad_norm": 0.22910326719284058, + "learning_rate": 6.24119353501865e-07, + "loss": 0.4747, + "step": 10560 + }, + { + "epoch": 0.8753418980522172, + "grad_norm": 0.23769104480743408, + "learning_rate": 6.237049316203896e-07, + "loss": 0.491, + "step": 10561 + }, + { + "epoch": 0.8754247824285122, + "grad_norm": 0.24516616761684418, + "learning_rate": 6.232905097389143e-07, + "loss": 0.4792, + "step": 10562 + }, + { + "epoch": 0.8755076668048073, + "grad_norm": 0.2379389852285385, + "learning_rate": 6.228760878574389e-07, + "loss": 0.481, + "step": 10563 + }, + { + "epoch": 0.8755905511811024, + "grad_norm": 0.31120458245277405, + "learning_rate": 6.224616659759636e-07, + "loss": 0.5461, + "step": 10564 + }, + { + "epoch": 0.8756734355573974, + "grad_norm": 0.22652584314346313, + "learning_rate": 6.220472440944882e-07, + "loss": 0.5081, + "step": 10565 + }, + { + "epoch": 0.8757563199336925, + "grad_norm": 0.2234327793121338, + "learning_rate": 6.216328222130129e-07, + "loss": 0.4335, + "step": 10566 + }, + { + "epoch": 0.8758392043099875, + "grad_norm": 0.2337707132101059, + "learning_rate": 6.212184003315375e-07, + "loss": 0.4922, + "step": 10567 + }, + { + "epoch": 0.8759220886862826, + "grad_norm": 0.24665385484695435, + "learning_rate": 6.208039784500621e-07, + "loss": 0.5165, + "step": 10568 + }, + { + "epoch": 0.8760049730625777, + "grad_norm": 0.2442847341299057, + "learning_rate": 6.203895565685869e-07, + "loss": 0.5166, + "step": 10569 + }, + { + "epoch": 0.8760878574388727, + "grad_norm": 0.23023630678653717, + "learning_rate": 6.199751346871115e-07, + "loss": 0.4681, + "step": 10570 + }, + { + "epoch": 0.8761707418151679, + "grad_norm": 0.23273538053035736, + "learning_rate": 6.195607128056362e-07, + "loss": 0.4661, + "step": 10571 + }, + { + "epoch": 0.876253626191463, + "grad_norm": 0.2422524243593216, + "learning_rate": 6.191462909241608e-07, + "loss": 0.4961, + "step": 10572 + }, + { + "epoch": 0.876336510567758, + "grad_norm": 0.25597327947616577, + "learning_rate": 6.187318690426855e-07, + "loss": 0.5341, + "step": 10573 + }, + { + "epoch": 0.8764193949440531, + "grad_norm": 0.24855856597423553, + "learning_rate": 6.183174471612101e-07, + "loss": 0.5294, + "step": 10574 + }, + { + "epoch": 0.8765022793203481, + "grad_norm": 0.23975889384746552, + "learning_rate": 6.179030252797348e-07, + "loss": 0.4714, + "step": 10575 + }, + { + "epoch": 0.8765851636966432, + "grad_norm": 0.22056008875370026, + "learning_rate": 6.174886033982595e-07, + "loss": 0.4466, + "step": 10576 + }, + { + "epoch": 0.8766680480729383, + "grad_norm": 0.24767206609249115, + "learning_rate": 6.170741815167842e-07, + "loss": 0.4739, + "step": 10577 + }, + { + "epoch": 0.8767509324492333, + "grad_norm": 0.247758150100708, + "learning_rate": 6.166597596353088e-07, + "loss": 0.49, + "step": 10578 + }, + { + "epoch": 0.8768338168255284, + "grad_norm": 0.2392001748085022, + "learning_rate": 6.162453377538334e-07, + "loss": 0.5092, + "step": 10579 + }, + { + "epoch": 0.8769167012018234, + "grad_norm": 0.2503005564212799, + "learning_rate": 6.158309158723581e-07, + "loss": 0.5486, + "step": 10580 + }, + { + "epoch": 0.8769995855781185, + "grad_norm": 0.2397831827402115, + "learning_rate": 6.154164939908828e-07, + "loss": 0.4753, + "step": 10581 + }, + { + "epoch": 0.8770824699544136, + "grad_norm": 0.23750650882720947, + "learning_rate": 6.150020721094075e-07, + "loss": 0.509, + "step": 10582 + }, + { + "epoch": 0.8771653543307086, + "grad_norm": 0.26563072204589844, + "learning_rate": 6.145876502279321e-07, + "loss": 0.5646, + "step": 10583 + }, + { + "epoch": 0.8772482387070037, + "grad_norm": 0.24446925520896912, + "learning_rate": 6.141732283464568e-07, + "loss": 0.5116, + "step": 10584 + }, + { + "epoch": 0.8773311230832987, + "grad_norm": 0.26253554224967957, + "learning_rate": 6.137588064649814e-07, + "loss": 0.5101, + "step": 10585 + }, + { + "epoch": 0.8774140074595939, + "grad_norm": 0.2384743094444275, + "learning_rate": 6.13344384583506e-07, + "loss": 0.4944, + "step": 10586 + }, + { + "epoch": 0.877496891835889, + "grad_norm": 0.24804633855819702, + "learning_rate": 6.129299627020307e-07, + "loss": 0.4929, + "step": 10587 + }, + { + "epoch": 0.877579776212184, + "grad_norm": 0.2343822717666626, + "learning_rate": 6.125155408205554e-07, + "loss": 0.497, + "step": 10588 + }, + { + "epoch": 0.8776626605884791, + "grad_norm": 0.2423899620771408, + "learning_rate": 6.121011189390801e-07, + "loss": 0.4906, + "step": 10589 + }, + { + "epoch": 0.8777455449647742, + "grad_norm": 0.2416868954896927, + "learning_rate": 6.116866970576047e-07, + "loss": 0.4681, + "step": 10590 + }, + { + "epoch": 0.8778284293410692, + "grad_norm": 0.40016359090805054, + "learning_rate": 6.112722751761294e-07, + "loss": 0.4696, + "step": 10591 + }, + { + "epoch": 0.8779113137173643, + "grad_norm": 0.2769862115383148, + "learning_rate": 6.10857853294654e-07, + "loss": 0.546, + "step": 10592 + }, + { + "epoch": 0.8779941980936593, + "grad_norm": 0.25176578760147095, + "learning_rate": 6.104434314131787e-07, + "loss": 0.4922, + "step": 10593 + }, + { + "epoch": 0.8780770824699544, + "grad_norm": 0.2389220893383026, + "learning_rate": 6.100290095317033e-07, + "loss": 0.5242, + "step": 10594 + }, + { + "epoch": 0.8781599668462495, + "grad_norm": 0.2427835613489151, + "learning_rate": 6.096145876502279e-07, + "loss": 0.4925, + "step": 10595 + }, + { + "epoch": 0.8782428512225445, + "grad_norm": 0.22767971456050873, + "learning_rate": 6.092001657687526e-07, + "loss": 0.4643, + "step": 10596 + }, + { + "epoch": 0.8783257355988396, + "grad_norm": 0.24275954067707062, + "learning_rate": 6.087857438872772e-07, + "loss": 0.4783, + "step": 10597 + }, + { + "epoch": 0.8784086199751346, + "grad_norm": 0.2400035858154297, + "learning_rate": 6.08371322005802e-07, + "loss": 0.519, + "step": 10598 + }, + { + "epoch": 0.8784915043514298, + "grad_norm": 0.23803068697452545, + "learning_rate": 6.079569001243266e-07, + "loss": 0.5065, + "step": 10599 + }, + { + "epoch": 0.8785743887277249, + "grad_norm": 0.2292993813753128, + "learning_rate": 6.075424782428513e-07, + "loss": 0.4502, + "step": 10600 + }, + { + "epoch": 0.8786572731040199, + "grad_norm": 0.24336911737918854, + "learning_rate": 6.071280563613759e-07, + "loss": 0.511, + "step": 10601 + }, + { + "epoch": 0.878740157480315, + "grad_norm": 0.23407939076423645, + "learning_rate": 6.067136344799006e-07, + "loss": 0.4723, + "step": 10602 + }, + { + "epoch": 0.87882304185661, + "grad_norm": 0.2615978419780731, + "learning_rate": 6.062992125984252e-07, + "loss": 0.5397, + "step": 10603 + }, + { + "epoch": 0.8789059262329051, + "grad_norm": 0.24471724033355713, + "learning_rate": 6.058847907169498e-07, + "loss": 0.5553, + "step": 10604 + }, + { + "epoch": 0.8789888106092002, + "grad_norm": 0.25241485238075256, + "learning_rate": 6.054703688354745e-07, + "loss": 0.535, + "step": 10605 + }, + { + "epoch": 0.8790716949854952, + "grad_norm": 0.24397018551826477, + "learning_rate": 6.050559469539992e-07, + "loss": 0.4788, + "step": 10606 + }, + { + "epoch": 0.8791545793617903, + "grad_norm": 0.23121164739131927, + "learning_rate": 6.046415250725239e-07, + "loss": 0.498, + "step": 10607 + }, + { + "epoch": 0.8792374637380854, + "grad_norm": 0.24014584720134735, + "learning_rate": 6.042271031910485e-07, + "loss": 0.5555, + "step": 10608 + }, + { + "epoch": 0.8793203481143804, + "grad_norm": 0.2435484081506729, + "learning_rate": 6.038126813095732e-07, + "loss": 0.4874, + "step": 10609 + }, + { + "epoch": 0.8794032324906755, + "grad_norm": 0.2498323768377304, + "learning_rate": 6.033982594280978e-07, + "loss": 0.5547, + "step": 10610 + }, + { + "epoch": 0.8794861168669705, + "grad_norm": 0.2278405874967575, + "learning_rate": 6.029838375466226e-07, + "loss": 0.4495, + "step": 10611 + }, + { + "epoch": 0.8795690012432656, + "grad_norm": 0.24638262391090393, + "learning_rate": 6.025694156651472e-07, + "loss": 0.4889, + "step": 10612 + }, + { + "epoch": 0.8796518856195608, + "grad_norm": 0.23271869122982025, + "learning_rate": 6.021549937836718e-07, + "loss": 0.4662, + "step": 10613 + }, + { + "epoch": 0.8797347699958558, + "grad_norm": 0.2300374060869217, + "learning_rate": 6.017405719021965e-07, + "loss": 0.4598, + "step": 10614 + }, + { + "epoch": 0.8798176543721509, + "grad_norm": 0.22762729227542877, + "learning_rate": 6.013261500207211e-07, + "loss": 0.5211, + "step": 10615 + }, + { + "epoch": 0.8799005387484459, + "grad_norm": 0.236034095287323, + "learning_rate": 6.009117281392458e-07, + "loss": 0.4839, + "step": 10616 + }, + { + "epoch": 0.879983423124741, + "grad_norm": 0.2509174346923828, + "learning_rate": 6.004973062577705e-07, + "loss": 0.504, + "step": 10617 + }, + { + "epoch": 0.8800663075010361, + "grad_norm": 0.26067832112312317, + "learning_rate": 6.000828843762952e-07, + "loss": 0.4612, + "step": 10618 + }, + { + "epoch": 0.8801491918773311, + "grad_norm": 0.24835748970508575, + "learning_rate": 5.996684624948198e-07, + "loss": 0.529, + "step": 10619 + }, + { + "epoch": 0.8802320762536262, + "grad_norm": 0.23650339245796204, + "learning_rate": 5.992540406133445e-07, + "loss": 0.4699, + "step": 10620 + }, + { + "epoch": 0.8803149606299212, + "grad_norm": 0.2361430525779724, + "learning_rate": 5.988396187318691e-07, + "loss": 0.5008, + "step": 10621 + }, + { + "epoch": 0.8803978450062163, + "grad_norm": 0.22231514751911163, + "learning_rate": 5.984251968503937e-07, + "loss": 0.4422, + "step": 10622 + }, + { + "epoch": 0.8804807293825114, + "grad_norm": 0.224776491522789, + "learning_rate": 5.980107749689184e-07, + "loss": 0.4583, + "step": 10623 + }, + { + "epoch": 0.8805636137588064, + "grad_norm": 0.23963885009288788, + "learning_rate": 5.97596353087443e-07, + "loss": 0.5109, + "step": 10624 + }, + { + "epoch": 0.8806464981351015, + "grad_norm": 0.23698106408119202, + "learning_rate": 5.971819312059677e-07, + "loss": 0.4884, + "step": 10625 + }, + { + "epoch": 0.8807293825113967, + "grad_norm": 0.24177274107933044, + "learning_rate": 5.967675093244924e-07, + "loss": 0.4894, + "step": 10626 + }, + { + "epoch": 0.8808122668876917, + "grad_norm": 0.24016137421131134, + "learning_rate": 5.963530874430171e-07, + "loss": 0.4926, + "step": 10627 + }, + { + "epoch": 0.8808951512639868, + "grad_norm": 0.2516383230686188, + "learning_rate": 5.959386655615417e-07, + "loss": 0.5161, + "step": 10628 + }, + { + "epoch": 0.8809780356402818, + "grad_norm": 0.23557835817337036, + "learning_rate": 5.955242436800664e-07, + "loss": 0.4953, + "step": 10629 + }, + { + "epoch": 0.8810609200165769, + "grad_norm": 0.2338138073682785, + "learning_rate": 5.95109821798591e-07, + "loss": 0.4872, + "step": 10630 + }, + { + "epoch": 0.881143804392872, + "grad_norm": 0.22945907711982727, + "learning_rate": 5.946953999171156e-07, + "loss": 0.4705, + "step": 10631 + }, + { + "epoch": 0.881226688769167, + "grad_norm": 0.24449028074741364, + "learning_rate": 5.942809780356403e-07, + "loss": 0.4946, + "step": 10632 + }, + { + "epoch": 0.8813095731454621, + "grad_norm": 0.25547993183135986, + "learning_rate": 5.938665561541649e-07, + "loss": 0.4994, + "step": 10633 + }, + { + "epoch": 0.8813924575217571, + "grad_norm": 0.2406463921070099, + "learning_rate": 5.934521342726896e-07, + "loss": 0.5308, + "step": 10634 + }, + { + "epoch": 0.8814753418980522, + "grad_norm": 0.25491687655448914, + "learning_rate": 5.930377123912143e-07, + "loss": 0.4866, + "step": 10635 + }, + { + "epoch": 0.8815582262743473, + "grad_norm": 0.23938272893428802, + "learning_rate": 5.92623290509739e-07, + "loss": 0.4881, + "step": 10636 + }, + { + "epoch": 0.8816411106506423, + "grad_norm": 0.2269599437713623, + "learning_rate": 5.922088686282636e-07, + "loss": 0.4731, + "step": 10637 + }, + { + "epoch": 0.8817239950269374, + "grad_norm": 0.2299308329820633, + "learning_rate": 5.917944467467883e-07, + "loss": 0.4682, + "step": 10638 + }, + { + "epoch": 0.8818068794032324, + "grad_norm": 0.24234677851200104, + "learning_rate": 5.913800248653129e-07, + "loss": 0.5296, + "step": 10639 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 0.23704466223716736, + "learning_rate": 5.909656029838376e-07, + "loss": 0.4588, + "step": 10640 + }, + { + "epoch": 0.8819726481558227, + "grad_norm": 0.2422153800725937, + "learning_rate": 5.905511811023623e-07, + "loss": 0.527, + "step": 10641 + }, + { + "epoch": 0.8820555325321177, + "grad_norm": 0.2563624382019043, + "learning_rate": 5.901367592208869e-07, + "loss": 0.5308, + "step": 10642 + }, + { + "epoch": 0.8821384169084128, + "grad_norm": 0.22462508082389832, + "learning_rate": 5.897223373394116e-07, + "loss": 0.4527, + "step": 10643 + }, + { + "epoch": 0.8822213012847079, + "grad_norm": 0.25808730721473694, + "learning_rate": 5.893079154579362e-07, + "loss": 0.4835, + "step": 10644 + }, + { + "epoch": 0.8823041856610029, + "grad_norm": 0.24309517443180084, + "learning_rate": 5.888934935764609e-07, + "loss": 0.4651, + "step": 10645 + }, + { + "epoch": 0.882387070037298, + "grad_norm": 0.23677220940589905, + "learning_rate": 5.884790716949856e-07, + "loss": 0.5022, + "step": 10646 + }, + { + "epoch": 0.882469954413593, + "grad_norm": 0.2437528818845749, + "learning_rate": 5.880646498135103e-07, + "loss": 0.4871, + "step": 10647 + }, + { + "epoch": 0.8825528387898881, + "grad_norm": 0.2411574423313141, + "learning_rate": 5.876502279320349e-07, + "loss": 0.4923, + "step": 10648 + }, + { + "epoch": 0.8826357231661832, + "grad_norm": 0.24421505630016327, + "learning_rate": 5.872358060505595e-07, + "loss": 0.4779, + "step": 10649 + }, + { + "epoch": 0.8827186075424782, + "grad_norm": 0.24436146020889282, + "learning_rate": 5.868213841690842e-07, + "loss": 0.4813, + "step": 10650 + }, + { + "epoch": 0.8828014919187733, + "grad_norm": 0.22965937852859497, + "learning_rate": 5.864069622876088e-07, + "loss": 0.4672, + "step": 10651 + }, + { + "epoch": 0.8828843762950683, + "grad_norm": 0.25509536266326904, + "learning_rate": 5.859925404061335e-07, + "loss": 0.507, + "step": 10652 + }, + { + "epoch": 0.8829672606713634, + "grad_norm": 0.21884776651859283, + "learning_rate": 5.855781185246581e-07, + "loss": 0.4717, + "step": 10653 + }, + { + "epoch": 0.8830501450476586, + "grad_norm": 0.2544578015804291, + "learning_rate": 5.851636966431828e-07, + "loss": 0.4915, + "step": 10654 + }, + { + "epoch": 0.8831330294239536, + "grad_norm": 0.2644634544849396, + "learning_rate": 5.847492747617075e-07, + "loss": 0.5103, + "step": 10655 + }, + { + "epoch": 0.8832159138002487, + "grad_norm": 0.24540428817272186, + "learning_rate": 5.843348528802322e-07, + "loss": 0.5357, + "step": 10656 + }, + { + "epoch": 0.8832987981765438, + "grad_norm": 0.23299342393875122, + "learning_rate": 5.839204309987568e-07, + "loss": 0.4834, + "step": 10657 + }, + { + "epoch": 0.8833816825528388, + "grad_norm": 0.23961471021175385, + "learning_rate": 5.835060091172814e-07, + "loss": 0.4681, + "step": 10658 + }, + { + "epoch": 0.8834645669291339, + "grad_norm": 0.2438056766986847, + "learning_rate": 5.830915872358061e-07, + "loss": 0.5125, + "step": 10659 + }, + { + "epoch": 0.8835474513054289, + "grad_norm": 0.25159749388694763, + "learning_rate": 5.826771653543307e-07, + "loss": 0.4986, + "step": 10660 + }, + { + "epoch": 0.883630335681724, + "grad_norm": 0.22830526530742645, + "learning_rate": 5.822627434728554e-07, + "loss": 0.4822, + "step": 10661 + }, + { + "epoch": 0.8837132200580191, + "grad_norm": 0.26311880350112915, + "learning_rate": 5.8184832159138e-07, + "loss": 0.5212, + "step": 10662 + }, + { + "epoch": 0.8837961044343141, + "grad_norm": 0.22749820351600647, + "learning_rate": 5.814338997099047e-07, + "loss": 0.4917, + "step": 10663 + }, + { + "epoch": 0.8838789888106092, + "grad_norm": 0.2537408769130707, + "learning_rate": 5.810194778284294e-07, + "loss": 0.5277, + "step": 10664 + }, + { + "epoch": 0.8839618731869042, + "grad_norm": 0.2234206646680832, + "learning_rate": 5.806050559469541e-07, + "loss": 0.4803, + "step": 10665 + }, + { + "epoch": 0.8840447575631993, + "grad_norm": 0.2282925844192505, + "learning_rate": 5.801906340654787e-07, + "loss": 0.4556, + "step": 10666 + }, + { + "epoch": 0.8841276419394944, + "grad_norm": 0.2692238986492157, + "learning_rate": 5.797762121840033e-07, + "loss": 0.4824, + "step": 10667 + }, + { + "epoch": 0.8842105263157894, + "grad_norm": 0.26156970858573914, + "learning_rate": 5.79361790302528e-07, + "loss": 0.4769, + "step": 10668 + }, + { + "epoch": 0.8842934106920846, + "grad_norm": 0.2380075305700302, + "learning_rate": 5.789473684210526e-07, + "loss": 0.5179, + "step": 10669 + }, + { + "epoch": 0.8843762950683796, + "grad_norm": 0.22906394302845, + "learning_rate": 5.785329465395774e-07, + "loss": 0.513, + "step": 10670 + }, + { + "epoch": 0.8844591794446747, + "grad_norm": 0.23783984780311584, + "learning_rate": 5.78118524658102e-07, + "loss": 0.5182, + "step": 10671 + }, + { + "epoch": 0.8845420638209698, + "grad_norm": 0.2447664737701416, + "learning_rate": 5.777041027766267e-07, + "loss": 0.4932, + "step": 10672 + }, + { + "epoch": 0.8846249481972648, + "grad_norm": 0.24222156405448914, + "learning_rate": 5.772896808951513e-07, + "loss": 0.5036, + "step": 10673 + }, + { + "epoch": 0.8847078325735599, + "grad_norm": 0.2528953552246094, + "learning_rate": 5.76875259013676e-07, + "loss": 0.4808, + "step": 10674 + }, + { + "epoch": 0.884790716949855, + "grad_norm": 0.2305499017238617, + "learning_rate": 5.764608371322007e-07, + "loss": 0.5007, + "step": 10675 + }, + { + "epoch": 0.88487360132615, + "grad_norm": 0.252806156873703, + "learning_rate": 5.760464152507253e-07, + "loss": 0.5148, + "step": 10676 + }, + { + "epoch": 0.8849564857024451, + "grad_norm": 0.23280666768550873, + "learning_rate": 5.7563199336925e-07, + "loss": 0.4505, + "step": 10677 + }, + { + "epoch": 0.8850393700787401, + "grad_norm": 0.24493597447872162, + "learning_rate": 5.752175714877746e-07, + "loss": 0.5344, + "step": 10678 + }, + { + "epoch": 0.8851222544550352, + "grad_norm": 0.23988313972949982, + "learning_rate": 5.748031496062993e-07, + "loss": 0.4942, + "step": 10679 + }, + { + "epoch": 0.8852051388313303, + "grad_norm": 0.24268558621406555, + "learning_rate": 5.743887277248239e-07, + "loss": 0.4911, + "step": 10680 + }, + { + "epoch": 0.8852880232076253, + "grad_norm": 0.22355258464813232, + "learning_rate": 5.739743058433486e-07, + "loss": 0.4419, + "step": 10681 + }, + { + "epoch": 0.8853709075839205, + "grad_norm": 0.23602676391601562, + "learning_rate": 5.735598839618732e-07, + "loss": 0.4962, + "step": 10682 + }, + { + "epoch": 0.8854537919602155, + "grad_norm": 0.2427944540977478, + "learning_rate": 5.73145462080398e-07, + "loss": 0.4818, + "step": 10683 + }, + { + "epoch": 0.8855366763365106, + "grad_norm": 0.24146044254302979, + "learning_rate": 5.727310401989226e-07, + "loss": 0.4891, + "step": 10684 + }, + { + "epoch": 0.8856195607128057, + "grad_norm": 0.2441176474094391, + "learning_rate": 5.723166183174472e-07, + "loss": 0.4988, + "step": 10685 + }, + { + "epoch": 0.8857024450891007, + "grad_norm": 0.24303144216537476, + "learning_rate": 5.719021964359719e-07, + "loss": 0.4877, + "step": 10686 + }, + { + "epoch": 0.8857853294653958, + "grad_norm": 0.23573584854602814, + "learning_rate": 5.714877745544965e-07, + "loss": 0.4824, + "step": 10687 + }, + { + "epoch": 0.8858682138416908, + "grad_norm": 0.24294672906398773, + "learning_rate": 5.710733526730212e-07, + "loss": 0.5294, + "step": 10688 + }, + { + "epoch": 0.8859510982179859, + "grad_norm": 0.2251405268907547, + "learning_rate": 5.706589307915458e-07, + "loss": 0.4497, + "step": 10689 + }, + { + "epoch": 0.886033982594281, + "grad_norm": 0.22936952114105225, + "learning_rate": 5.702445089100705e-07, + "loss": 0.4883, + "step": 10690 + }, + { + "epoch": 0.886116866970576, + "grad_norm": 0.21791782975196838, + "learning_rate": 5.698300870285951e-07, + "loss": 0.4196, + "step": 10691 + }, + { + "epoch": 0.8861997513468711, + "grad_norm": 0.23805610835552216, + "learning_rate": 5.694156651471197e-07, + "loss": 0.4986, + "step": 10692 + }, + { + "epoch": 0.8862826357231662, + "grad_norm": 0.2370453178882599, + "learning_rate": 5.690012432656445e-07, + "loss": 0.4942, + "step": 10693 + }, + { + "epoch": 0.8863655200994612, + "grad_norm": 0.24263042211532593, + "learning_rate": 5.685868213841691e-07, + "loss": 0.5066, + "step": 10694 + }, + { + "epoch": 0.8864484044757563, + "grad_norm": 0.23506022989749908, + "learning_rate": 5.681723995026938e-07, + "loss": 0.4894, + "step": 10695 + }, + { + "epoch": 0.8865312888520513, + "grad_norm": 0.264437735080719, + "learning_rate": 5.677579776212184e-07, + "loss": 0.5027, + "step": 10696 + }, + { + "epoch": 0.8866141732283465, + "grad_norm": 0.2403205931186676, + "learning_rate": 5.673435557397431e-07, + "loss": 0.4821, + "step": 10697 + }, + { + "epoch": 0.8866970576046416, + "grad_norm": 0.2616799771785736, + "learning_rate": 5.669291338582677e-07, + "loss": 0.5352, + "step": 10698 + }, + { + "epoch": 0.8867799419809366, + "grad_norm": 0.2597057819366455, + "learning_rate": 5.665147119767924e-07, + "loss": 0.4905, + "step": 10699 + }, + { + "epoch": 0.8868628263572317, + "grad_norm": 0.22728809714317322, + "learning_rate": 5.661002900953171e-07, + "loss": 0.4849, + "step": 10700 + }, + { + "epoch": 0.8869457107335267, + "grad_norm": 0.232818603515625, + "learning_rate": 5.656858682138418e-07, + "loss": 0.4864, + "step": 10701 + }, + { + "epoch": 0.8870285951098218, + "grad_norm": 0.23412619531154633, + "learning_rate": 5.652714463323664e-07, + "loss": 0.4924, + "step": 10702 + }, + { + "epoch": 0.8871114794861169, + "grad_norm": 0.24098847806453705, + "learning_rate": 5.64857024450891e-07, + "loss": 0.5505, + "step": 10703 + }, + { + "epoch": 0.8871943638624119, + "grad_norm": 0.23295345902442932, + "learning_rate": 5.644426025694157e-07, + "loss": 0.4424, + "step": 10704 + }, + { + "epoch": 0.887277248238707, + "grad_norm": 0.223934605717659, + "learning_rate": 5.640281806879404e-07, + "loss": 0.4391, + "step": 10705 + }, + { + "epoch": 0.887360132615002, + "grad_norm": 0.2418489158153534, + "learning_rate": 5.636137588064651e-07, + "loss": 0.4926, + "step": 10706 + }, + { + "epoch": 0.8874430169912971, + "grad_norm": 0.26012131571769714, + "learning_rate": 5.631993369249897e-07, + "loss": 0.5095, + "step": 10707 + }, + { + "epoch": 0.8875259013675922, + "grad_norm": 0.23982642590999603, + "learning_rate": 5.627849150435144e-07, + "loss": 0.5024, + "step": 10708 + }, + { + "epoch": 0.8876087857438872, + "grad_norm": 0.25661739706993103, + "learning_rate": 5.62370493162039e-07, + "loss": 0.4811, + "step": 10709 + }, + { + "epoch": 0.8876916701201824, + "grad_norm": 0.21891717612743378, + "learning_rate": 5.619560712805636e-07, + "loss": 0.4638, + "step": 10710 + }, + { + "epoch": 0.8877745544964775, + "grad_norm": 0.2507716715335846, + "learning_rate": 5.615416493990883e-07, + "loss": 0.5066, + "step": 10711 + }, + { + "epoch": 0.8878574388727725, + "grad_norm": 0.23220467567443848, + "learning_rate": 5.61127227517613e-07, + "loss": 0.4643, + "step": 10712 + }, + { + "epoch": 0.8879403232490676, + "grad_norm": 0.25273850560188293, + "learning_rate": 5.607128056361377e-07, + "loss": 0.5526, + "step": 10713 + }, + { + "epoch": 0.8880232076253626, + "grad_norm": 0.24481527507305145, + "learning_rate": 5.602983837546623e-07, + "loss": 0.5222, + "step": 10714 + }, + { + "epoch": 0.8881060920016577, + "grad_norm": 0.2511640787124634, + "learning_rate": 5.59883961873187e-07, + "loss": 0.4846, + "step": 10715 + }, + { + "epoch": 0.8881889763779528, + "grad_norm": 0.2533699572086334, + "learning_rate": 5.594695399917116e-07, + "loss": 0.5147, + "step": 10716 + }, + { + "epoch": 0.8882718607542478, + "grad_norm": 0.24780084192752838, + "learning_rate": 5.590551181102363e-07, + "loss": 0.5201, + "step": 10717 + }, + { + "epoch": 0.8883547451305429, + "grad_norm": 0.23405271768569946, + "learning_rate": 5.586406962287609e-07, + "loss": 0.485, + "step": 10718 + }, + { + "epoch": 0.8884376295068379, + "grad_norm": 0.25072386860847473, + "learning_rate": 5.582262743472855e-07, + "loss": 0.5131, + "step": 10719 + }, + { + "epoch": 0.888520513883133, + "grad_norm": 0.253900945186615, + "learning_rate": 5.578118524658102e-07, + "loss": 0.5468, + "step": 10720 + }, + { + "epoch": 0.8886033982594281, + "grad_norm": 0.22505991160869598, + "learning_rate": 5.573974305843349e-07, + "loss": 0.4694, + "step": 10721 + }, + { + "epoch": 0.8886862826357231, + "grad_norm": 0.25273397564888, + "learning_rate": 5.569830087028596e-07, + "loss": 0.4974, + "step": 10722 + }, + { + "epoch": 0.8887691670120182, + "grad_norm": 0.26308807730674744, + "learning_rate": 5.565685868213842e-07, + "loss": 0.5922, + "step": 10723 + }, + { + "epoch": 0.8888520513883134, + "grad_norm": 0.2420249879360199, + "learning_rate": 5.561541649399089e-07, + "loss": 0.4756, + "step": 10724 + }, + { + "epoch": 0.8889349357646084, + "grad_norm": 0.22396574914455414, + "learning_rate": 5.557397430584335e-07, + "loss": 0.4827, + "step": 10725 + }, + { + "epoch": 0.8890178201409035, + "grad_norm": 0.2471519559621811, + "learning_rate": 5.553253211769582e-07, + "loss": 0.4994, + "step": 10726 + }, + { + "epoch": 0.8891007045171985, + "grad_norm": 0.2598514258861542, + "learning_rate": 5.549108992954828e-07, + "loss": 0.5034, + "step": 10727 + }, + { + "epoch": 0.8891835888934936, + "grad_norm": 0.2427518516778946, + "learning_rate": 5.544964774140074e-07, + "loss": 0.5002, + "step": 10728 + }, + { + "epoch": 0.8892664732697887, + "grad_norm": 0.2363189309835434, + "learning_rate": 5.540820555325321e-07, + "loss": 0.5059, + "step": 10729 + }, + { + "epoch": 0.8893493576460837, + "grad_norm": 0.2437649518251419, + "learning_rate": 5.536676336510568e-07, + "loss": 0.4914, + "step": 10730 + }, + { + "epoch": 0.8894322420223788, + "grad_norm": 0.25876322388648987, + "learning_rate": 5.532532117695815e-07, + "loss": 0.4951, + "step": 10731 + }, + { + "epoch": 0.8895151263986738, + "grad_norm": 0.23620498180389404, + "learning_rate": 5.528387898881061e-07, + "loss": 0.5067, + "step": 10732 + }, + { + "epoch": 0.8895980107749689, + "grad_norm": 0.24749110639095306, + "learning_rate": 5.524243680066308e-07, + "loss": 0.4627, + "step": 10733 + }, + { + "epoch": 0.889680895151264, + "grad_norm": 0.23722179234027863, + "learning_rate": 5.520099461251554e-07, + "loss": 0.4827, + "step": 10734 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 0.24744519591331482, + "learning_rate": 5.515955242436802e-07, + "loss": 0.4938, + "step": 10735 + }, + { + "epoch": 0.8898466639038541, + "grad_norm": 0.2359021008014679, + "learning_rate": 5.511811023622048e-07, + "loss": 0.4986, + "step": 10736 + }, + { + "epoch": 0.8899295482801491, + "grad_norm": 0.23165254294872284, + "learning_rate": 5.507666804807294e-07, + "loss": 0.448, + "step": 10737 + }, + { + "epoch": 0.8900124326564443, + "grad_norm": 0.21194201707839966, + "learning_rate": 5.503522585992541e-07, + "loss": 0.4267, + "step": 10738 + }, + { + "epoch": 0.8900953170327394, + "grad_norm": 0.2405976951122284, + "learning_rate": 5.499378367177787e-07, + "loss": 0.4831, + "step": 10739 + }, + { + "epoch": 0.8901782014090344, + "grad_norm": 0.26871412992477417, + "learning_rate": 5.495234148363035e-07, + "loss": 0.5245, + "step": 10740 + }, + { + "epoch": 0.8902610857853295, + "grad_norm": 0.2314576357603073, + "learning_rate": 5.491089929548281e-07, + "loss": 0.4895, + "step": 10741 + }, + { + "epoch": 0.8903439701616246, + "grad_norm": 0.230809286236763, + "learning_rate": 5.486945710733528e-07, + "loss": 0.493, + "step": 10742 + }, + { + "epoch": 0.8904268545379196, + "grad_norm": 0.24209772050380707, + "learning_rate": 5.482801491918774e-07, + "loss": 0.4699, + "step": 10743 + }, + { + "epoch": 0.8905097389142147, + "grad_norm": 0.23218651115894318, + "learning_rate": 5.478657273104021e-07, + "loss": 0.4718, + "step": 10744 + }, + { + "epoch": 0.8905926232905097, + "grad_norm": 0.23119528591632843, + "learning_rate": 5.474513054289267e-07, + "loss": 0.4614, + "step": 10745 + }, + { + "epoch": 0.8906755076668048, + "grad_norm": 0.23331212997436523, + "learning_rate": 5.470368835474513e-07, + "loss": 0.4517, + "step": 10746 + }, + { + "epoch": 0.8907583920430999, + "grad_norm": 0.2514331638813019, + "learning_rate": 5.46622461665976e-07, + "loss": 0.5283, + "step": 10747 + }, + { + "epoch": 0.8908412764193949, + "grad_norm": 0.2337503731250763, + "learning_rate": 5.462080397845006e-07, + "loss": 0.4839, + "step": 10748 + }, + { + "epoch": 0.89092416079569, + "grad_norm": 0.23835307359695435, + "learning_rate": 5.457936179030253e-07, + "loss": 0.4968, + "step": 10749 + }, + { + "epoch": 0.891007045171985, + "grad_norm": 0.22773876786231995, + "learning_rate": 5.4537919602155e-07, + "loss": 0.492, + "step": 10750 + }, + { + "epoch": 0.8910899295482801, + "grad_norm": 0.2335827797651291, + "learning_rate": 5.449647741400747e-07, + "loss": 0.4795, + "step": 10751 + }, + { + "epoch": 0.8911728139245753, + "grad_norm": 0.2540830075740814, + "learning_rate": 5.445503522585993e-07, + "loss": 0.4811, + "step": 10752 + }, + { + "epoch": 0.8912556983008703, + "grad_norm": 0.2422071248292923, + "learning_rate": 5.44135930377124e-07, + "loss": 0.5204, + "step": 10753 + }, + { + "epoch": 0.8913385826771654, + "grad_norm": 0.23971301317214966, + "learning_rate": 5.437215084956486e-07, + "loss": 0.4756, + "step": 10754 + }, + { + "epoch": 0.8914214670534604, + "grad_norm": 0.23819608986377716, + "learning_rate": 5.433070866141732e-07, + "loss": 0.4876, + "step": 10755 + }, + { + "epoch": 0.8915043514297555, + "grad_norm": 0.23742875456809998, + "learning_rate": 5.428926647326979e-07, + "loss": 0.4927, + "step": 10756 + }, + { + "epoch": 0.8915872358060506, + "grad_norm": 0.23962098360061646, + "learning_rate": 5.424782428512225e-07, + "loss": 0.4863, + "step": 10757 + }, + { + "epoch": 0.8916701201823456, + "grad_norm": 0.2715975344181061, + "learning_rate": 5.420638209697472e-07, + "loss": 0.4901, + "step": 10758 + }, + { + "epoch": 0.8917530045586407, + "grad_norm": 0.22676469385623932, + "learning_rate": 5.416493990882719e-07, + "loss": 0.4477, + "step": 10759 + }, + { + "epoch": 0.8918358889349358, + "grad_norm": 0.22708937525749207, + "learning_rate": 5.412349772067966e-07, + "loss": 0.5121, + "step": 10760 + }, + { + "epoch": 0.8919187733112308, + "grad_norm": 0.23648354411125183, + "learning_rate": 5.408205553253212e-07, + "loss": 0.459, + "step": 10761 + }, + { + "epoch": 0.8920016576875259, + "grad_norm": 0.2350785732269287, + "learning_rate": 5.404061334438459e-07, + "loss": 0.4878, + "step": 10762 + }, + { + "epoch": 0.8920845420638209, + "grad_norm": 0.25399836897850037, + "learning_rate": 5.399917115623705e-07, + "loss": 0.5023, + "step": 10763 + }, + { + "epoch": 0.892167426440116, + "grad_norm": 0.24812524020671844, + "learning_rate": 5.395772896808951e-07, + "loss": 0.4823, + "step": 10764 + }, + { + "epoch": 0.8922503108164112, + "grad_norm": 0.24020975828170776, + "learning_rate": 5.391628677994199e-07, + "loss": 0.4818, + "step": 10765 + }, + { + "epoch": 0.8923331951927062, + "grad_norm": 0.21812188625335693, + "learning_rate": 5.387484459179445e-07, + "loss": 0.4675, + "step": 10766 + }, + { + "epoch": 0.8924160795690013, + "grad_norm": 0.2424517124891281, + "learning_rate": 5.383340240364692e-07, + "loss": 0.5049, + "step": 10767 + }, + { + "epoch": 0.8924989639452963, + "grad_norm": 0.25410300493240356, + "learning_rate": 5.379196021549938e-07, + "loss": 0.545, + "step": 10768 + }, + { + "epoch": 0.8925818483215914, + "grad_norm": 0.2443411946296692, + "learning_rate": 5.375051802735185e-07, + "loss": 0.5095, + "step": 10769 + }, + { + "epoch": 0.8926647326978865, + "grad_norm": 0.255367636680603, + "learning_rate": 5.370907583920432e-07, + "loss": 0.4773, + "step": 10770 + }, + { + "epoch": 0.8927476170741815, + "grad_norm": 0.23145882785320282, + "learning_rate": 5.366763365105679e-07, + "loss": 0.455, + "step": 10771 + }, + { + "epoch": 0.8928305014504766, + "grad_norm": 0.23535621166229248, + "learning_rate": 5.362619146290925e-07, + "loss": 0.4584, + "step": 10772 + }, + { + "epoch": 0.8929133858267716, + "grad_norm": 0.24614159762859344, + "learning_rate": 5.358474927476171e-07, + "loss": 0.5171, + "step": 10773 + }, + { + "epoch": 0.8929962702030667, + "grad_norm": 0.22960823774337769, + "learning_rate": 5.354330708661418e-07, + "loss": 0.4467, + "step": 10774 + }, + { + "epoch": 0.8930791545793618, + "grad_norm": 0.23340818285942078, + "learning_rate": 5.350186489846664e-07, + "loss": 0.4528, + "step": 10775 + }, + { + "epoch": 0.8931620389556568, + "grad_norm": 0.24899651110172272, + "learning_rate": 5.346042271031911e-07, + "loss": 0.4724, + "step": 10776 + }, + { + "epoch": 0.8932449233319519, + "grad_norm": 0.2282993197441101, + "learning_rate": 5.341898052217157e-07, + "loss": 0.4742, + "step": 10777 + }, + { + "epoch": 0.893327807708247, + "grad_norm": 0.2402859926223755, + "learning_rate": 5.337753833402405e-07, + "loss": 0.4999, + "step": 10778 + }, + { + "epoch": 0.893410692084542, + "grad_norm": 0.2345011979341507, + "learning_rate": 5.333609614587651e-07, + "loss": 0.5054, + "step": 10779 + }, + { + "epoch": 0.8934935764608372, + "grad_norm": 0.23628173768520355, + "learning_rate": 5.329465395772898e-07, + "loss": 0.4829, + "step": 10780 + }, + { + "epoch": 0.8935764608371322, + "grad_norm": 0.23907588422298431, + "learning_rate": 5.325321176958144e-07, + "loss": 0.512, + "step": 10781 + }, + { + "epoch": 0.8936593452134273, + "grad_norm": 0.24183708429336548, + "learning_rate": 5.32117695814339e-07, + "loss": 0.4871, + "step": 10782 + }, + { + "epoch": 0.8937422295897224, + "grad_norm": 0.25206008553504944, + "learning_rate": 5.317032739328637e-07, + "loss": 0.5499, + "step": 10783 + }, + { + "epoch": 0.8938251139660174, + "grad_norm": 0.2599303424358368, + "learning_rate": 5.312888520513883e-07, + "loss": 0.4942, + "step": 10784 + }, + { + "epoch": 0.8939079983423125, + "grad_norm": 0.23435001075267792, + "learning_rate": 5.30874430169913e-07, + "loss": 0.5125, + "step": 10785 + }, + { + "epoch": 0.8939908827186075, + "grad_norm": 0.22439377009868622, + "learning_rate": 5.304600082884376e-07, + "loss": 0.4404, + "step": 10786 + }, + { + "epoch": 0.8940737670949026, + "grad_norm": 0.24961915612220764, + "learning_rate": 5.300455864069623e-07, + "loss": 0.5424, + "step": 10787 + }, + { + "epoch": 0.8941566514711977, + "grad_norm": 0.2439393550157547, + "learning_rate": 5.29631164525487e-07, + "loss": 0.5251, + "step": 10788 + }, + { + "epoch": 0.8942395358474927, + "grad_norm": 0.22673293948173523, + "learning_rate": 5.292167426440117e-07, + "loss": 0.4781, + "step": 10789 + }, + { + "epoch": 0.8943224202237878, + "grad_norm": 0.23729918897151947, + "learning_rate": 5.288023207625363e-07, + "loss": 0.493, + "step": 10790 + }, + { + "epoch": 0.8944053046000828, + "grad_norm": 0.22990156710147858, + "learning_rate": 5.283878988810609e-07, + "loss": 0.4741, + "step": 10791 + }, + { + "epoch": 0.8944881889763779, + "grad_norm": 0.23511724174022675, + "learning_rate": 5.279734769995856e-07, + "loss": 0.4966, + "step": 10792 + }, + { + "epoch": 0.894571073352673, + "grad_norm": 0.23322124779224396, + "learning_rate": 5.275590551181102e-07, + "loss": 0.4882, + "step": 10793 + }, + { + "epoch": 0.894653957728968, + "grad_norm": 0.24519787728786469, + "learning_rate": 5.271446332366349e-07, + "loss": 0.4731, + "step": 10794 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.26534736156463623, + "learning_rate": 5.267302113551596e-07, + "loss": 0.5115, + "step": 10795 + }, + { + "epoch": 0.8948197264815583, + "grad_norm": 0.2646810710430145, + "learning_rate": 5.263157894736843e-07, + "loss": 0.5434, + "step": 10796 + }, + { + "epoch": 0.8949026108578533, + "grad_norm": 0.24568018317222595, + "learning_rate": 5.259013675922089e-07, + "loss": 0.463, + "step": 10797 + }, + { + "epoch": 0.8949854952341484, + "grad_norm": 0.21836183965206146, + "learning_rate": 5.254869457107336e-07, + "loss": 0.4672, + "step": 10798 + }, + { + "epoch": 0.8950683796104434, + "grad_norm": 0.24072711169719696, + "learning_rate": 5.250725238292582e-07, + "loss": 0.4943, + "step": 10799 + }, + { + "epoch": 0.8951512639867385, + "grad_norm": 0.23408745229244232, + "learning_rate": 5.246581019477829e-07, + "loss": 0.5076, + "step": 10800 + }, + { + "epoch": 0.8952341483630336, + "grad_norm": 0.2278391718864441, + "learning_rate": 5.242436800663076e-07, + "loss": 0.4602, + "step": 10801 + }, + { + "epoch": 0.8953170327393286, + "grad_norm": 0.24560204148292542, + "learning_rate": 5.238292581848322e-07, + "loss": 0.4496, + "step": 10802 + }, + { + "epoch": 0.8953999171156237, + "grad_norm": 0.24233335256576538, + "learning_rate": 5.234148363033569e-07, + "loss": 0.5358, + "step": 10803 + }, + { + "epoch": 0.8954828014919187, + "grad_norm": 0.23423293232917786, + "learning_rate": 5.230004144218815e-07, + "loss": 0.4962, + "step": 10804 + }, + { + "epoch": 0.8955656858682138, + "grad_norm": 0.23673170804977417, + "learning_rate": 5.225859925404062e-07, + "loss": 0.4646, + "step": 10805 + }, + { + "epoch": 0.895648570244509, + "grad_norm": 0.21675118803977966, + "learning_rate": 5.221715706589308e-07, + "loss": 0.4772, + "step": 10806 + }, + { + "epoch": 0.895731454620804, + "grad_norm": 0.2513522803783417, + "learning_rate": 5.217571487774556e-07, + "loss": 0.5172, + "step": 10807 + }, + { + "epoch": 0.8958143389970991, + "grad_norm": 0.30200567841529846, + "learning_rate": 5.213427268959802e-07, + "loss": 0.5041, + "step": 10808 + }, + { + "epoch": 0.8958972233733942, + "grad_norm": 0.2397511899471283, + "learning_rate": 5.209283050145048e-07, + "loss": 0.4962, + "step": 10809 + }, + { + "epoch": 0.8959801077496892, + "grad_norm": 0.24056436121463776, + "learning_rate": 5.205138831330295e-07, + "loss": 0.5112, + "step": 10810 + }, + { + "epoch": 0.8960629921259843, + "grad_norm": 0.2447928637266159, + "learning_rate": 5.200994612515541e-07, + "loss": 0.5142, + "step": 10811 + }, + { + "epoch": 0.8961458765022793, + "grad_norm": 0.2574746608734131, + "learning_rate": 5.196850393700788e-07, + "loss": 0.512, + "step": 10812 + }, + { + "epoch": 0.8962287608785744, + "grad_norm": 0.24214062094688416, + "learning_rate": 5.192706174886034e-07, + "loss": 0.4959, + "step": 10813 + }, + { + "epoch": 0.8963116452548695, + "grad_norm": 0.24790170788764954, + "learning_rate": 5.188561956071281e-07, + "loss": 0.4505, + "step": 10814 + }, + { + "epoch": 0.8963945296311645, + "grad_norm": 0.23655910789966583, + "learning_rate": 5.184417737256527e-07, + "loss": 0.4794, + "step": 10815 + }, + { + "epoch": 0.8964774140074596, + "grad_norm": 0.23315152525901794, + "learning_rate": 5.180273518441775e-07, + "loss": 0.4828, + "step": 10816 + }, + { + "epoch": 0.8965602983837546, + "grad_norm": 0.24725022912025452, + "learning_rate": 5.176129299627021e-07, + "loss": 0.5109, + "step": 10817 + }, + { + "epoch": 0.8966431827600497, + "grad_norm": 0.23293758928775787, + "learning_rate": 5.171985080812267e-07, + "loss": 0.4614, + "step": 10818 + }, + { + "epoch": 0.8967260671363448, + "grad_norm": 0.23191820085048676, + "learning_rate": 5.167840861997514e-07, + "loss": 0.4615, + "step": 10819 + }, + { + "epoch": 0.8968089515126398, + "grad_norm": 0.24501320719718933, + "learning_rate": 5.16369664318276e-07, + "loss": 0.5086, + "step": 10820 + }, + { + "epoch": 0.896891835888935, + "grad_norm": 0.234775573015213, + "learning_rate": 5.159552424368007e-07, + "loss": 0.5098, + "step": 10821 + }, + { + "epoch": 0.89697472026523, + "grad_norm": 0.23653241991996765, + "learning_rate": 5.155408205553253e-07, + "loss": 0.4836, + "step": 10822 + }, + { + "epoch": 0.8970576046415251, + "grad_norm": 0.2227669656276703, + "learning_rate": 5.1512639867385e-07, + "loss": 0.4649, + "step": 10823 + }, + { + "epoch": 0.8971404890178202, + "grad_norm": 0.24491889774799347, + "learning_rate": 5.147119767923746e-07, + "loss": 0.4968, + "step": 10824 + }, + { + "epoch": 0.8972233733941152, + "grad_norm": 0.2625403106212616, + "learning_rate": 5.142975549108994e-07, + "loss": 0.5262, + "step": 10825 + }, + { + "epoch": 0.8973062577704103, + "grad_norm": 0.22527523338794708, + "learning_rate": 5.13883133029424e-07, + "loss": 0.4528, + "step": 10826 + }, + { + "epoch": 0.8973891421467054, + "grad_norm": 0.23062975704669952, + "learning_rate": 5.134687111479486e-07, + "loss": 0.4865, + "step": 10827 + }, + { + "epoch": 0.8974720265230004, + "grad_norm": 0.23366010189056396, + "learning_rate": 5.130542892664733e-07, + "loss": 0.507, + "step": 10828 + }, + { + "epoch": 0.8975549108992955, + "grad_norm": 0.2549523413181305, + "learning_rate": 5.126398673849979e-07, + "loss": 0.5474, + "step": 10829 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 0.24089917540550232, + "learning_rate": 5.122254455035227e-07, + "loss": 0.5063, + "step": 10830 + }, + { + "epoch": 0.8977206796518856, + "grad_norm": 0.23597843945026398, + "learning_rate": 5.118110236220473e-07, + "loss": 0.4918, + "step": 10831 + }, + { + "epoch": 0.8978035640281807, + "grad_norm": 0.23485201597213745, + "learning_rate": 5.11396601740572e-07, + "loss": 0.4751, + "step": 10832 + }, + { + "epoch": 0.8978864484044757, + "grad_norm": 0.25433650612831116, + "learning_rate": 5.109821798590966e-07, + "loss": 0.4839, + "step": 10833 + }, + { + "epoch": 0.8979693327807708, + "grad_norm": 0.23527461290359497, + "learning_rate": 5.105677579776212e-07, + "loss": 0.4958, + "step": 10834 + }, + { + "epoch": 0.8980522171570658, + "grad_norm": 0.2340749204158783, + "learning_rate": 5.10153336096146e-07, + "loss": 0.5158, + "step": 10835 + }, + { + "epoch": 0.898135101533361, + "grad_norm": 0.22979024052619934, + "learning_rate": 5.097389142146706e-07, + "loss": 0.4867, + "step": 10836 + }, + { + "epoch": 0.8982179859096561, + "grad_norm": 0.2536781430244446, + "learning_rate": 5.093244923331953e-07, + "loss": 0.5203, + "step": 10837 + }, + { + "epoch": 0.8983008702859511, + "grad_norm": 0.22414061427116394, + "learning_rate": 5.089100704517199e-07, + "loss": 0.4581, + "step": 10838 + }, + { + "epoch": 0.8983837546622462, + "grad_norm": 0.25290578603744507, + "learning_rate": 5.084956485702446e-07, + "loss": 0.5272, + "step": 10839 + }, + { + "epoch": 0.8984666390385412, + "grad_norm": 0.22891680896282196, + "learning_rate": 5.080812266887692e-07, + "loss": 0.4787, + "step": 10840 + }, + { + "epoch": 0.8985495234148363, + "grad_norm": 0.25564810633659363, + "learning_rate": 5.076668048072939e-07, + "loss": 0.5131, + "step": 10841 + }, + { + "epoch": 0.8986324077911314, + "grad_norm": 0.22934098541736603, + "learning_rate": 5.072523829258185e-07, + "loss": 0.4468, + "step": 10842 + }, + { + "epoch": 0.8987152921674264, + "grad_norm": 0.26213544607162476, + "learning_rate": 5.068379610443431e-07, + "loss": 0.555, + "step": 10843 + }, + { + "epoch": 0.8987981765437215, + "grad_norm": 0.2526857852935791, + "learning_rate": 5.064235391628678e-07, + "loss": 0.4982, + "step": 10844 + }, + { + "epoch": 0.8988810609200166, + "grad_norm": 0.24173378944396973, + "learning_rate": 5.060091172813925e-07, + "loss": 0.5289, + "step": 10845 + }, + { + "epoch": 0.8989639452963116, + "grad_norm": 0.24359142780303955, + "learning_rate": 5.055946953999172e-07, + "loss": 0.5228, + "step": 10846 + }, + { + "epoch": 0.8990468296726067, + "grad_norm": 0.24199385941028595, + "learning_rate": 5.051802735184418e-07, + "loss": 0.5195, + "step": 10847 + }, + { + "epoch": 0.8991297140489017, + "grad_norm": 0.24004870653152466, + "learning_rate": 5.047658516369665e-07, + "loss": 0.4673, + "step": 10848 + }, + { + "epoch": 0.8992125984251969, + "grad_norm": 0.24023838341236115, + "learning_rate": 5.043514297554911e-07, + "loss": 0.4951, + "step": 10849 + }, + { + "epoch": 0.899295482801492, + "grad_norm": 0.22772784531116486, + "learning_rate": 5.039370078740158e-07, + "loss": 0.4807, + "step": 10850 + }, + { + "epoch": 0.899378367177787, + "grad_norm": 0.23504990339279175, + "learning_rate": 5.035225859925404e-07, + "loss": 0.4743, + "step": 10851 + }, + { + "epoch": 0.8994612515540821, + "grad_norm": 0.24009297788143158, + "learning_rate": 5.03108164111065e-07, + "loss": 0.5014, + "step": 10852 + }, + { + "epoch": 0.8995441359303771, + "grad_norm": 0.2637585997581482, + "learning_rate": 5.026937422295897e-07, + "loss": 0.5311, + "step": 10853 + }, + { + "epoch": 0.8996270203066722, + "grad_norm": 0.24932758510112762, + "learning_rate": 5.022793203481144e-07, + "loss": 0.4953, + "step": 10854 + }, + { + "epoch": 0.8997099046829673, + "grad_norm": 0.24345025420188904, + "learning_rate": 5.018648984666391e-07, + "loss": 0.4712, + "step": 10855 + }, + { + "epoch": 0.8997927890592623, + "grad_norm": 0.2894948124885559, + "learning_rate": 5.014504765851637e-07, + "loss": 0.5091, + "step": 10856 + }, + { + "epoch": 0.8998756734355574, + "grad_norm": 0.23045775294303894, + "learning_rate": 5.010360547036884e-07, + "loss": 0.4872, + "step": 10857 + }, + { + "epoch": 0.8999585578118524, + "grad_norm": 0.23909680545330048, + "learning_rate": 5.00621632822213e-07, + "loss": 0.4843, + "step": 10858 + }, + { + "epoch": 0.9000414421881475, + "grad_norm": 0.24288761615753174, + "learning_rate": 5.002072109407378e-07, + "loss": 0.4656, + "step": 10859 + }, + { + "epoch": 0.9001243265644426, + "grad_norm": 0.23338380455970764, + "learning_rate": 4.997927890592624e-07, + "loss": 0.4406, + "step": 10860 + }, + { + "epoch": 0.9002072109407376, + "grad_norm": 0.24005860090255737, + "learning_rate": 4.99378367177787e-07, + "loss": 0.5091, + "step": 10861 + }, + { + "epoch": 0.9002900953170327, + "grad_norm": 0.23955252766609192, + "learning_rate": 4.989639452963117e-07, + "loss": 0.4829, + "step": 10862 + }, + { + "epoch": 0.9003729796933279, + "grad_norm": 0.22985517978668213, + "learning_rate": 4.985495234148363e-07, + "loss": 0.4794, + "step": 10863 + }, + { + "epoch": 0.9004558640696229, + "grad_norm": 0.2458392232656479, + "learning_rate": 4.98135101533361e-07, + "loss": 0.4997, + "step": 10864 + }, + { + "epoch": 0.900538748445918, + "grad_norm": 0.2416633516550064, + "learning_rate": 4.977206796518857e-07, + "loss": 0.4893, + "step": 10865 + }, + { + "epoch": 0.900621632822213, + "grad_norm": 0.2569166123867035, + "learning_rate": 4.973062577704104e-07, + "loss": 0.5122, + "step": 10866 + }, + { + "epoch": 0.9007045171985081, + "grad_norm": 0.21681518852710724, + "learning_rate": 4.96891835888935e-07, + "loss": 0.4597, + "step": 10867 + }, + { + "epoch": 0.9007874015748032, + "grad_norm": 0.23014265298843384, + "learning_rate": 4.964774140074597e-07, + "loss": 0.4575, + "step": 10868 + }, + { + "epoch": 0.9008702859510982, + "grad_norm": 0.35238176584243774, + "learning_rate": 4.960629921259843e-07, + "loss": 0.5004, + "step": 10869 + }, + { + "epoch": 0.9009531703273933, + "grad_norm": 0.312906950712204, + "learning_rate": 4.956485702445089e-07, + "loss": 0.4767, + "step": 10870 + }, + { + "epoch": 0.9010360547036883, + "grad_norm": 0.23805195093154907, + "learning_rate": 4.952341483630336e-07, + "loss": 0.4744, + "step": 10871 + }, + { + "epoch": 0.9011189390799834, + "grad_norm": 0.22713008522987366, + "learning_rate": 4.948197264815582e-07, + "loss": 0.4817, + "step": 10872 + }, + { + "epoch": 0.9012018234562785, + "grad_norm": 0.2336534559726715, + "learning_rate": 4.94405304600083e-07, + "loss": 0.4735, + "step": 10873 + }, + { + "epoch": 0.9012847078325735, + "grad_norm": 0.24436140060424805, + "learning_rate": 4.939908827186076e-07, + "loss": 0.515, + "step": 10874 + }, + { + "epoch": 0.9013675922088686, + "grad_norm": 0.22647570073604584, + "learning_rate": 4.935764608371323e-07, + "loss": 0.487, + "step": 10875 + }, + { + "epoch": 0.9014504765851638, + "grad_norm": 0.21725626289844513, + "learning_rate": 4.931620389556569e-07, + "loss": 0.4511, + "step": 10876 + }, + { + "epoch": 0.9015333609614588, + "grad_norm": 0.21905630826950073, + "learning_rate": 4.927476170741816e-07, + "loss": 0.4615, + "step": 10877 + }, + { + "epoch": 0.9016162453377539, + "grad_norm": 0.23917719721794128, + "learning_rate": 4.923331951927062e-07, + "loss": 0.4865, + "step": 10878 + }, + { + "epoch": 0.9016991297140489, + "grad_norm": 0.2403797209262848, + "learning_rate": 4.919187733112308e-07, + "loss": 0.5032, + "step": 10879 + }, + { + "epoch": 0.901782014090344, + "grad_norm": 0.22361236810684204, + "learning_rate": 4.915043514297555e-07, + "loss": 0.4858, + "step": 10880 + }, + { + "epoch": 0.9018648984666391, + "grad_norm": 0.23185405135154724, + "learning_rate": 4.910899295482801e-07, + "loss": 0.4805, + "step": 10881 + }, + { + "epoch": 0.9019477828429341, + "grad_norm": 0.30702486634254456, + "learning_rate": 4.906755076668048e-07, + "loss": 0.488, + "step": 10882 + }, + { + "epoch": 0.9020306672192292, + "grad_norm": 0.24255244433879852, + "learning_rate": 4.902610857853295e-07, + "loss": 0.5109, + "step": 10883 + }, + { + "epoch": 0.9021135515955242, + "grad_norm": 0.2377278357744217, + "learning_rate": 4.898466639038542e-07, + "loss": 0.4803, + "step": 10884 + }, + { + "epoch": 0.9021964359718193, + "grad_norm": 0.22743555903434753, + "learning_rate": 4.894322420223788e-07, + "loss": 0.4707, + "step": 10885 + }, + { + "epoch": 0.9022793203481144, + "grad_norm": 0.22784222662448883, + "learning_rate": 4.890178201409035e-07, + "loss": 0.4612, + "step": 10886 + }, + { + "epoch": 0.9023622047244094, + "grad_norm": 0.24301831424236298, + "learning_rate": 4.886033982594281e-07, + "loss": 0.4919, + "step": 10887 + }, + { + "epoch": 0.9024450891007045, + "grad_norm": 0.24771247804164886, + "learning_rate": 4.881889763779527e-07, + "loss": 0.5344, + "step": 10888 + }, + { + "epoch": 0.9025279734769995, + "grad_norm": 0.2323460578918457, + "learning_rate": 4.877745544964775e-07, + "loss": 0.4632, + "step": 10889 + }, + { + "epoch": 0.9026108578532946, + "grad_norm": 0.25613266229629517, + "learning_rate": 4.873601326150021e-07, + "loss": 0.5516, + "step": 10890 + }, + { + "epoch": 0.9026937422295898, + "grad_norm": 0.23125258088111877, + "learning_rate": 4.869457107335268e-07, + "loss": 0.4636, + "step": 10891 + }, + { + "epoch": 0.9027766266058848, + "grad_norm": 0.22329364717006683, + "learning_rate": 4.865312888520514e-07, + "loss": 0.4574, + "step": 10892 + }, + { + "epoch": 0.9028595109821799, + "grad_norm": 0.23252147436141968, + "learning_rate": 4.861168669705761e-07, + "loss": 0.4942, + "step": 10893 + }, + { + "epoch": 0.902942395358475, + "grad_norm": 0.23799176514148712, + "learning_rate": 4.857024450891008e-07, + "loss": 0.4683, + "step": 10894 + }, + { + "epoch": 0.90302527973477, + "grad_norm": 0.2503929138183594, + "learning_rate": 4.852880232076255e-07, + "loss": 0.5499, + "step": 10895 + }, + { + "epoch": 0.9031081641110651, + "grad_norm": 0.23867660760879517, + "learning_rate": 4.848736013261501e-07, + "loss": 0.4821, + "step": 10896 + }, + { + "epoch": 0.9031910484873601, + "grad_norm": 0.2374333292245865, + "learning_rate": 4.844591794446747e-07, + "loss": 0.4679, + "step": 10897 + }, + { + "epoch": 0.9032739328636552, + "grad_norm": 0.2358364313840866, + "learning_rate": 4.840447575631994e-07, + "loss": 0.4896, + "step": 10898 + }, + { + "epoch": 0.9033568172399503, + "grad_norm": 0.23147575557231903, + "learning_rate": 4.83630335681724e-07, + "loss": 0.4353, + "step": 10899 + }, + { + "epoch": 0.9034397016162453, + "grad_norm": 0.24745018780231476, + "learning_rate": 4.832159138002487e-07, + "loss": 0.5532, + "step": 10900 + }, + { + "epoch": 0.9035225859925404, + "grad_norm": 0.23558852076530457, + "learning_rate": 4.828014919187733e-07, + "loss": 0.4485, + "step": 10901 + }, + { + "epoch": 0.9036054703688354, + "grad_norm": 0.25235700607299805, + "learning_rate": 4.823870700372981e-07, + "loss": 0.4998, + "step": 10902 + }, + { + "epoch": 0.9036883547451305, + "grad_norm": 0.21449601650238037, + "learning_rate": 4.819726481558227e-07, + "loss": 0.444, + "step": 10903 + }, + { + "epoch": 0.9037712391214257, + "grad_norm": 0.21530939638614655, + "learning_rate": 4.815582262743474e-07, + "loss": 0.4355, + "step": 10904 + }, + { + "epoch": 0.9038541234977207, + "grad_norm": 0.2329769730567932, + "learning_rate": 4.81143804392872e-07, + "loss": 0.4789, + "step": 10905 + }, + { + "epoch": 0.9039370078740158, + "grad_norm": 0.2400415539741516, + "learning_rate": 4.807293825113966e-07, + "loss": 0.4763, + "step": 10906 + }, + { + "epoch": 0.9040198922503108, + "grad_norm": 0.23377983272075653, + "learning_rate": 4.803149606299213e-07, + "loss": 0.4621, + "step": 10907 + }, + { + "epoch": 0.9041027766266059, + "grad_norm": 0.25369179248809814, + "learning_rate": 4.799005387484459e-07, + "loss": 0.535, + "step": 10908 + }, + { + "epoch": 0.904185661002901, + "grad_norm": 0.25172820687294006, + "learning_rate": 4.794861168669706e-07, + "loss": 0.5121, + "step": 10909 + }, + { + "epoch": 0.904268545379196, + "grad_norm": 0.2240159660577774, + "learning_rate": 4.790716949854952e-07, + "loss": 0.4791, + "step": 10910 + }, + { + "epoch": 0.9043514297554911, + "grad_norm": 0.242887943983078, + "learning_rate": 4.7865727310402e-07, + "loss": 0.512, + "step": 10911 + }, + { + "epoch": 0.9044343141317862, + "grad_norm": 0.24285130202770233, + "learning_rate": 4.782428512225446e-07, + "loss": 0.5424, + "step": 10912 + }, + { + "epoch": 0.9045171985080812, + "grad_norm": 0.23282217979431152, + "learning_rate": 4.778284293410693e-07, + "loss": 0.4765, + "step": 10913 + }, + { + "epoch": 0.9046000828843763, + "grad_norm": 0.22260735929012299, + "learning_rate": 4.774140074595939e-07, + "loss": 0.469, + "step": 10914 + }, + { + "epoch": 0.9046829672606713, + "grad_norm": 0.2281429022550583, + "learning_rate": 4.769995855781185e-07, + "loss": 0.4654, + "step": 10915 + }, + { + "epoch": 0.9047658516369664, + "grad_norm": 0.23574180901050568, + "learning_rate": 4.7658516369664324e-07, + "loss": 0.4807, + "step": 10916 + }, + { + "epoch": 0.9048487360132615, + "grad_norm": 0.23751553893089294, + "learning_rate": 4.761707418151679e-07, + "loss": 0.4334, + "step": 10917 + }, + { + "epoch": 0.9049316203895565, + "grad_norm": 0.23020526766777039, + "learning_rate": 4.757563199336925e-07, + "loss": 0.4869, + "step": 10918 + }, + { + "epoch": 0.9050145047658517, + "grad_norm": 0.24443848431110382, + "learning_rate": 4.7534189805221717e-07, + "loss": 0.5226, + "step": 10919 + }, + { + "epoch": 0.9050973891421467, + "grad_norm": 0.23569336533546448, + "learning_rate": 4.749274761707418e-07, + "loss": 0.5082, + "step": 10920 + }, + { + "epoch": 0.9051802735184418, + "grad_norm": 0.2447376549243927, + "learning_rate": 4.7451305428926656e-07, + "loss": 0.4571, + "step": 10921 + }, + { + "epoch": 0.9052631578947369, + "grad_norm": 0.2325635552406311, + "learning_rate": 4.740986324077912e-07, + "loss": 0.4728, + "step": 10922 + }, + { + "epoch": 0.9053460422710319, + "grad_norm": 0.2845239043235779, + "learning_rate": 4.7368421052631585e-07, + "loss": 0.5117, + "step": 10923 + }, + { + "epoch": 0.905428926647327, + "grad_norm": 0.23723797500133514, + "learning_rate": 4.732697886448405e-07, + "loss": 0.506, + "step": 10924 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 0.24770815670490265, + "learning_rate": 4.7285536676336514e-07, + "loss": 0.4387, + "step": 10925 + }, + { + "epoch": 0.9055946953999171, + "grad_norm": 0.23491592705249786, + "learning_rate": 4.724409448818898e-07, + "loss": 0.4892, + "step": 10926 + }, + { + "epoch": 0.9056775797762122, + "grad_norm": 0.22424089908599854, + "learning_rate": 4.720265230004144e-07, + "loss": 0.4686, + "step": 10927 + }, + { + "epoch": 0.9057604641525072, + "grad_norm": 0.22551612555980682, + "learning_rate": 4.716121011189391e-07, + "loss": 0.459, + "step": 10928 + }, + { + "epoch": 0.9058433485288023, + "grad_norm": 0.2312525361776352, + "learning_rate": 4.7119767923746376e-07, + "loss": 0.4399, + "step": 10929 + }, + { + "epoch": 0.9059262329050974, + "grad_norm": 0.23445436358451843, + "learning_rate": 4.7078325735598846e-07, + "loss": 0.5261, + "step": 10930 + }, + { + "epoch": 0.9060091172813924, + "grad_norm": 0.23796546459197998, + "learning_rate": 4.703688354745131e-07, + "loss": 0.5123, + "step": 10931 + }, + { + "epoch": 0.9060920016576876, + "grad_norm": 0.2683394253253937, + "learning_rate": 4.6995441359303775e-07, + "loss": 0.5052, + "step": 10932 + }, + { + "epoch": 0.9061748860339826, + "grad_norm": 0.24134576320648193, + "learning_rate": 4.6953999171156244e-07, + "loss": 0.4893, + "step": 10933 + }, + { + "epoch": 0.9062577704102777, + "grad_norm": 0.2336481809616089, + "learning_rate": 4.691255698300871e-07, + "loss": 0.4907, + "step": 10934 + }, + { + "epoch": 0.9063406547865728, + "grad_norm": 0.24559226632118225, + "learning_rate": 4.6871114794861173e-07, + "loss": 0.492, + "step": 10935 + }, + { + "epoch": 0.9064235391628678, + "grad_norm": 0.24145425856113434, + "learning_rate": 4.682967260671364e-07, + "loss": 0.5007, + "step": 10936 + }, + { + "epoch": 0.9065064235391629, + "grad_norm": 0.23306132853031158, + "learning_rate": 4.67882304185661e-07, + "loss": 0.4907, + "step": 10937 + }, + { + "epoch": 0.9065893079154579, + "grad_norm": 0.24706463515758514, + "learning_rate": 4.6746788230418566e-07, + "loss": 0.5018, + "step": 10938 + }, + { + "epoch": 0.906672192291753, + "grad_norm": 0.24187105894088745, + "learning_rate": 4.670534604227103e-07, + "loss": 0.5047, + "step": 10939 + }, + { + "epoch": 0.9067550766680481, + "grad_norm": 0.24156548082828522, + "learning_rate": 4.6663903854123505e-07, + "loss": 0.4713, + "step": 10940 + }, + { + "epoch": 0.9068379610443431, + "grad_norm": 0.24669930338859558, + "learning_rate": 4.662246166597597e-07, + "loss": 0.4901, + "step": 10941 + }, + { + "epoch": 0.9069208454206382, + "grad_norm": 0.2541485130786896, + "learning_rate": 4.6581019477828434e-07, + "loss": 0.498, + "step": 10942 + }, + { + "epoch": 0.9070037297969333, + "grad_norm": 0.24615710973739624, + "learning_rate": 4.65395772896809e-07, + "loss": 0.5298, + "step": 10943 + }, + { + "epoch": 0.9070866141732283, + "grad_norm": 0.22534741461277008, + "learning_rate": 4.6498135101533363e-07, + "loss": 0.4692, + "step": 10944 + }, + { + "epoch": 0.9071694985495234, + "grad_norm": 0.24131768941879272, + "learning_rate": 4.645669291338583e-07, + "loss": 0.5061, + "step": 10945 + }, + { + "epoch": 0.9072523829258184, + "grad_norm": 0.23704352974891663, + "learning_rate": 4.6415250725238297e-07, + "loss": 0.5087, + "step": 10946 + }, + { + "epoch": 0.9073352673021136, + "grad_norm": 0.2319975048303604, + "learning_rate": 4.637380853709076e-07, + "loss": 0.4707, + "step": 10947 + }, + { + "epoch": 0.9074181516784087, + "grad_norm": 0.23308846354484558, + "learning_rate": 4.6332366348943226e-07, + "loss": 0.5164, + "step": 10948 + }, + { + "epoch": 0.9075010360547037, + "grad_norm": 0.24756279587745667, + "learning_rate": 4.6290924160795695e-07, + "loss": 0.4695, + "step": 10949 + }, + { + "epoch": 0.9075839204309988, + "grad_norm": 0.2422063946723938, + "learning_rate": 4.6249481972648165e-07, + "loss": 0.5038, + "step": 10950 + }, + { + "epoch": 0.9076668048072938, + "grad_norm": 0.23598553240299225, + "learning_rate": 4.620803978450063e-07, + "loss": 0.4941, + "step": 10951 + }, + { + "epoch": 0.9077496891835889, + "grad_norm": 0.2230931520462036, + "learning_rate": 4.6166597596353094e-07, + "loss": 0.4551, + "step": 10952 + }, + { + "epoch": 0.907832573559884, + "grad_norm": 0.23124967515468597, + "learning_rate": 4.612515540820556e-07, + "loss": 0.4604, + "step": 10953 + }, + { + "epoch": 0.907915457936179, + "grad_norm": 0.2689432203769684, + "learning_rate": 4.608371322005802e-07, + "loss": 0.4726, + "step": 10954 + }, + { + "epoch": 0.9079983423124741, + "grad_norm": 0.24698667228221893, + "learning_rate": 4.6042271031910487e-07, + "loss": 0.4698, + "step": 10955 + }, + { + "epoch": 0.9080812266887691, + "grad_norm": 0.25274428725242615, + "learning_rate": 4.600082884376295e-07, + "loss": 0.5074, + "step": 10956 + }, + { + "epoch": 0.9081641110650642, + "grad_norm": 0.24628865718841553, + "learning_rate": 4.5959386655615415e-07, + "loss": 0.526, + "step": 10957 + }, + { + "epoch": 0.9082469954413593, + "grad_norm": 0.23419222235679626, + "learning_rate": 4.5917944467467885e-07, + "loss": 0.5271, + "step": 10958 + }, + { + "epoch": 0.9083298798176543, + "grad_norm": 0.23166537284851074, + "learning_rate": 4.5876502279320355e-07, + "loss": 0.5042, + "step": 10959 + }, + { + "epoch": 0.9084127641939495, + "grad_norm": 0.22999286651611328, + "learning_rate": 4.583506009117282e-07, + "loss": 0.4788, + "step": 10960 + }, + { + "epoch": 0.9084956485702446, + "grad_norm": 0.24300500750541687, + "learning_rate": 4.5793617903025283e-07, + "loss": 0.5304, + "step": 10961 + }, + { + "epoch": 0.9085785329465396, + "grad_norm": 0.23281770944595337, + "learning_rate": 4.575217571487775e-07, + "loss": 0.473, + "step": 10962 + }, + { + "epoch": 0.9086614173228347, + "grad_norm": 0.22614452242851257, + "learning_rate": 4.5710733526730217e-07, + "loss": 0.4911, + "step": 10963 + }, + { + "epoch": 0.9087443016991297, + "grad_norm": 0.27260512113571167, + "learning_rate": 4.566929133858268e-07, + "loss": 0.5222, + "step": 10964 + }, + { + "epoch": 0.9088271860754248, + "grad_norm": 0.24739179015159607, + "learning_rate": 4.5627849150435146e-07, + "loss": 0.4549, + "step": 10965 + }, + { + "epoch": 0.9089100704517199, + "grad_norm": 0.24995802342891693, + "learning_rate": 4.558640696228761e-07, + "loss": 0.5049, + "step": 10966 + }, + { + "epoch": 0.9089929548280149, + "grad_norm": 0.2352139800786972, + "learning_rate": 4.5544964774140075e-07, + "loss": 0.4921, + "step": 10967 + }, + { + "epoch": 0.90907583920431, + "grad_norm": 0.23694993555545807, + "learning_rate": 4.550352258599255e-07, + "loss": 0.474, + "step": 10968 + }, + { + "epoch": 0.909158723580605, + "grad_norm": 0.2307981252670288, + "learning_rate": 4.5462080397845014e-07, + "loss": 0.4823, + "step": 10969 + }, + { + "epoch": 0.9092416079569001, + "grad_norm": 0.2320312261581421, + "learning_rate": 4.542063820969748e-07, + "loss": 0.4814, + "step": 10970 + }, + { + "epoch": 0.9093244923331952, + "grad_norm": 0.23090991377830505, + "learning_rate": 4.5379196021549943e-07, + "loss": 0.4659, + "step": 10971 + }, + { + "epoch": 0.9094073767094902, + "grad_norm": 0.2295662760734558, + "learning_rate": 4.5337753833402407e-07, + "loss": 0.4791, + "step": 10972 + }, + { + "epoch": 0.9094902610857853, + "grad_norm": 0.22418063879013062, + "learning_rate": 4.529631164525487e-07, + "loss": 0.4915, + "step": 10973 + }, + { + "epoch": 0.9095731454620803, + "grad_norm": 0.23829056322574615, + "learning_rate": 4.5254869457107336e-07, + "loss": 0.4852, + "step": 10974 + }, + { + "epoch": 0.9096560298383755, + "grad_norm": 0.2685638964176178, + "learning_rate": 4.5213427268959805e-07, + "loss": 0.5179, + "step": 10975 + }, + { + "epoch": 0.9097389142146706, + "grad_norm": 0.24074482917785645, + "learning_rate": 4.517198508081227e-07, + "loss": 0.5108, + "step": 10976 + }, + { + "epoch": 0.9098217985909656, + "grad_norm": 0.2258349359035492, + "learning_rate": 4.5130542892664734e-07, + "loss": 0.4546, + "step": 10977 + }, + { + "epoch": 0.9099046829672607, + "grad_norm": 0.23570765554904938, + "learning_rate": 4.5089100704517204e-07, + "loss": 0.4562, + "step": 10978 + }, + { + "epoch": 0.9099875673435558, + "grad_norm": 0.25645822286605835, + "learning_rate": 4.504765851636967e-07, + "loss": 0.5044, + "step": 10979 + }, + { + "epoch": 0.9100704517198508, + "grad_norm": 0.2289789617061615, + "learning_rate": 4.500621632822214e-07, + "loss": 0.4832, + "step": 10980 + }, + { + "epoch": 0.9101533360961459, + "grad_norm": 0.23880693316459656, + "learning_rate": 4.49647741400746e-07, + "loss": 0.5285, + "step": 10981 + }, + { + "epoch": 0.9102362204724409, + "grad_norm": 0.26638928055763245, + "learning_rate": 4.4923331951927067e-07, + "loss": 0.5447, + "step": 10982 + }, + { + "epoch": 0.910319104848736, + "grad_norm": 0.2347477376461029, + "learning_rate": 4.488188976377953e-07, + "loss": 0.5043, + "step": 10983 + }, + { + "epoch": 0.9104019892250311, + "grad_norm": 0.2238667905330658, + "learning_rate": 4.4840447575631995e-07, + "loss": 0.5046, + "step": 10984 + }, + { + "epoch": 0.9104848736013261, + "grad_norm": 0.23763594031333923, + "learning_rate": 4.479900538748446e-07, + "loss": 0.4986, + "step": 10985 + }, + { + "epoch": 0.9105677579776212, + "grad_norm": 0.22650988399982452, + "learning_rate": 4.4757563199336924e-07, + "loss": 0.4603, + "step": 10986 + }, + { + "epoch": 0.9106506423539162, + "grad_norm": 0.23195183277130127, + "learning_rate": 4.47161210111894e-07, + "loss": 0.505, + "step": 10987 + }, + { + "epoch": 0.9107335267302114, + "grad_norm": 0.2393152415752411, + "learning_rate": 4.4674678823041863e-07, + "loss": 0.5029, + "step": 10988 + }, + { + "epoch": 0.9108164111065065, + "grad_norm": 0.22865426540374756, + "learning_rate": 4.463323663489433e-07, + "loss": 0.4591, + "step": 10989 + }, + { + "epoch": 0.9108992954828015, + "grad_norm": 0.23854056000709534, + "learning_rate": 4.459179444674679e-07, + "loss": 0.5222, + "step": 10990 + }, + { + "epoch": 0.9109821798590966, + "grad_norm": 0.2505294680595398, + "learning_rate": 4.4550352258599256e-07, + "loss": 0.5308, + "step": 10991 + }, + { + "epoch": 0.9110650642353916, + "grad_norm": 0.2547774910926819, + "learning_rate": 4.450891007045172e-07, + "loss": 0.5233, + "step": 10992 + }, + { + "epoch": 0.9111479486116867, + "grad_norm": 0.24363413453102112, + "learning_rate": 4.446746788230419e-07, + "loss": 0.5206, + "step": 10993 + }, + { + "epoch": 0.9112308329879818, + "grad_norm": 0.24616016447544098, + "learning_rate": 4.4426025694156655e-07, + "loss": 0.5115, + "step": 10994 + }, + { + "epoch": 0.9113137173642768, + "grad_norm": 0.24003401398658752, + "learning_rate": 4.438458350600912e-07, + "loss": 0.4651, + "step": 10995 + }, + { + "epoch": 0.9113966017405719, + "grad_norm": 0.25433316826820374, + "learning_rate": 4.4343141317861583e-07, + "loss": 0.5405, + "step": 10996 + }, + { + "epoch": 0.911479486116867, + "grad_norm": 0.22444547712802887, + "learning_rate": 4.4301699129714053e-07, + "loss": 0.4306, + "step": 10997 + }, + { + "epoch": 0.911562370493162, + "grad_norm": 0.2617519199848175, + "learning_rate": 4.4260256941566523e-07, + "loss": 0.5521, + "step": 10998 + }, + { + "epoch": 0.9116452548694571, + "grad_norm": 0.2521141767501831, + "learning_rate": 4.4218814753418987e-07, + "loss": 0.5393, + "step": 10999 + }, + { + "epoch": 0.9117281392457521, + "grad_norm": 0.24342943727970123, + "learning_rate": 4.417737256527145e-07, + "loss": 0.4941, + "step": 11000 + }, + { + "epoch": 0.9118110236220472, + "grad_norm": 0.24154704809188843, + "learning_rate": 4.4135930377123916e-07, + "loss": 0.5294, + "step": 11001 + }, + { + "epoch": 0.9118939079983424, + "grad_norm": 0.2391587644815445, + "learning_rate": 4.409448818897638e-07, + "loss": 0.4643, + "step": 11002 + }, + { + "epoch": 0.9119767923746374, + "grad_norm": 0.23846614360809326, + "learning_rate": 4.4053046000828844e-07, + "loss": 0.5023, + "step": 11003 + }, + { + "epoch": 0.9120596767509325, + "grad_norm": 0.23168134689331055, + "learning_rate": 4.401160381268131e-07, + "loss": 0.4571, + "step": 11004 + }, + { + "epoch": 0.9121425611272275, + "grad_norm": 0.22803866863250732, + "learning_rate": 4.397016162453378e-07, + "loss": 0.4661, + "step": 11005 + }, + { + "epoch": 0.9122254455035226, + "grad_norm": 0.23144495487213135, + "learning_rate": 4.392871943638625e-07, + "loss": 0.4646, + "step": 11006 + }, + { + "epoch": 0.9123083298798177, + "grad_norm": 0.23491816222667694, + "learning_rate": 4.388727724823871e-07, + "loss": 0.5025, + "step": 11007 + }, + { + "epoch": 0.9123912142561127, + "grad_norm": 0.23690174520015717, + "learning_rate": 4.3845835060091177e-07, + "loss": 0.4566, + "step": 11008 + }, + { + "epoch": 0.9124740986324078, + "grad_norm": 0.235037699341774, + "learning_rate": 4.380439287194364e-07, + "loss": 0.503, + "step": 11009 + }, + { + "epoch": 0.9125569830087028, + "grad_norm": 0.24966146051883698, + "learning_rate": 4.376295068379611e-07, + "loss": 0.4943, + "step": 11010 + }, + { + "epoch": 0.9126398673849979, + "grad_norm": 0.2492164522409439, + "learning_rate": 4.3721508495648575e-07, + "loss": 0.5071, + "step": 11011 + }, + { + "epoch": 0.912722751761293, + "grad_norm": 0.23997749388217926, + "learning_rate": 4.368006630750104e-07, + "loss": 0.48, + "step": 11012 + }, + { + "epoch": 0.912805636137588, + "grad_norm": 0.22804851830005646, + "learning_rate": 4.3638624119353504e-07, + "loss": 0.4419, + "step": 11013 + }, + { + "epoch": 0.9128885205138831, + "grad_norm": 0.22866804897785187, + "learning_rate": 4.359718193120597e-07, + "loss": 0.4661, + "step": 11014 + }, + { + "epoch": 0.9129714048901783, + "grad_norm": 0.23590469360351562, + "learning_rate": 4.355573974305843e-07, + "loss": 0.4886, + "step": 11015 + }, + { + "epoch": 0.9130542892664733, + "grad_norm": 0.23509038984775543, + "learning_rate": 4.351429755491091e-07, + "loss": 0.4911, + "step": 11016 + }, + { + "epoch": 0.9131371736427684, + "grad_norm": 0.2365441918373108, + "learning_rate": 4.347285536676337e-07, + "loss": 0.465, + "step": 11017 + }, + { + "epoch": 0.9132200580190634, + "grad_norm": 0.24740833044052124, + "learning_rate": 4.3431413178615836e-07, + "loss": 0.5232, + "step": 11018 + }, + { + "epoch": 0.9133029423953585, + "grad_norm": 0.23147925734519958, + "learning_rate": 4.33899709904683e-07, + "loss": 0.4457, + "step": 11019 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 0.2449006587266922, + "learning_rate": 4.3348528802320765e-07, + "loss": 0.5091, + "step": 11020 + }, + { + "epoch": 0.9134687111479486, + "grad_norm": 0.2730006277561188, + "learning_rate": 4.330708661417323e-07, + "loss": 0.5039, + "step": 11021 + }, + { + "epoch": 0.9135515955242437, + "grad_norm": 0.25047987699508667, + "learning_rate": 4.32656444260257e-07, + "loss": 0.4896, + "step": 11022 + }, + { + "epoch": 0.9136344799005387, + "grad_norm": 0.23884820938110352, + "learning_rate": 4.3224202237878163e-07, + "loss": 0.4977, + "step": 11023 + }, + { + "epoch": 0.9137173642768338, + "grad_norm": 0.22832998633384705, + "learning_rate": 4.318276004973063e-07, + "loss": 0.4705, + "step": 11024 + }, + { + "epoch": 0.9138002486531289, + "grad_norm": 0.22351710498332977, + "learning_rate": 4.31413178615831e-07, + "loss": 0.4924, + "step": 11025 + }, + { + "epoch": 0.9138831330294239, + "grad_norm": 0.24869027733802795, + "learning_rate": 4.309987567343556e-07, + "loss": 0.5168, + "step": 11026 + }, + { + "epoch": 0.913966017405719, + "grad_norm": 0.22954705357551575, + "learning_rate": 4.3058433485288026e-07, + "loss": 0.4582, + "step": 11027 + }, + { + "epoch": 0.9140489017820141, + "grad_norm": 0.21562515199184418, + "learning_rate": 4.3016991297140496e-07, + "loss": 0.4349, + "step": 11028 + }, + { + "epoch": 0.9141317861583091, + "grad_norm": 0.24853605031967163, + "learning_rate": 4.297554910899296e-07, + "loss": 0.5081, + "step": 11029 + }, + { + "epoch": 0.9142146705346043, + "grad_norm": 0.2353130280971527, + "learning_rate": 4.2934106920845424e-07, + "loss": 0.484, + "step": 11030 + }, + { + "epoch": 0.9142975549108993, + "grad_norm": 0.2302992343902588, + "learning_rate": 4.289266473269789e-07, + "loss": 0.497, + "step": 11031 + }, + { + "epoch": 0.9143804392871944, + "grad_norm": 0.2987944781780243, + "learning_rate": 4.2851222544550353e-07, + "loss": 0.5071, + "step": 11032 + }, + { + "epoch": 0.9144633236634895, + "grad_norm": 0.24360565841197968, + "learning_rate": 4.280978035640282e-07, + "loss": 0.5034, + "step": 11033 + }, + { + "epoch": 0.9145462080397845, + "grad_norm": 0.2484310269355774, + "learning_rate": 4.276833816825528e-07, + "loss": 0.5048, + "step": 11034 + }, + { + "epoch": 0.9146290924160796, + "grad_norm": 0.257424533367157, + "learning_rate": 4.2726895980107757e-07, + "loss": 0.4756, + "step": 11035 + }, + { + "epoch": 0.9147119767923746, + "grad_norm": 0.24065619707107544, + "learning_rate": 4.268545379196022e-07, + "loss": 0.4975, + "step": 11036 + }, + { + "epoch": 0.9147948611686697, + "grad_norm": 0.23235008120536804, + "learning_rate": 4.2644011603812686e-07, + "loss": 0.4734, + "step": 11037 + }, + { + "epoch": 0.9148777455449648, + "grad_norm": 0.24285493791103363, + "learning_rate": 4.260256941566515e-07, + "loss": 0.494, + "step": 11038 + }, + { + "epoch": 0.9149606299212598, + "grad_norm": 0.23833894729614258, + "learning_rate": 4.2561127227517614e-07, + "loss": 0.4791, + "step": 11039 + }, + { + "epoch": 0.9150435142975549, + "grad_norm": 0.2155248522758484, + "learning_rate": 4.2519685039370084e-07, + "loss": 0.4276, + "step": 11040 + }, + { + "epoch": 0.9151263986738499, + "grad_norm": 0.22739562392234802, + "learning_rate": 4.247824285122255e-07, + "loss": 0.4779, + "step": 11041 + }, + { + "epoch": 0.915209283050145, + "grad_norm": 0.22370420396327972, + "learning_rate": 4.243680066307501e-07, + "loss": 0.468, + "step": 11042 + }, + { + "epoch": 0.9152921674264402, + "grad_norm": 0.22527705132961273, + "learning_rate": 4.2395358474927477e-07, + "loss": 0.4819, + "step": 11043 + }, + { + "epoch": 0.9153750518027352, + "grad_norm": 0.23141930997371674, + "learning_rate": 4.2353916286779947e-07, + "loss": 0.4621, + "step": 11044 + }, + { + "epoch": 0.9154579361790303, + "grad_norm": 0.24166293442249298, + "learning_rate": 4.2312474098632416e-07, + "loss": 0.472, + "step": 11045 + }, + { + "epoch": 0.9155408205553254, + "grad_norm": 0.23996485769748688, + "learning_rate": 4.227103191048488e-07, + "loss": 0.4639, + "step": 11046 + }, + { + "epoch": 0.9156237049316204, + "grad_norm": 0.24313612282276154, + "learning_rate": 4.2229589722337345e-07, + "loss": 0.5161, + "step": 11047 + }, + { + "epoch": 0.9157065893079155, + "grad_norm": 0.23758000135421753, + "learning_rate": 4.218814753418981e-07, + "loss": 0.513, + "step": 11048 + }, + { + "epoch": 0.9157894736842105, + "grad_norm": 0.22149726748466492, + "learning_rate": 4.2146705346042274e-07, + "loss": 0.4646, + "step": 11049 + }, + { + "epoch": 0.9158723580605056, + "grad_norm": 0.24895307421684265, + "learning_rate": 4.210526315789474e-07, + "loss": 0.5148, + "step": 11050 + }, + { + "epoch": 0.9159552424368007, + "grad_norm": 0.23414193093776703, + "learning_rate": 4.20638209697472e-07, + "loss": 0.4995, + "step": 11051 + }, + { + "epoch": 0.9160381268130957, + "grad_norm": 0.23294557631015778, + "learning_rate": 4.202237878159967e-07, + "loss": 0.511, + "step": 11052 + }, + { + "epoch": 0.9161210111893908, + "grad_norm": 0.2436961680650711, + "learning_rate": 4.1980936593452136e-07, + "loss": 0.5089, + "step": 11053 + }, + { + "epoch": 0.9162038955656858, + "grad_norm": 0.23179000616073608, + "learning_rate": 4.1939494405304606e-07, + "loss": 0.4609, + "step": 11054 + }, + { + "epoch": 0.9162867799419809, + "grad_norm": 0.22537337243556976, + "learning_rate": 4.189805221715707e-07, + "loss": 0.4742, + "step": 11055 + }, + { + "epoch": 0.916369664318276, + "grad_norm": 0.23588937520980835, + "learning_rate": 4.1856610029009535e-07, + "loss": 0.4946, + "step": 11056 + }, + { + "epoch": 0.916452548694571, + "grad_norm": 0.2603192925453186, + "learning_rate": 4.1815167840862004e-07, + "loss": 0.5175, + "step": 11057 + }, + { + "epoch": 0.9165354330708662, + "grad_norm": 0.23582054674625397, + "learning_rate": 4.177372565271447e-07, + "loss": 0.4548, + "step": 11058 + }, + { + "epoch": 0.9166183174471612, + "grad_norm": 0.24791020154953003, + "learning_rate": 4.1732283464566933e-07, + "loss": 0.5336, + "step": 11059 + }, + { + "epoch": 0.9167012018234563, + "grad_norm": 0.23614144325256348, + "learning_rate": 4.16908412764194e-07, + "loss": 0.4587, + "step": 11060 + }, + { + "epoch": 0.9167840861997514, + "grad_norm": 0.22884690761566162, + "learning_rate": 4.164939908827186e-07, + "loss": 0.5059, + "step": 11061 + }, + { + "epoch": 0.9168669705760464, + "grad_norm": 0.2484837770462036, + "learning_rate": 4.1607956900124326e-07, + "loss": 0.4711, + "step": 11062 + }, + { + "epoch": 0.9169498549523415, + "grad_norm": 0.23269113898277283, + "learning_rate": 4.15665147119768e-07, + "loss": 0.485, + "step": 11063 + }, + { + "epoch": 0.9170327393286366, + "grad_norm": 0.23583769798278809, + "learning_rate": 4.1525072523829265e-07, + "loss": 0.4864, + "step": 11064 + }, + { + "epoch": 0.9171156237049316, + "grad_norm": 0.25085800886154175, + "learning_rate": 4.148363033568173e-07, + "loss": 0.4897, + "step": 11065 + }, + { + "epoch": 0.9171985080812267, + "grad_norm": 0.23713766038417816, + "learning_rate": 4.1442188147534194e-07, + "loss": 0.5035, + "step": 11066 + }, + { + "epoch": 0.9172813924575217, + "grad_norm": 0.24524535238742828, + "learning_rate": 4.140074595938666e-07, + "loss": 0.4824, + "step": 11067 + }, + { + "epoch": 0.9173642768338168, + "grad_norm": 0.238126739859581, + "learning_rate": 4.1359303771239123e-07, + "loss": 0.5278, + "step": 11068 + }, + { + "epoch": 0.9174471612101119, + "grad_norm": 0.22071614861488342, + "learning_rate": 4.1317861583091587e-07, + "loss": 0.4783, + "step": 11069 + }, + { + "epoch": 0.9175300455864069, + "grad_norm": 0.21649613976478577, + "learning_rate": 4.1276419394944057e-07, + "loss": 0.4658, + "step": 11070 + }, + { + "epoch": 0.917612929962702, + "grad_norm": 0.2412692755460739, + "learning_rate": 4.123497720679652e-07, + "loss": 0.5207, + "step": 11071 + }, + { + "epoch": 0.917695814338997, + "grad_norm": 0.23762720823287964, + "learning_rate": 4.1193535018648986e-07, + "loss": 0.4514, + "step": 11072 + }, + { + "epoch": 0.9177786987152922, + "grad_norm": 0.24469518661499023, + "learning_rate": 4.1152092830501455e-07, + "loss": 0.4891, + "step": 11073 + }, + { + "epoch": 0.9178615830915873, + "grad_norm": 0.24984674155712128, + "learning_rate": 4.111065064235392e-07, + "loss": 0.526, + "step": 11074 + }, + { + "epoch": 0.9179444674678823, + "grad_norm": 0.25536179542541504, + "learning_rate": 4.106920845420639e-07, + "loss": 0.5123, + "step": 11075 + }, + { + "epoch": 0.9180273518441774, + "grad_norm": 0.258391410112381, + "learning_rate": 4.1027766266058854e-07, + "loss": 0.5111, + "step": 11076 + }, + { + "epoch": 0.9181102362204724, + "grad_norm": 0.23326535522937775, + "learning_rate": 4.098632407791132e-07, + "loss": 0.4972, + "step": 11077 + }, + { + "epoch": 0.9181931205967675, + "grad_norm": 0.23962651193141937, + "learning_rate": 4.094488188976378e-07, + "loss": 0.4608, + "step": 11078 + }, + { + "epoch": 0.9182760049730626, + "grad_norm": 0.23339369893074036, + "learning_rate": 4.0903439701616247e-07, + "loss": 0.4957, + "step": 11079 + }, + { + "epoch": 0.9183588893493576, + "grad_norm": 0.23469966650009155, + "learning_rate": 4.086199751346871e-07, + "loss": 0.5182, + "step": 11080 + }, + { + "epoch": 0.9184417737256527, + "grad_norm": 0.2410801202058792, + "learning_rate": 4.0820555325321175e-07, + "loss": 0.4834, + "step": 11081 + }, + { + "epoch": 0.9185246581019478, + "grad_norm": 0.25718486309051514, + "learning_rate": 4.077911313717365e-07, + "loss": 0.4991, + "step": 11082 + }, + { + "epoch": 0.9186075424782428, + "grad_norm": 0.23811542987823486, + "learning_rate": 4.0737670949026115e-07, + "loss": 0.4661, + "step": 11083 + }, + { + "epoch": 0.918690426854538, + "grad_norm": 0.25624561309814453, + "learning_rate": 4.069622876087858e-07, + "loss": 0.5428, + "step": 11084 + }, + { + "epoch": 0.918773311230833, + "grad_norm": 0.23816578090190887, + "learning_rate": 4.0654786572731043e-07, + "loss": 0.4695, + "step": 11085 + }, + { + "epoch": 0.9188561956071281, + "grad_norm": 0.23226766288280487, + "learning_rate": 4.061334438458351e-07, + "loss": 0.4959, + "step": 11086 + }, + { + "epoch": 0.9189390799834232, + "grad_norm": 0.23301517963409424, + "learning_rate": 4.057190219643598e-07, + "loss": 0.4711, + "step": 11087 + }, + { + "epoch": 0.9190219643597182, + "grad_norm": 0.22410933673381805, + "learning_rate": 4.053046000828844e-07, + "loss": 0.4451, + "step": 11088 + }, + { + "epoch": 0.9191048487360133, + "grad_norm": 0.21824310719966888, + "learning_rate": 4.0489017820140906e-07, + "loss": 0.4127, + "step": 11089 + }, + { + "epoch": 0.9191877331123083, + "grad_norm": 0.23802269995212555, + "learning_rate": 4.044757563199337e-07, + "loss": 0.4551, + "step": 11090 + }, + { + "epoch": 0.9192706174886034, + "grad_norm": 0.26297086477279663, + "learning_rate": 4.0406133443845835e-07, + "loss": 0.527, + "step": 11091 + }, + { + "epoch": 0.9193535018648985, + "grad_norm": 0.23661471903324127, + "learning_rate": 4.036469125569831e-07, + "loss": 0.5045, + "step": 11092 + }, + { + "epoch": 0.9194363862411935, + "grad_norm": 0.23147587478160858, + "learning_rate": 4.0323249067550774e-07, + "loss": 0.4834, + "step": 11093 + }, + { + "epoch": 0.9195192706174886, + "grad_norm": 0.2427765280008316, + "learning_rate": 4.028180687940324e-07, + "loss": 0.5081, + "step": 11094 + }, + { + "epoch": 0.9196021549937837, + "grad_norm": 0.25318270921707153, + "learning_rate": 4.0240364691255703e-07, + "loss": 0.5469, + "step": 11095 + }, + { + "epoch": 0.9196850393700787, + "grad_norm": 0.2417239099740982, + "learning_rate": 4.0198922503108167e-07, + "loss": 0.4799, + "step": 11096 + }, + { + "epoch": 0.9197679237463738, + "grad_norm": 0.2322654128074646, + "learning_rate": 4.015748031496063e-07, + "loss": 0.4979, + "step": 11097 + }, + { + "epoch": 0.9198508081226688, + "grad_norm": 0.24573124945163727, + "learning_rate": 4.0116038126813096e-07, + "loss": 0.5352, + "step": 11098 + }, + { + "epoch": 0.919933692498964, + "grad_norm": 0.24537549912929535, + "learning_rate": 4.007459593866556e-07, + "loss": 0.4885, + "step": 11099 + }, + { + "epoch": 0.9200165768752591, + "grad_norm": 0.22548002004623413, + "learning_rate": 4.003315375051803e-07, + "loss": 0.4781, + "step": 11100 + }, + { + "epoch": 0.9200994612515541, + "grad_norm": 0.24484321475028992, + "learning_rate": 3.99917115623705e-07, + "loss": 0.4914, + "step": 11101 + }, + { + "epoch": 0.9201823456278492, + "grad_norm": 0.23593951761722565, + "learning_rate": 3.9950269374222964e-07, + "loss": 0.5008, + "step": 11102 + }, + { + "epoch": 0.9202652300041442, + "grad_norm": 0.24976293742656708, + "learning_rate": 3.990882718607543e-07, + "loss": 0.53, + "step": 11103 + }, + { + "epoch": 0.9203481143804393, + "grad_norm": 0.23272788524627686, + "learning_rate": 3.986738499792789e-07, + "loss": 0.4889, + "step": 11104 + }, + { + "epoch": 0.9204309987567344, + "grad_norm": 0.24582833051681519, + "learning_rate": 3.982594280978036e-07, + "loss": 0.468, + "step": 11105 + }, + { + "epoch": 0.9205138831330294, + "grad_norm": 0.2185535430908203, + "learning_rate": 3.9784500621632827e-07, + "loss": 0.4473, + "step": 11106 + }, + { + "epoch": 0.9205967675093245, + "grad_norm": 0.2531065046787262, + "learning_rate": 3.974305843348529e-07, + "loss": 0.4887, + "step": 11107 + }, + { + "epoch": 0.9206796518856195, + "grad_norm": 0.2626193165779114, + "learning_rate": 3.9701616245337755e-07, + "loss": 0.5011, + "step": 11108 + }, + { + "epoch": 0.9207625362619146, + "grad_norm": 0.49525186419487, + "learning_rate": 3.966017405719022e-07, + "loss": 0.5426, + "step": 11109 + }, + { + "epoch": 0.9208454206382097, + "grad_norm": 0.24178172647953033, + "learning_rate": 3.9618731869042684e-07, + "loss": 0.4833, + "step": 11110 + }, + { + "epoch": 0.9209283050145047, + "grad_norm": 0.2608765959739685, + "learning_rate": 3.957728968089516e-07, + "loss": 0.5292, + "step": 11111 + }, + { + "epoch": 0.9210111893907998, + "grad_norm": 0.2347300797700882, + "learning_rate": 3.9535847492747623e-07, + "loss": 0.4666, + "step": 11112 + }, + { + "epoch": 0.921094073767095, + "grad_norm": 0.24867142736911774, + "learning_rate": 3.949440530460009e-07, + "loss": 0.5018, + "step": 11113 + }, + { + "epoch": 0.92117695814339, + "grad_norm": 0.23412634432315826, + "learning_rate": 3.945296311645255e-07, + "loss": 0.4886, + "step": 11114 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 0.25324225425720215, + "learning_rate": 3.9411520928305016e-07, + "loss": 0.5693, + "step": 11115 + }, + { + "epoch": 0.9213427268959801, + "grad_norm": 0.24770738184452057, + "learning_rate": 3.937007874015748e-07, + "loss": 0.4833, + "step": 11116 + }, + { + "epoch": 0.9214256112722752, + "grad_norm": 0.24025753140449524, + "learning_rate": 3.932863655200995e-07, + "loss": 0.5055, + "step": 11117 + }, + { + "epoch": 0.9215084956485703, + "grad_norm": 0.24431435763835907, + "learning_rate": 3.9287194363862415e-07, + "loss": 0.4852, + "step": 11118 + }, + { + "epoch": 0.9215913800248653, + "grad_norm": 0.24431009590625763, + "learning_rate": 3.924575217571488e-07, + "loss": 0.5796, + "step": 11119 + }, + { + "epoch": 0.9216742644011604, + "grad_norm": 0.24044209718704224, + "learning_rate": 3.9204309987567343e-07, + "loss": 0.4803, + "step": 11120 + }, + { + "epoch": 0.9217571487774554, + "grad_norm": 0.2884967625141144, + "learning_rate": 3.9162867799419813e-07, + "loss": 0.5264, + "step": 11121 + }, + { + "epoch": 0.9218400331537505, + "grad_norm": 0.2395898997783661, + "learning_rate": 3.9121425611272283e-07, + "loss": 0.4507, + "step": 11122 + }, + { + "epoch": 0.9219229175300456, + "grad_norm": 0.23173901438713074, + "learning_rate": 3.9079983423124747e-07, + "loss": 0.4651, + "step": 11123 + }, + { + "epoch": 0.9220058019063406, + "grad_norm": 0.22781696915626526, + "learning_rate": 3.903854123497721e-07, + "loss": 0.5142, + "step": 11124 + }, + { + "epoch": 0.9220886862826357, + "grad_norm": 0.23169469833374023, + "learning_rate": 3.8997099046829676e-07, + "loss": 0.4768, + "step": 11125 + }, + { + "epoch": 0.9221715706589307, + "grad_norm": 0.24143877625465393, + "learning_rate": 3.895565685868214e-07, + "loss": 0.4977, + "step": 11126 + }, + { + "epoch": 0.9222544550352259, + "grad_norm": 0.2316935509443283, + "learning_rate": 3.8914214670534605e-07, + "loss": 0.4665, + "step": 11127 + }, + { + "epoch": 0.922337339411521, + "grad_norm": 0.24115169048309326, + "learning_rate": 3.887277248238707e-07, + "loss": 0.4843, + "step": 11128 + }, + { + "epoch": 0.922420223787816, + "grad_norm": 0.22892886400222778, + "learning_rate": 3.883133029423954e-07, + "loss": 0.4702, + "step": 11129 + }, + { + "epoch": 0.9225031081641111, + "grad_norm": 0.22872675955295563, + "learning_rate": 3.878988810609201e-07, + "loss": 0.4688, + "step": 11130 + }, + { + "epoch": 0.9225859925404062, + "grad_norm": 0.23492605984210968, + "learning_rate": 3.874844591794447e-07, + "loss": 0.4886, + "step": 11131 + }, + { + "epoch": 0.9226688769167012, + "grad_norm": 0.24219000339508057, + "learning_rate": 3.8707003729796937e-07, + "loss": 0.5119, + "step": 11132 + }, + { + "epoch": 0.9227517612929963, + "grad_norm": 0.24671132862567902, + "learning_rate": 3.86655615416494e-07, + "loss": 0.4866, + "step": 11133 + }, + { + "epoch": 0.9228346456692913, + "grad_norm": 0.24693648517131805, + "learning_rate": 3.8624119353501866e-07, + "loss": 0.4932, + "step": 11134 + }, + { + "epoch": 0.9229175300455864, + "grad_norm": 0.24232572317123413, + "learning_rate": 3.8582677165354335e-07, + "loss": 0.4696, + "step": 11135 + }, + { + "epoch": 0.9230004144218815, + "grad_norm": 0.23034384846687317, + "learning_rate": 3.85412349772068e-07, + "loss": 0.5003, + "step": 11136 + }, + { + "epoch": 0.9230832987981765, + "grad_norm": 0.22721923887729645, + "learning_rate": 3.8499792789059264e-07, + "loss": 0.4734, + "step": 11137 + }, + { + "epoch": 0.9231661831744716, + "grad_norm": 0.25368547439575195, + "learning_rate": 3.845835060091173e-07, + "loss": 0.4885, + "step": 11138 + }, + { + "epoch": 0.9232490675507666, + "grad_norm": 0.23171868920326233, + "learning_rate": 3.8416908412764193e-07, + "loss": 0.5252, + "step": 11139 + }, + { + "epoch": 0.9233319519270617, + "grad_norm": 0.2360566258430481, + "learning_rate": 3.837546622461667e-07, + "loss": 0.5079, + "step": 11140 + }, + { + "epoch": 0.9234148363033569, + "grad_norm": 0.25571009516716003, + "learning_rate": 3.833402403646913e-07, + "loss": 0.5079, + "step": 11141 + }, + { + "epoch": 0.9234977206796519, + "grad_norm": 0.23847892880439758, + "learning_rate": 3.8292581848321596e-07, + "loss": 0.4484, + "step": 11142 + }, + { + "epoch": 0.923580605055947, + "grad_norm": 0.2395121157169342, + "learning_rate": 3.825113966017406e-07, + "loss": 0.5061, + "step": 11143 + }, + { + "epoch": 0.923663489432242, + "grad_norm": 0.2296699732542038, + "learning_rate": 3.8209697472026525e-07, + "loss": 0.4977, + "step": 11144 + }, + { + "epoch": 0.9237463738085371, + "grad_norm": 0.2236250936985016, + "learning_rate": 3.816825528387899e-07, + "loss": 0.4676, + "step": 11145 + }, + { + "epoch": 0.9238292581848322, + "grad_norm": 0.22323252260684967, + "learning_rate": 3.8126813095731454e-07, + "loss": 0.4391, + "step": 11146 + }, + { + "epoch": 0.9239121425611272, + "grad_norm": 0.23708045482635498, + "learning_rate": 3.8085370907583923e-07, + "loss": 0.4991, + "step": 11147 + }, + { + "epoch": 0.9239950269374223, + "grad_norm": 0.24494558572769165, + "learning_rate": 3.804392871943639e-07, + "loss": 0.5204, + "step": 11148 + }, + { + "epoch": 0.9240779113137174, + "grad_norm": 0.27014073729515076, + "learning_rate": 3.800248653128886e-07, + "loss": 0.4848, + "step": 11149 + }, + { + "epoch": 0.9241607956900124, + "grad_norm": 0.24518825113773346, + "learning_rate": 3.796104434314132e-07, + "loss": 0.4937, + "step": 11150 + }, + { + "epoch": 0.9242436800663075, + "grad_norm": 0.24161642789840698, + "learning_rate": 3.7919602154993786e-07, + "loss": 0.5233, + "step": 11151 + }, + { + "epoch": 0.9243265644426025, + "grad_norm": 0.234612375497818, + "learning_rate": 3.7878159966846256e-07, + "loss": 0.4762, + "step": 11152 + }, + { + "epoch": 0.9244094488188976, + "grad_norm": 0.23455077409744263, + "learning_rate": 3.783671777869872e-07, + "loss": 0.447, + "step": 11153 + }, + { + "epoch": 0.9244923331951928, + "grad_norm": 0.24152648448944092, + "learning_rate": 3.7795275590551184e-07, + "loss": 0.4746, + "step": 11154 + }, + { + "epoch": 0.9245752175714878, + "grad_norm": 0.2350127398967743, + "learning_rate": 3.775383340240365e-07, + "loss": 0.5025, + "step": 11155 + }, + { + "epoch": 0.9246581019477829, + "grad_norm": 0.24485044181346893, + "learning_rate": 3.7712391214256113e-07, + "loss": 0.5219, + "step": 11156 + }, + { + "epoch": 0.9247409863240779, + "grad_norm": 0.23815180361270905, + "learning_rate": 3.767094902610858e-07, + "loss": 0.4778, + "step": 11157 + }, + { + "epoch": 0.924823870700373, + "grad_norm": 0.24947313964366913, + "learning_rate": 3.762950683796104e-07, + "loss": 0.496, + "step": 11158 + }, + { + "epoch": 0.9249067550766681, + "grad_norm": 0.23273958265781403, + "learning_rate": 3.7588064649813517e-07, + "loss": 0.4946, + "step": 11159 + }, + { + "epoch": 0.9249896394529631, + "grad_norm": 0.22741517424583435, + "learning_rate": 3.754662246166598e-07, + "loss": 0.4688, + "step": 11160 + }, + { + "epoch": 0.9250725238292582, + "grad_norm": 0.23086191713809967, + "learning_rate": 3.7505180273518446e-07, + "loss": 0.4856, + "step": 11161 + }, + { + "epoch": 0.9251554082055532, + "grad_norm": 0.24560540914535522, + "learning_rate": 3.746373808537091e-07, + "loss": 0.485, + "step": 11162 + }, + { + "epoch": 0.9252382925818483, + "grad_norm": 0.2387905716896057, + "learning_rate": 3.7422295897223374e-07, + "loss": 0.4926, + "step": 11163 + }, + { + "epoch": 0.9253211769581434, + "grad_norm": 0.2379046380519867, + "learning_rate": 3.7380853709075844e-07, + "loss": 0.5138, + "step": 11164 + }, + { + "epoch": 0.9254040613344384, + "grad_norm": 0.24842263758182526, + "learning_rate": 3.733941152092831e-07, + "loss": 0.5107, + "step": 11165 + }, + { + "epoch": 0.9254869457107335, + "grad_norm": 0.21848748624324799, + "learning_rate": 3.729796933278077e-07, + "loss": 0.4684, + "step": 11166 + }, + { + "epoch": 0.9255698300870286, + "grad_norm": 0.2361539900302887, + "learning_rate": 3.7256527144633237e-07, + "loss": 0.5021, + "step": 11167 + }, + { + "epoch": 0.9256527144633236, + "grad_norm": 0.23145844042301178, + "learning_rate": 3.7215084956485707e-07, + "loss": 0.4857, + "step": 11168 + }, + { + "epoch": 0.9257355988396188, + "grad_norm": 0.22928859293460846, + "learning_rate": 3.7173642768338176e-07, + "loss": 0.4388, + "step": 11169 + }, + { + "epoch": 0.9258184832159138, + "grad_norm": 0.2463676780462265, + "learning_rate": 3.713220058019064e-07, + "loss": 0.4926, + "step": 11170 + }, + { + "epoch": 0.9259013675922089, + "grad_norm": 0.2420349270105362, + "learning_rate": 3.7090758392043105e-07, + "loss": 0.5082, + "step": 11171 + }, + { + "epoch": 0.925984251968504, + "grad_norm": 0.2417917549610138, + "learning_rate": 3.704931620389557e-07, + "loss": 0.5124, + "step": 11172 + }, + { + "epoch": 0.926067136344799, + "grad_norm": 0.23810958862304688, + "learning_rate": 3.7007874015748034e-07, + "loss": 0.4881, + "step": 11173 + }, + { + "epoch": 0.9261500207210941, + "grad_norm": 0.22809641063213348, + "learning_rate": 3.69664318276005e-07, + "loss": 0.4699, + "step": 11174 + }, + { + "epoch": 0.9262329050973891, + "grad_norm": 0.24859033524990082, + "learning_rate": 3.692498963945296e-07, + "loss": 0.5074, + "step": 11175 + }, + { + "epoch": 0.9263157894736842, + "grad_norm": 0.21964886784553528, + "learning_rate": 3.6883547451305427e-07, + "loss": 0.4741, + "step": 11176 + }, + { + "epoch": 0.9263986738499793, + "grad_norm": 0.24429212510585785, + "learning_rate": 3.6842105263157896e-07, + "loss": 0.4907, + "step": 11177 + }, + { + "epoch": 0.9264815582262743, + "grad_norm": 0.23666945099830627, + "learning_rate": 3.6800663075010366e-07, + "loss": 0.4929, + "step": 11178 + }, + { + "epoch": 0.9265644426025694, + "grad_norm": 0.27709126472473145, + "learning_rate": 3.675922088686283e-07, + "loss": 0.4641, + "step": 11179 + }, + { + "epoch": 0.9266473269788645, + "grad_norm": 0.2347952127456665, + "learning_rate": 3.6717778698715295e-07, + "loss": 0.4707, + "step": 11180 + }, + { + "epoch": 0.9267302113551595, + "grad_norm": 0.23867560923099518, + "learning_rate": 3.667633651056776e-07, + "loss": 0.5294, + "step": 11181 + }, + { + "epoch": 0.9268130957314547, + "grad_norm": 0.2435847371816635, + "learning_rate": 3.663489432242023e-07, + "loss": 0.5016, + "step": 11182 + }, + { + "epoch": 0.9268959801077497, + "grad_norm": 0.23584628105163574, + "learning_rate": 3.6593452134272693e-07, + "loss": 0.5007, + "step": 11183 + }, + { + "epoch": 0.9269788644840448, + "grad_norm": 0.23451852798461914, + "learning_rate": 3.655200994612516e-07, + "loss": 0.4672, + "step": 11184 + }, + { + "epoch": 0.9270617488603399, + "grad_norm": 0.2357151061296463, + "learning_rate": 3.651056775797762e-07, + "loss": 0.4994, + "step": 11185 + }, + { + "epoch": 0.9271446332366349, + "grad_norm": 0.2473149299621582, + "learning_rate": 3.6469125569830086e-07, + "loss": 0.4907, + "step": 11186 + }, + { + "epoch": 0.92722751761293, + "grad_norm": 0.24088925123214722, + "learning_rate": 3.642768338168256e-07, + "loss": 0.5084, + "step": 11187 + }, + { + "epoch": 0.927310401989225, + "grad_norm": 0.23840613663196564, + "learning_rate": 3.6386241193535025e-07, + "loss": 0.4932, + "step": 11188 + }, + { + "epoch": 0.9273932863655201, + "grad_norm": 0.23818664252758026, + "learning_rate": 3.634479900538749e-07, + "loss": 0.4783, + "step": 11189 + }, + { + "epoch": 0.9274761707418152, + "grad_norm": 0.2299446165561676, + "learning_rate": 3.6303356817239954e-07, + "loss": 0.4785, + "step": 11190 + }, + { + "epoch": 0.9275590551181102, + "grad_norm": 0.24015890061855316, + "learning_rate": 3.626191462909242e-07, + "loss": 0.5, + "step": 11191 + }, + { + "epoch": 0.9276419394944053, + "grad_norm": 0.21823595464229584, + "learning_rate": 3.6220472440944883e-07, + "loss": 0.492, + "step": 11192 + }, + { + "epoch": 0.9277248238707003, + "grad_norm": 0.2330089807510376, + "learning_rate": 3.6179030252797347e-07, + "loss": 0.4547, + "step": 11193 + }, + { + "epoch": 0.9278077082469954, + "grad_norm": 0.25333648920059204, + "learning_rate": 3.6137588064649817e-07, + "loss": 0.5469, + "step": 11194 + }, + { + "epoch": 0.9278905926232905, + "grad_norm": 0.23064100742340088, + "learning_rate": 3.609614587650228e-07, + "loss": 0.4803, + "step": 11195 + }, + { + "epoch": 0.9279734769995855, + "grad_norm": 0.23315918445587158, + "learning_rate": 3.6054703688354746e-07, + "loss": 0.4794, + "step": 11196 + }, + { + "epoch": 0.9280563613758807, + "grad_norm": 0.23694494366645813, + "learning_rate": 3.6013261500207215e-07, + "loss": 0.4771, + "step": 11197 + }, + { + "epoch": 0.9281392457521758, + "grad_norm": 0.24583002924919128, + "learning_rate": 3.597181931205968e-07, + "loss": 0.5111, + "step": 11198 + }, + { + "epoch": 0.9282221301284708, + "grad_norm": 0.25696033239364624, + "learning_rate": 3.593037712391215e-07, + "loss": 0.5157, + "step": 11199 + }, + { + "epoch": 0.9283050145047659, + "grad_norm": 0.2607084810733795, + "learning_rate": 3.5888934935764614e-07, + "loss": 0.5012, + "step": 11200 + }, + { + "epoch": 0.9283878988810609, + "grad_norm": 0.2457619458436966, + "learning_rate": 3.584749274761708e-07, + "loss": 0.4997, + "step": 11201 + }, + { + "epoch": 0.928470783257356, + "grad_norm": 0.24187098443508148, + "learning_rate": 3.580605055946954e-07, + "loss": 0.5035, + "step": 11202 + }, + { + "epoch": 0.9285536676336511, + "grad_norm": 0.26001134514808655, + "learning_rate": 3.5764608371322007e-07, + "loss": 0.5222, + "step": 11203 + }, + { + "epoch": 0.9286365520099461, + "grad_norm": 0.2528049349784851, + "learning_rate": 3.572316618317447e-07, + "loss": 0.5519, + "step": 11204 + }, + { + "epoch": 0.9287194363862412, + "grad_norm": 0.2408076971769333, + "learning_rate": 3.5681723995026935e-07, + "loss": 0.4897, + "step": 11205 + }, + { + "epoch": 0.9288023207625362, + "grad_norm": 0.2219531387090683, + "learning_rate": 3.564028180687941e-07, + "loss": 0.4547, + "step": 11206 + }, + { + "epoch": 0.9288852051388313, + "grad_norm": 0.2514455020427704, + "learning_rate": 3.5598839618731875e-07, + "loss": 0.4608, + "step": 11207 + }, + { + "epoch": 0.9289680895151264, + "grad_norm": 0.2485358715057373, + "learning_rate": 3.555739743058434e-07, + "loss": 0.4689, + "step": 11208 + }, + { + "epoch": 0.9290509738914214, + "grad_norm": 0.2438933253288269, + "learning_rate": 3.5515955242436803e-07, + "loss": 0.4976, + "step": 11209 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 0.24874745309352875, + "learning_rate": 3.547451305428927e-07, + "loss": 0.5219, + "step": 11210 + }, + { + "epoch": 0.9292167426440116, + "grad_norm": 0.23263797163963318, + "learning_rate": 3.543307086614173e-07, + "loss": 0.459, + "step": 11211 + }, + { + "epoch": 0.9292996270203067, + "grad_norm": 0.24161632359027863, + "learning_rate": 3.53916286779942e-07, + "loss": 0.4723, + "step": 11212 + }, + { + "epoch": 0.9293825113966018, + "grad_norm": 0.24588090181350708, + "learning_rate": 3.5350186489846666e-07, + "loss": 0.4564, + "step": 11213 + }, + { + "epoch": 0.9294653957728968, + "grad_norm": 0.22436431050300598, + "learning_rate": 3.530874430169913e-07, + "loss": 0.4778, + "step": 11214 + }, + { + "epoch": 0.9295482801491919, + "grad_norm": 0.24168072640895844, + "learning_rate": 3.5267302113551595e-07, + "loss": 0.467, + "step": 11215 + }, + { + "epoch": 0.929631164525487, + "grad_norm": 0.22763808071613312, + "learning_rate": 3.5225859925404065e-07, + "loss": 0.4971, + "step": 11216 + }, + { + "epoch": 0.929714048901782, + "grad_norm": 0.24790440499782562, + "learning_rate": 3.5184417737256534e-07, + "loss": 0.4955, + "step": 11217 + }, + { + "epoch": 0.9297969332780771, + "grad_norm": 0.2301531881093979, + "learning_rate": 3.5142975549109e-07, + "loss": 0.4324, + "step": 11218 + }, + { + "epoch": 0.9298798176543721, + "grad_norm": 0.23525504767894745, + "learning_rate": 3.5101533360961463e-07, + "loss": 0.4817, + "step": 11219 + }, + { + "epoch": 0.9299627020306672, + "grad_norm": 0.2564193606376648, + "learning_rate": 3.5060091172813927e-07, + "loss": 0.5009, + "step": 11220 + }, + { + "epoch": 0.9300455864069623, + "grad_norm": 0.2468210905790329, + "learning_rate": 3.501864898466639e-07, + "loss": 0.5115, + "step": 11221 + }, + { + "epoch": 0.9301284707832573, + "grad_norm": 0.24615085124969482, + "learning_rate": 3.4977206796518856e-07, + "loss": 0.5301, + "step": 11222 + }, + { + "epoch": 0.9302113551595524, + "grad_norm": 0.23498456180095673, + "learning_rate": 3.493576460837132e-07, + "loss": 0.501, + "step": 11223 + }, + { + "epoch": 0.9302942395358474, + "grad_norm": 0.24732758104801178, + "learning_rate": 3.489432242022379e-07, + "loss": 0.4803, + "step": 11224 + }, + { + "epoch": 0.9303771239121426, + "grad_norm": 0.24329234659671783, + "learning_rate": 3.485288023207626e-07, + "loss": 0.5453, + "step": 11225 + }, + { + "epoch": 0.9304600082884377, + "grad_norm": 0.23988981544971466, + "learning_rate": 3.4811438043928724e-07, + "loss": 0.5003, + "step": 11226 + }, + { + "epoch": 0.9305428926647327, + "grad_norm": 0.23182249069213867, + "learning_rate": 3.476999585578119e-07, + "loss": 0.4903, + "step": 11227 + }, + { + "epoch": 0.9306257770410278, + "grad_norm": 0.2229297161102295, + "learning_rate": 3.4728553667633653e-07, + "loss": 0.4308, + "step": 11228 + }, + { + "epoch": 0.9307086614173228, + "grad_norm": 0.23795372247695923, + "learning_rate": 3.468711147948612e-07, + "loss": 0.487, + "step": 11229 + }, + { + "epoch": 0.9307915457936179, + "grad_norm": 0.21828609704971313, + "learning_rate": 3.4645669291338587e-07, + "loss": 0.4507, + "step": 11230 + }, + { + "epoch": 0.930874430169913, + "grad_norm": 0.2321958839893341, + "learning_rate": 3.460422710319105e-07, + "loss": 0.4871, + "step": 11231 + }, + { + "epoch": 0.930957314546208, + "grad_norm": 0.24411503970623016, + "learning_rate": 3.4562784915043515e-07, + "loss": 0.5112, + "step": 11232 + }, + { + "epoch": 0.9310401989225031, + "grad_norm": 0.21815542876720428, + "learning_rate": 3.452134272689598e-07, + "loss": 0.4439, + "step": 11233 + }, + { + "epoch": 0.9311230832987982, + "grad_norm": 0.2374567985534668, + "learning_rate": 3.4479900538748444e-07, + "loss": 0.4929, + "step": 11234 + }, + { + "epoch": 0.9312059676750932, + "grad_norm": 0.25023728609085083, + "learning_rate": 3.443845835060092e-07, + "loss": 0.4835, + "step": 11235 + }, + { + "epoch": 0.9312888520513883, + "grad_norm": 0.2345862090587616, + "learning_rate": 3.4397016162453383e-07, + "loss": 0.4948, + "step": 11236 + }, + { + "epoch": 0.9313717364276833, + "grad_norm": 0.24823027849197388, + "learning_rate": 3.435557397430585e-07, + "loss": 0.4846, + "step": 11237 + }, + { + "epoch": 0.9314546208039785, + "grad_norm": 0.2528548538684845, + "learning_rate": 3.431413178615831e-07, + "loss": 0.5037, + "step": 11238 + }, + { + "epoch": 0.9315375051802736, + "grad_norm": 0.22841860353946686, + "learning_rate": 3.4272689598010776e-07, + "loss": 0.4588, + "step": 11239 + }, + { + "epoch": 0.9316203895565686, + "grad_norm": 0.23064358532428741, + "learning_rate": 3.423124740986324e-07, + "loss": 0.5012, + "step": 11240 + }, + { + "epoch": 0.9317032739328637, + "grad_norm": 0.23623430728912354, + "learning_rate": 3.418980522171571e-07, + "loss": 0.4893, + "step": 11241 + }, + { + "epoch": 0.9317861583091587, + "grad_norm": 0.2486286610364914, + "learning_rate": 3.4148363033568175e-07, + "loss": 0.4838, + "step": 11242 + }, + { + "epoch": 0.9318690426854538, + "grad_norm": 0.2375546544790268, + "learning_rate": 3.410692084542064e-07, + "loss": 0.4996, + "step": 11243 + }, + { + "epoch": 0.9319519270617489, + "grad_norm": 0.24490223824977875, + "learning_rate": 3.406547865727311e-07, + "loss": 0.5081, + "step": 11244 + }, + { + "epoch": 0.9320348114380439, + "grad_norm": 0.24364334344863892, + "learning_rate": 3.4024036469125573e-07, + "loss": 0.5247, + "step": 11245 + }, + { + "epoch": 0.932117695814339, + "grad_norm": 0.23198702931404114, + "learning_rate": 3.398259428097804e-07, + "loss": 0.5191, + "step": 11246 + }, + { + "epoch": 0.9322005801906341, + "grad_norm": 0.24211400747299194, + "learning_rate": 3.3941152092830507e-07, + "loss": 0.5018, + "step": 11247 + }, + { + "epoch": 0.9322834645669291, + "grad_norm": 0.24442802369594574, + "learning_rate": 3.389970990468297e-07, + "loss": 0.4741, + "step": 11248 + }, + { + "epoch": 0.9323663489432242, + "grad_norm": 0.23168928921222687, + "learning_rate": 3.3858267716535436e-07, + "loss": 0.46, + "step": 11249 + }, + { + "epoch": 0.9324492333195192, + "grad_norm": 0.24978038668632507, + "learning_rate": 3.38168255283879e-07, + "loss": 0.5258, + "step": 11250 + }, + { + "epoch": 0.9325321176958143, + "grad_norm": 0.22539407014846802, + "learning_rate": 3.3775383340240365e-07, + "loss": 0.4716, + "step": 11251 + }, + { + "epoch": 0.9326150020721095, + "grad_norm": 0.2344532310962677, + "learning_rate": 3.373394115209283e-07, + "loss": 0.4735, + "step": 11252 + }, + { + "epoch": 0.9326978864484045, + "grad_norm": 0.2539006173610687, + "learning_rate": 3.3692498963945293e-07, + "loss": 0.5011, + "step": 11253 + }, + { + "epoch": 0.9327807708246996, + "grad_norm": 0.24847322702407837, + "learning_rate": 3.365105677579777e-07, + "loss": 0.5144, + "step": 11254 + }, + { + "epoch": 0.9328636552009946, + "grad_norm": 0.24110731482505798, + "learning_rate": 3.360961458765023e-07, + "loss": 0.5001, + "step": 11255 + }, + { + "epoch": 0.9329465395772897, + "grad_norm": 0.23564350605010986, + "learning_rate": 3.3568172399502697e-07, + "loss": 0.4751, + "step": 11256 + }, + { + "epoch": 0.9330294239535848, + "grad_norm": 0.2250724732875824, + "learning_rate": 3.352673021135516e-07, + "loss": 0.472, + "step": 11257 + }, + { + "epoch": 0.9331123083298798, + "grad_norm": 0.22544091939926147, + "learning_rate": 3.3485288023207626e-07, + "loss": 0.4727, + "step": 11258 + }, + { + "epoch": 0.9331951927061749, + "grad_norm": 0.24081385135650635, + "learning_rate": 3.3443845835060095e-07, + "loss": 0.5009, + "step": 11259 + }, + { + "epoch": 0.9332780770824699, + "grad_norm": 0.22727665305137634, + "learning_rate": 3.340240364691256e-07, + "loss": 0.4782, + "step": 11260 + }, + { + "epoch": 0.933360961458765, + "grad_norm": 0.23674726486206055, + "learning_rate": 3.3360961458765024e-07, + "loss": 0.4692, + "step": 11261 + }, + { + "epoch": 0.9334438458350601, + "grad_norm": 0.23004698753356934, + "learning_rate": 3.331951927061749e-07, + "loss": 0.4594, + "step": 11262 + }, + { + "epoch": 0.9335267302113551, + "grad_norm": 0.24865040183067322, + "learning_rate": 3.327807708246996e-07, + "loss": 0.4907, + "step": 11263 + }, + { + "epoch": 0.9336096145876502, + "grad_norm": 0.22917591035366058, + "learning_rate": 3.323663489432243e-07, + "loss": 0.4835, + "step": 11264 + }, + { + "epoch": 0.9336924989639453, + "grad_norm": 0.2260863482952118, + "learning_rate": 3.319519270617489e-07, + "loss": 0.4613, + "step": 11265 + }, + { + "epoch": 0.9337753833402404, + "grad_norm": 0.24515078961849213, + "learning_rate": 3.3153750518027356e-07, + "loss": 0.4909, + "step": 11266 + }, + { + "epoch": 0.9338582677165355, + "grad_norm": 0.23420089483261108, + "learning_rate": 3.311230832987982e-07, + "loss": 0.4852, + "step": 11267 + }, + { + "epoch": 0.9339411520928305, + "grad_norm": 0.22917091846466064, + "learning_rate": 3.3070866141732285e-07, + "loss": 0.4828, + "step": 11268 + }, + { + "epoch": 0.9340240364691256, + "grad_norm": 0.2336301952600479, + "learning_rate": 3.302942395358475e-07, + "loss": 0.4774, + "step": 11269 + }, + { + "epoch": 0.9341069208454207, + "grad_norm": 0.24243170022964478, + "learning_rate": 3.2987981765437214e-07, + "loss": 0.5129, + "step": 11270 + }, + { + "epoch": 0.9341898052217157, + "grad_norm": 0.248285710811615, + "learning_rate": 3.2946539577289683e-07, + "loss": 0.5089, + "step": 11271 + }, + { + "epoch": 0.9342726895980108, + "grad_norm": 0.2271849364042282, + "learning_rate": 3.290509738914215e-07, + "loss": 0.4601, + "step": 11272 + }, + { + "epoch": 0.9343555739743058, + "grad_norm": 0.23860277235507965, + "learning_rate": 3.286365520099462e-07, + "loss": 0.5173, + "step": 11273 + }, + { + "epoch": 0.9344384583506009, + "grad_norm": 0.24538613855838776, + "learning_rate": 3.282221301284708e-07, + "loss": 0.482, + "step": 11274 + }, + { + "epoch": 0.934521342726896, + "grad_norm": 0.24949806928634644, + "learning_rate": 3.2780770824699546e-07, + "loss": 0.4969, + "step": 11275 + }, + { + "epoch": 0.934604227103191, + "grad_norm": 0.22444139420986176, + "learning_rate": 3.2739328636552016e-07, + "loss": 0.4714, + "step": 11276 + }, + { + "epoch": 0.9346871114794861, + "grad_norm": 0.2417127639055252, + "learning_rate": 3.269788644840448e-07, + "loss": 0.4909, + "step": 11277 + }, + { + "epoch": 0.9347699958557811, + "grad_norm": 0.23894456028938293, + "learning_rate": 3.2656444260256945e-07, + "loss": 0.4823, + "step": 11278 + }, + { + "epoch": 0.9348528802320762, + "grad_norm": 0.2389015406370163, + "learning_rate": 3.261500207210941e-07, + "loss": 0.4687, + "step": 11279 + }, + { + "epoch": 0.9349357646083714, + "grad_norm": 0.23098696768283844, + "learning_rate": 3.2573559883961873e-07, + "loss": 0.4645, + "step": 11280 + }, + { + "epoch": 0.9350186489846664, + "grad_norm": 0.2228226512670517, + "learning_rate": 3.253211769581434e-07, + "loss": 0.4475, + "step": 11281 + }, + { + "epoch": 0.9351015333609615, + "grad_norm": 0.2300451248884201, + "learning_rate": 3.249067550766681e-07, + "loss": 0.4796, + "step": 11282 + }, + { + "epoch": 0.9351844177372566, + "grad_norm": 0.24004310369491577, + "learning_rate": 3.2449233319519277e-07, + "loss": 0.4936, + "step": 11283 + }, + { + "epoch": 0.9352673021135516, + "grad_norm": 0.24720869958400726, + "learning_rate": 3.240779113137174e-07, + "loss": 0.5231, + "step": 11284 + }, + { + "epoch": 0.9353501864898467, + "grad_norm": 0.22399084270000458, + "learning_rate": 3.2366348943224206e-07, + "loss": 0.4351, + "step": 11285 + }, + { + "epoch": 0.9354330708661417, + "grad_norm": 0.2449546456336975, + "learning_rate": 3.232490675507667e-07, + "loss": 0.5178, + "step": 11286 + }, + { + "epoch": 0.9355159552424368, + "grad_norm": 0.24975912272930145, + "learning_rate": 3.2283464566929134e-07, + "loss": 0.5393, + "step": 11287 + }, + { + "epoch": 0.9355988396187319, + "grad_norm": 0.24193406105041504, + "learning_rate": 3.22420223787816e-07, + "loss": 0.494, + "step": 11288 + }, + { + "epoch": 0.9356817239950269, + "grad_norm": 0.2385571151971817, + "learning_rate": 3.220058019063407e-07, + "loss": 0.5062, + "step": 11289 + }, + { + "epoch": 0.935764608371322, + "grad_norm": 0.2204095423221588, + "learning_rate": 3.2159138002486533e-07, + "loss": 0.4523, + "step": 11290 + }, + { + "epoch": 0.935847492747617, + "grad_norm": 0.2262590080499649, + "learning_rate": 3.2117695814338997e-07, + "loss": 0.4821, + "step": 11291 + }, + { + "epoch": 0.9359303771239121, + "grad_norm": 0.2472125142812729, + "learning_rate": 3.2076253626191467e-07, + "loss": 0.5272, + "step": 11292 + }, + { + "epoch": 0.9360132615002072, + "grad_norm": 0.2468128800392151, + "learning_rate": 3.203481143804393e-07, + "loss": 0.4915, + "step": 11293 + }, + { + "epoch": 0.9360961458765023, + "grad_norm": 0.24511125683784485, + "learning_rate": 3.19933692498964e-07, + "loss": 0.492, + "step": 11294 + }, + { + "epoch": 0.9361790302527974, + "grad_norm": 0.23942899703979492, + "learning_rate": 3.1951927061748865e-07, + "loss": 0.4964, + "step": 11295 + }, + { + "epoch": 0.9362619146290924, + "grad_norm": 0.2211521565914154, + "learning_rate": 3.191048487360133e-07, + "loss": 0.4871, + "step": 11296 + }, + { + "epoch": 0.9363447990053875, + "grad_norm": 0.22001545131206512, + "learning_rate": 3.1869042685453794e-07, + "loss": 0.4708, + "step": 11297 + }, + { + "epoch": 0.9364276833816826, + "grad_norm": 0.24112670123577118, + "learning_rate": 3.182760049730626e-07, + "loss": 0.5304, + "step": 11298 + }, + { + "epoch": 0.9365105677579776, + "grad_norm": 0.25422558188438416, + "learning_rate": 3.178615830915872e-07, + "loss": 0.5033, + "step": 11299 + }, + { + "epoch": 0.9365934521342727, + "grad_norm": 0.23685111105442047, + "learning_rate": 3.1744716121011187e-07, + "loss": 0.4815, + "step": 11300 + }, + { + "epoch": 0.9366763365105678, + "grad_norm": 0.24656155705451965, + "learning_rate": 3.170327393286366e-07, + "loss": 0.4863, + "step": 11301 + }, + { + "epoch": 0.9367592208868628, + "grad_norm": 0.24169433116912842, + "learning_rate": 3.1661831744716126e-07, + "loss": 0.514, + "step": 11302 + }, + { + "epoch": 0.9368421052631579, + "grad_norm": 0.2400898039340973, + "learning_rate": 3.162038955656859e-07, + "loss": 0.5019, + "step": 11303 + }, + { + "epoch": 0.9369249896394529, + "grad_norm": 0.2449289709329605, + "learning_rate": 3.1578947368421055e-07, + "loss": 0.49, + "step": 11304 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 0.22506403923034668, + "learning_rate": 3.153750518027352e-07, + "loss": 0.4379, + "step": 11305 + }, + { + "epoch": 0.9370907583920431, + "grad_norm": 0.24068528413772583, + "learning_rate": 3.149606299212599e-07, + "loss": 0.4943, + "step": 11306 + }, + { + "epoch": 0.9371736427683381, + "grad_norm": 0.2578258812427521, + "learning_rate": 3.1454620803978453e-07, + "loss": 0.5064, + "step": 11307 + }, + { + "epoch": 0.9372565271446333, + "grad_norm": 0.23578381538391113, + "learning_rate": 3.141317861583092e-07, + "loss": 0.4865, + "step": 11308 + }, + { + "epoch": 0.9373394115209283, + "grad_norm": 0.2432047724723816, + "learning_rate": 3.137173642768338e-07, + "loss": 0.4754, + "step": 11309 + }, + { + "epoch": 0.9374222958972234, + "grad_norm": 0.22130563855171204, + "learning_rate": 3.1330294239535846e-07, + "loss": 0.4728, + "step": 11310 + }, + { + "epoch": 0.9375051802735185, + "grad_norm": 0.22070161998271942, + "learning_rate": 3.128885205138832e-07, + "loss": 0.4504, + "step": 11311 + }, + { + "epoch": 0.9375880646498135, + "grad_norm": 0.2393118143081665, + "learning_rate": 3.124740986324078e-07, + "loss": 0.4829, + "step": 11312 + }, + { + "epoch": 0.9376709490261086, + "grad_norm": 0.26888424158096313, + "learning_rate": 3.120596767509325e-07, + "loss": 0.4893, + "step": 11313 + }, + { + "epoch": 0.9377538334024036, + "grad_norm": 0.2132347822189331, + "learning_rate": 3.1164525486945714e-07, + "loss": 0.4606, + "step": 11314 + }, + { + "epoch": 0.9378367177786987, + "grad_norm": 0.2436733841896057, + "learning_rate": 3.112308329879818e-07, + "loss": 0.4817, + "step": 11315 + }, + { + "epoch": 0.9379196021549938, + "grad_norm": 0.22751177847385406, + "learning_rate": 3.1081641110650643e-07, + "loss": 0.467, + "step": 11316 + }, + { + "epoch": 0.9380024865312888, + "grad_norm": 0.24104782938957214, + "learning_rate": 3.1040198922503107e-07, + "loss": 0.5129, + "step": 11317 + }, + { + "epoch": 0.9380853709075839, + "grad_norm": 0.24536724388599396, + "learning_rate": 3.0998756734355577e-07, + "loss": 0.5202, + "step": 11318 + }, + { + "epoch": 0.938168255283879, + "grad_norm": 0.2526252567768097, + "learning_rate": 3.095731454620804e-07, + "loss": 0.5179, + "step": 11319 + }, + { + "epoch": 0.938251139660174, + "grad_norm": 0.22274385392665863, + "learning_rate": 3.0915872358060506e-07, + "loss": 0.453, + "step": 11320 + }, + { + "epoch": 0.9383340240364691, + "grad_norm": 0.22125916182994843, + "learning_rate": 3.0874430169912975e-07, + "loss": 0.4717, + "step": 11321 + }, + { + "epoch": 0.9384169084127642, + "grad_norm": 0.2408905178308487, + "learning_rate": 3.083298798176544e-07, + "loss": 0.4965, + "step": 11322 + }, + { + "epoch": 0.9384997927890593, + "grad_norm": 0.24302947521209717, + "learning_rate": 3.0791545793617904e-07, + "loss": 0.507, + "step": 11323 + }, + { + "epoch": 0.9385826771653544, + "grad_norm": 0.22437255084514618, + "learning_rate": 3.0750103605470374e-07, + "loss": 0.4705, + "step": 11324 + }, + { + "epoch": 0.9386655615416494, + "grad_norm": 0.23459234833717346, + "learning_rate": 3.070866141732284e-07, + "loss": 0.4681, + "step": 11325 + }, + { + "epoch": 0.9387484459179445, + "grad_norm": 0.24213743209838867, + "learning_rate": 3.06672192291753e-07, + "loss": 0.4646, + "step": 11326 + }, + { + "epoch": 0.9388313302942395, + "grad_norm": 0.25729629397392273, + "learning_rate": 3.062577704102777e-07, + "loss": 0.5471, + "step": 11327 + }, + { + "epoch": 0.9389142146705346, + "grad_norm": 0.24193982779979706, + "learning_rate": 3.0584334852880236e-07, + "loss": 0.4556, + "step": 11328 + }, + { + "epoch": 0.9389970990468297, + "grad_norm": 0.22777986526489258, + "learning_rate": 3.05428926647327e-07, + "loss": 0.4746, + "step": 11329 + }, + { + "epoch": 0.9390799834231247, + "grad_norm": 0.25708168745040894, + "learning_rate": 3.0501450476585165e-07, + "loss": 0.5124, + "step": 11330 + }, + { + "epoch": 0.9391628677994198, + "grad_norm": 0.23659205436706543, + "learning_rate": 3.046000828843763e-07, + "loss": 0.4732, + "step": 11331 + }, + { + "epoch": 0.9392457521757149, + "grad_norm": 0.23233358561992645, + "learning_rate": 3.04185661002901e-07, + "loss": 0.4607, + "step": 11332 + }, + { + "epoch": 0.9393286365520099, + "grad_norm": 0.22662292420864105, + "learning_rate": 3.0377123912142563e-07, + "loss": 0.4858, + "step": 11333 + }, + { + "epoch": 0.939411520928305, + "grad_norm": 0.23942895233631134, + "learning_rate": 3.033568172399503e-07, + "loss": 0.5111, + "step": 11334 + }, + { + "epoch": 0.9394944053046, + "grad_norm": 0.22836396098136902, + "learning_rate": 3.029423953584749e-07, + "loss": 0.4714, + "step": 11335 + }, + { + "epoch": 0.9395772896808952, + "grad_norm": 0.22767077386379242, + "learning_rate": 3.025279734769996e-07, + "loss": 0.4265, + "step": 11336 + }, + { + "epoch": 0.9396601740571903, + "grad_norm": 0.2424718290567398, + "learning_rate": 3.0211355159552426e-07, + "loss": 0.4833, + "step": 11337 + }, + { + "epoch": 0.9397430584334853, + "grad_norm": 0.23284046351909637, + "learning_rate": 3.016991297140489e-07, + "loss": 0.512, + "step": 11338 + }, + { + "epoch": 0.9398259428097804, + "grad_norm": 0.2598162889480591, + "learning_rate": 3.012847078325736e-07, + "loss": 0.5313, + "step": 11339 + }, + { + "epoch": 0.9399088271860754, + "grad_norm": 0.23142947256565094, + "learning_rate": 3.0087028595109825e-07, + "loss": 0.5176, + "step": 11340 + }, + { + "epoch": 0.9399917115623705, + "grad_norm": 0.2542782127857208, + "learning_rate": 3.004558640696229e-07, + "loss": 0.5205, + "step": 11341 + }, + { + "epoch": 0.9400745959386656, + "grad_norm": 0.24072414636611938, + "learning_rate": 3.000414421881476e-07, + "loss": 0.4931, + "step": 11342 + }, + { + "epoch": 0.9401574803149606, + "grad_norm": 0.23721016943454742, + "learning_rate": 2.9962702030667223e-07, + "loss": 0.4576, + "step": 11343 + }, + { + "epoch": 0.9402403646912557, + "grad_norm": 0.25096338987350464, + "learning_rate": 2.9921259842519687e-07, + "loss": 0.5061, + "step": 11344 + }, + { + "epoch": 0.9403232490675507, + "grad_norm": 0.2582290768623352, + "learning_rate": 2.987981765437215e-07, + "loss": 0.527, + "step": 11345 + }, + { + "epoch": 0.9404061334438458, + "grad_norm": 0.25055477023124695, + "learning_rate": 2.983837546622462e-07, + "loss": 0.4756, + "step": 11346 + }, + { + "epoch": 0.9404890178201409, + "grad_norm": 0.24350477755069733, + "learning_rate": 2.9796933278077086e-07, + "loss": 0.4997, + "step": 11347 + }, + { + "epoch": 0.9405719021964359, + "grad_norm": 0.23364165425300598, + "learning_rate": 2.975549108992955e-07, + "loss": 0.5008, + "step": 11348 + }, + { + "epoch": 0.940654786572731, + "grad_norm": 0.26279783248901367, + "learning_rate": 2.9714048901782014e-07, + "loss": 0.4795, + "step": 11349 + }, + { + "epoch": 0.9407376709490262, + "grad_norm": 0.2358272671699524, + "learning_rate": 2.967260671363448e-07, + "loss": 0.4962, + "step": 11350 + }, + { + "epoch": 0.9408205553253212, + "grad_norm": 0.26198670268058777, + "learning_rate": 2.963116452548695e-07, + "loss": 0.5061, + "step": 11351 + }, + { + "epoch": 0.9409034397016163, + "grad_norm": 0.24441006779670715, + "learning_rate": 2.9589722337339413e-07, + "loss": 0.4998, + "step": 11352 + }, + { + "epoch": 0.9409863240779113, + "grad_norm": 0.22339724004268646, + "learning_rate": 2.954828014919188e-07, + "loss": 0.4479, + "step": 11353 + }, + { + "epoch": 0.9410692084542064, + "grad_norm": 0.24314050376415253, + "learning_rate": 2.9506837961044347e-07, + "loss": 0.5336, + "step": 11354 + }, + { + "epoch": 0.9411520928305015, + "grad_norm": 0.23736721277236938, + "learning_rate": 2.946539577289681e-07, + "loss": 0.4626, + "step": 11355 + }, + { + "epoch": 0.9412349772067965, + "grad_norm": 0.24808703362941742, + "learning_rate": 2.942395358474928e-07, + "loss": 0.5331, + "step": 11356 + }, + { + "epoch": 0.9413178615830916, + "grad_norm": 0.2347869724035263, + "learning_rate": 2.9382511396601745e-07, + "loss": 0.4771, + "step": 11357 + }, + { + "epoch": 0.9414007459593866, + "grad_norm": 0.23697887361049652, + "learning_rate": 2.934106920845421e-07, + "loss": 0.4942, + "step": 11358 + }, + { + "epoch": 0.9414836303356817, + "grad_norm": 0.22283531725406647, + "learning_rate": 2.9299627020306674e-07, + "loss": 0.4578, + "step": 11359 + }, + { + "epoch": 0.9415665147119768, + "grad_norm": 0.21367840468883514, + "learning_rate": 2.925818483215914e-07, + "loss": 0.439, + "step": 11360 + }, + { + "epoch": 0.9416493990882718, + "grad_norm": 0.23132097721099854, + "learning_rate": 2.921674264401161e-07, + "loss": 0.4719, + "step": 11361 + }, + { + "epoch": 0.9417322834645669, + "grad_norm": 0.2306050956249237, + "learning_rate": 2.917530045586407e-07, + "loss": 0.4878, + "step": 11362 + }, + { + "epoch": 0.9418151678408619, + "grad_norm": 0.2302679419517517, + "learning_rate": 2.9133858267716536e-07, + "loss": 0.4586, + "step": 11363 + }, + { + "epoch": 0.9418980522171571, + "grad_norm": 0.23493757843971252, + "learning_rate": 2.9092416079569e-07, + "loss": 0.52, + "step": 11364 + }, + { + "epoch": 0.9419809365934522, + "grad_norm": 0.2530137300491333, + "learning_rate": 2.905097389142147e-07, + "loss": 0.488, + "step": 11365 + }, + { + "epoch": 0.9420638209697472, + "grad_norm": 0.2384846955537796, + "learning_rate": 2.9009531703273935e-07, + "loss": 0.4962, + "step": 11366 + }, + { + "epoch": 0.9421467053460423, + "grad_norm": 0.23425965011119843, + "learning_rate": 2.89680895151264e-07, + "loss": 0.4869, + "step": 11367 + }, + { + "epoch": 0.9422295897223374, + "grad_norm": 0.24937152862548828, + "learning_rate": 2.892664732697887e-07, + "loss": 0.5143, + "step": 11368 + }, + { + "epoch": 0.9423124740986324, + "grad_norm": 0.2213941514492035, + "learning_rate": 2.8885205138831333e-07, + "loss": 0.4633, + "step": 11369 + }, + { + "epoch": 0.9423953584749275, + "grad_norm": 0.24144314229488373, + "learning_rate": 2.88437629506838e-07, + "loss": 0.4917, + "step": 11370 + }, + { + "epoch": 0.9424782428512225, + "grad_norm": 0.24271917343139648, + "learning_rate": 2.8802320762536267e-07, + "loss": 0.501, + "step": 11371 + }, + { + "epoch": 0.9425611272275176, + "grad_norm": 0.22196322679519653, + "learning_rate": 2.876087857438873e-07, + "loss": 0.4579, + "step": 11372 + }, + { + "epoch": 0.9426440116038127, + "grad_norm": 0.209156796336174, + "learning_rate": 2.8719436386241196e-07, + "loss": 0.4629, + "step": 11373 + }, + { + "epoch": 0.9427268959801077, + "grad_norm": 0.22612427175045013, + "learning_rate": 2.867799419809366e-07, + "loss": 0.4677, + "step": 11374 + }, + { + "epoch": 0.9428097803564028, + "grad_norm": 0.24067126214504242, + "learning_rate": 2.863655200994613e-07, + "loss": 0.4709, + "step": 11375 + }, + { + "epoch": 0.9428926647326978, + "grad_norm": 0.22671155631542206, + "learning_rate": 2.8595109821798594e-07, + "loss": 0.5042, + "step": 11376 + }, + { + "epoch": 0.942975549108993, + "grad_norm": 0.23921671509742737, + "learning_rate": 2.855366763365106e-07, + "loss": 0.4898, + "step": 11377 + }, + { + "epoch": 0.9430584334852881, + "grad_norm": 0.23017215728759766, + "learning_rate": 2.8512225445503523e-07, + "loss": 0.5055, + "step": 11378 + }, + { + "epoch": 0.9431413178615831, + "grad_norm": 0.237370103597641, + "learning_rate": 2.8470783257355987e-07, + "loss": 0.4742, + "step": 11379 + }, + { + "epoch": 0.9432242022378782, + "grad_norm": 0.26176655292510986, + "learning_rate": 2.8429341069208457e-07, + "loss": 0.4884, + "step": 11380 + }, + { + "epoch": 0.9433070866141732, + "grad_norm": 0.22737936675548553, + "learning_rate": 2.838789888106092e-07, + "loss": 0.4674, + "step": 11381 + }, + { + "epoch": 0.9433899709904683, + "grad_norm": 0.2278457134962082, + "learning_rate": 2.8346456692913386e-07, + "loss": 0.4479, + "step": 11382 + }, + { + "epoch": 0.9434728553667634, + "grad_norm": 0.2417507767677307, + "learning_rate": 2.8305014504765855e-07, + "loss": 0.4943, + "step": 11383 + }, + { + "epoch": 0.9435557397430584, + "grad_norm": 0.24250000715255737, + "learning_rate": 2.826357231661832e-07, + "loss": 0.4939, + "step": 11384 + }, + { + "epoch": 0.9436386241193535, + "grad_norm": 0.2167283147573471, + "learning_rate": 2.8222130128470784e-07, + "loss": 0.4703, + "step": 11385 + }, + { + "epoch": 0.9437215084956486, + "grad_norm": 0.24004505574703217, + "learning_rate": 2.8180687940323254e-07, + "loss": 0.5061, + "step": 11386 + }, + { + "epoch": 0.9438043928719436, + "grad_norm": 0.2395767718553543, + "learning_rate": 2.813924575217572e-07, + "loss": 0.5012, + "step": 11387 + }, + { + "epoch": 0.9438872772482387, + "grad_norm": 0.32118624448776245, + "learning_rate": 2.809780356402818e-07, + "loss": 0.5537, + "step": 11388 + }, + { + "epoch": 0.9439701616245337, + "grad_norm": 0.22297286987304688, + "learning_rate": 2.805636137588065e-07, + "loss": 0.481, + "step": 11389 + }, + { + "epoch": 0.9440530460008288, + "grad_norm": 0.21917064487934113, + "learning_rate": 2.8014919187733116e-07, + "loss": 0.4146, + "step": 11390 + }, + { + "epoch": 0.944135930377124, + "grad_norm": 0.22087238729000092, + "learning_rate": 2.797347699958558e-07, + "loss": 0.4575, + "step": 11391 + }, + { + "epoch": 0.944218814753419, + "grad_norm": 0.2413453608751297, + "learning_rate": 2.7932034811438045e-07, + "loss": 0.5, + "step": 11392 + }, + { + "epoch": 0.9443016991297141, + "grad_norm": 0.23249198496341705, + "learning_rate": 2.789059262329051e-07, + "loss": 0.4527, + "step": 11393 + }, + { + "epoch": 0.9443845835060091, + "grad_norm": 0.2319396436214447, + "learning_rate": 2.784915043514298e-07, + "loss": 0.446, + "step": 11394 + }, + { + "epoch": 0.9444674678823042, + "grad_norm": 0.24005712568759918, + "learning_rate": 2.7807708246995444e-07, + "loss": 0.5027, + "step": 11395 + }, + { + "epoch": 0.9445503522585993, + "grad_norm": 0.22261539101600647, + "learning_rate": 2.776626605884791e-07, + "loss": 0.4588, + "step": 11396 + }, + { + "epoch": 0.9446332366348943, + "grad_norm": 0.2883044481277466, + "learning_rate": 2.772482387070037e-07, + "loss": 0.4493, + "step": 11397 + }, + { + "epoch": 0.9447161210111894, + "grad_norm": 0.24070174992084503, + "learning_rate": 2.768338168255284e-07, + "loss": 0.4775, + "step": 11398 + }, + { + "epoch": 0.9447990053874845, + "grad_norm": 0.23063258826732635, + "learning_rate": 2.7641939494405306e-07, + "loss": 0.4586, + "step": 11399 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 0.2444099336862564, + "learning_rate": 2.760049730625777e-07, + "loss": 0.5311, + "step": 11400 + }, + { + "epoch": 0.9449647741400746, + "grad_norm": 0.22428850829601288, + "learning_rate": 2.755905511811024e-07, + "loss": 0.4476, + "step": 11401 + }, + { + "epoch": 0.9450476585163696, + "grad_norm": 0.2607802748680115, + "learning_rate": 2.7517612929962705e-07, + "loss": 0.4797, + "step": 11402 + }, + { + "epoch": 0.9451305428926647, + "grad_norm": 0.2296757847070694, + "learning_rate": 2.7476170741815174e-07, + "loss": 0.4868, + "step": 11403 + }, + { + "epoch": 0.9452134272689598, + "grad_norm": 0.23974984884262085, + "learning_rate": 2.743472855366764e-07, + "loss": 0.5251, + "step": 11404 + }, + { + "epoch": 0.9452963116452548, + "grad_norm": 0.23913857340812683, + "learning_rate": 2.7393286365520103e-07, + "loss": 0.46, + "step": 11405 + }, + { + "epoch": 0.94537919602155, + "grad_norm": 0.2418556809425354, + "learning_rate": 2.7351844177372567e-07, + "loss": 0.5071, + "step": 11406 + }, + { + "epoch": 0.945462080397845, + "grad_norm": 0.2437872439622879, + "learning_rate": 2.731040198922503e-07, + "loss": 0.478, + "step": 11407 + }, + { + "epoch": 0.9455449647741401, + "grad_norm": 0.2305273711681366, + "learning_rate": 2.72689598010775e-07, + "loss": 0.4487, + "step": 11408 + }, + { + "epoch": 0.9456278491504352, + "grad_norm": 0.23889665305614471, + "learning_rate": 2.7227517612929966e-07, + "loss": 0.5135, + "step": 11409 + }, + { + "epoch": 0.9457107335267302, + "grad_norm": 0.23581673204898834, + "learning_rate": 2.718607542478243e-07, + "loss": 0.4963, + "step": 11410 + }, + { + "epoch": 0.9457936179030253, + "grad_norm": 0.2370905727148056, + "learning_rate": 2.7144633236634894e-07, + "loss": 0.5005, + "step": 11411 + }, + { + "epoch": 0.9458765022793203, + "grad_norm": 0.2441290318965912, + "learning_rate": 2.710319104848736e-07, + "loss": 0.497, + "step": 11412 + }, + { + "epoch": 0.9459593866556154, + "grad_norm": 0.2332525998353958, + "learning_rate": 2.706174886033983e-07, + "loss": 0.472, + "step": 11413 + }, + { + "epoch": 0.9460422710319105, + "grad_norm": 0.21993742883205414, + "learning_rate": 2.7020306672192293e-07, + "loss": 0.4658, + "step": 11414 + }, + { + "epoch": 0.9461251554082055, + "grad_norm": 0.22566840052604675, + "learning_rate": 2.6978864484044757e-07, + "loss": 0.4529, + "step": 11415 + }, + { + "epoch": 0.9462080397845006, + "grad_norm": 0.24049082398414612, + "learning_rate": 2.6937422295897227e-07, + "loss": 0.4888, + "step": 11416 + }, + { + "epoch": 0.9462909241607957, + "grad_norm": 0.22641272842884064, + "learning_rate": 2.689598010774969e-07, + "loss": 0.4778, + "step": 11417 + }, + { + "epoch": 0.9463738085370907, + "grad_norm": 0.23470020294189453, + "learning_rate": 2.685453791960216e-07, + "loss": 0.5214, + "step": 11418 + }, + { + "epoch": 0.9464566929133859, + "grad_norm": 0.2407173216342926, + "learning_rate": 2.6813095731454625e-07, + "loss": 0.5094, + "step": 11419 + }, + { + "epoch": 0.9465395772896809, + "grad_norm": 0.9775302410125732, + "learning_rate": 2.677165354330709e-07, + "loss": 0.5005, + "step": 11420 + }, + { + "epoch": 0.946622461665976, + "grad_norm": 0.2431705743074417, + "learning_rate": 2.6730211355159554e-07, + "loss": 0.4817, + "step": 11421 + }, + { + "epoch": 0.9467053460422711, + "grad_norm": 0.2645508050918579, + "learning_rate": 2.6688769167012023e-07, + "loss": 0.5659, + "step": 11422 + }, + { + "epoch": 0.9467882304185661, + "grad_norm": 0.22001007199287415, + "learning_rate": 2.664732697886449e-07, + "loss": 0.4682, + "step": 11423 + }, + { + "epoch": 0.9468711147948612, + "grad_norm": 0.24462006986141205, + "learning_rate": 2.660588479071695e-07, + "loss": 0.5372, + "step": 11424 + }, + { + "epoch": 0.9469539991711562, + "grad_norm": 0.24845804274082184, + "learning_rate": 2.6564442602569417e-07, + "loss": 0.5066, + "step": 11425 + }, + { + "epoch": 0.9470368835474513, + "grad_norm": 0.23198600113391876, + "learning_rate": 2.652300041442188e-07, + "loss": 0.4258, + "step": 11426 + }, + { + "epoch": 0.9471197679237464, + "grad_norm": 0.23040978610515594, + "learning_rate": 2.648155822627435e-07, + "loss": 0.5097, + "step": 11427 + }, + { + "epoch": 0.9472026523000414, + "grad_norm": 0.2463664561510086, + "learning_rate": 2.6440116038126815e-07, + "loss": 0.5458, + "step": 11428 + }, + { + "epoch": 0.9472855366763365, + "grad_norm": 0.23443837463855743, + "learning_rate": 2.639867384997928e-07, + "loss": 0.4947, + "step": 11429 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 0.23702092468738556, + "learning_rate": 2.6357231661831744e-07, + "loss": 0.4916, + "step": 11430 + }, + { + "epoch": 0.9474513054289266, + "grad_norm": 0.24235638976097107, + "learning_rate": 2.6315789473684213e-07, + "loss": 0.4629, + "step": 11431 + }, + { + "epoch": 0.9475341898052217, + "grad_norm": 0.2208169549703598, + "learning_rate": 2.627434728553668e-07, + "loss": 0.4804, + "step": 11432 + }, + { + "epoch": 0.9476170741815167, + "grad_norm": 0.24623514711856842, + "learning_rate": 2.6232905097389147e-07, + "loss": 0.5052, + "step": 11433 + }, + { + "epoch": 0.9476999585578119, + "grad_norm": 0.22313140332698822, + "learning_rate": 2.619146290924161e-07, + "loss": 0.4881, + "step": 11434 + }, + { + "epoch": 0.947782842934107, + "grad_norm": 0.235682874917984, + "learning_rate": 2.6150020721094076e-07, + "loss": 0.5098, + "step": 11435 + }, + { + "epoch": 0.947865727310402, + "grad_norm": 0.22384148836135864, + "learning_rate": 2.610857853294654e-07, + "loss": 0.4644, + "step": 11436 + }, + { + "epoch": 0.9479486116866971, + "grad_norm": 0.27198439836502075, + "learning_rate": 2.606713634479901e-07, + "loss": 0.5626, + "step": 11437 + }, + { + "epoch": 0.9480314960629921, + "grad_norm": 0.23093442618846893, + "learning_rate": 2.6025694156651474e-07, + "loss": 0.5129, + "step": 11438 + }, + { + "epoch": 0.9481143804392872, + "grad_norm": 0.23826265335083008, + "learning_rate": 2.598425196850394e-07, + "loss": 0.4718, + "step": 11439 + }, + { + "epoch": 0.9481972648155823, + "grad_norm": 0.2321157455444336, + "learning_rate": 2.5942809780356403e-07, + "loss": 0.4892, + "step": 11440 + }, + { + "epoch": 0.9482801491918773, + "grad_norm": 0.2344782054424286, + "learning_rate": 2.5901367592208873e-07, + "loss": 0.4951, + "step": 11441 + }, + { + "epoch": 0.9483630335681724, + "grad_norm": 0.25227710604667664, + "learning_rate": 2.5859925404061337e-07, + "loss": 0.5224, + "step": 11442 + }, + { + "epoch": 0.9484459179444674, + "grad_norm": 0.23970411717891693, + "learning_rate": 2.58184832159138e-07, + "loss": 0.5106, + "step": 11443 + }, + { + "epoch": 0.9485288023207625, + "grad_norm": 0.23761184513568878, + "learning_rate": 2.5777041027766266e-07, + "loss": 0.5029, + "step": 11444 + }, + { + "epoch": 0.9486116866970576, + "grad_norm": 0.22483637928962708, + "learning_rate": 2.573559883961873e-07, + "loss": 0.4712, + "step": 11445 + }, + { + "epoch": 0.9486945710733526, + "grad_norm": 0.26245516538619995, + "learning_rate": 2.56941566514712e-07, + "loss": 0.5189, + "step": 11446 + }, + { + "epoch": 0.9487774554496478, + "grad_norm": 0.2404133826494217, + "learning_rate": 2.5652714463323664e-07, + "loss": 0.496, + "step": 11447 + }, + { + "epoch": 0.9488603398259428, + "grad_norm": 0.23976914584636688, + "learning_rate": 2.5611272275176134e-07, + "loss": 0.4656, + "step": 11448 + }, + { + "epoch": 0.9489432242022379, + "grad_norm": 0.23578912019729614, + "learning_rate": 2.55698300870286e-07, + "loss": 0.4481, + "step": 11449 + }, + { + "epoch": 0.949026108578533, + "grad_norm": 0.2291191667318344, + "learning_rate": 2.552838789888106e-07, + "loss": 0.4552, + "step": 11450 + }, + { + "epoch": 0.949108992954828, + "grad_norm": 0.24327057600021362, + "learning_rate": 2.548694571073353e-07, + "loss": 0.4643, + "step": 11451 + }, + { + "epoch": 0.9491918773311231, + "grad_norm": 0.2500194311141968, + "learning_rate": 2.5445503522585996e-07, + "loss": 0.501, + "step": 11452 + }, + { + "epoch": 0.9492747617074182, + "grad_norm": 0.23323304951190948, + "learning_rate": 2.540406133443846e-07, + "loss": 0.4709, + "step": 11453 + }, + { + "epoch": 0.9493576460837132, + "grad_norm": 0.2322712242603302, + "learning_rate": 2.5362619146290925e-07, + "loss": 0.4848, + "step": 11454 + }, + { + "epoch": 0.9494405304600083, + "grad_norm": 0.2299736887216568, + "learning_rate": 2.532117695814339e-07, + "loss": 0.4707, + "step": 11455 + }, + { + "epoch": 0.9495234148363033, + "grad_norm": 0.23793654143810272, + "learning_rate": 2.527973476999586e-07, + "loss": 0.5244, + "step": 11456 + }, + { + "epoch": 0.9496062992125984, + "grad_norm": 0.22004368901252747, + "learning_rate": 2.5238292581848324e-07, + "loss": 0.4597, + "step": 11457 + }, + { + "epoch": 0.9496891835888935, + "grad_norm": 0.23030921816825867, + "learning_rate": 2.519685039370079e-07, + "loss": 0.4642, + "step": 11458 + }, + { + "epoch": 0.9497720679651885, + "grad_norm": 0.2504120171070099, + "learning_rate": 2.515540820555325e-07, + "loss": 0.4807, + "step": 11459 + }, + { + "epoch": 0.9498549523414836, + "grad_norm": 0.2422703355550766, + "learning_rate": 2.511396601740572e-07, + "loss": 0.4761, + "step": 11460 + }, + { + "epoch": 0.9499378367177786, + "grad_norm": 0.2204795777797699, + "learning_rate": 2.5072523829258186e-07, + "loss": 0.4607, + "step": 11461 + }, + { + "epoch": 0.9500207210940738, + "grad_norm": 0.2514796257019043, + "learning_rate": 2.503108164111065e-07, + "loss": 0.4645, + "step": 11462 + }, + { + "epoch": 0.9501036054703689, + "grad_norm": 0.2392696887254715, + "learning_rate": 2.498963945296312e-07, + "loss": 0.4775, + "step": 11463 + }, + { + "epoch": 0.9501864898466639, + "grad_norm": 0.22657640278339386, + "learning_rate": 2.4948197264815585e-07, + "loss": 0.4722, + "step": 11464 + }, + { + "epoch": 0.950269374222959, + "grad_norm": 0.2585420608520508, + "learning_rate": 2.490675507666805e-07, + "loss": 0.5088, + "step": 11465 + }, + { + "epoch": 0.9503522585992541, + "grad_norm": 0.24173583090305328, + "learning_rate": 2.486531288852052e-07, + "loss": 0.5335, + "step": 11466 + }, + { + "epoch": 0.9504351429755491, + "grad_norm": 0.24362331628799438, + "learning_rate": 2.4823870700372983e-07, + "loss": 0.4995, + "step": 11467 + }, + { + "epoch": 0.9505180273518442, + "grad_norm": 0.22758938372135162, + "learning_rate": 2.4782428512225447e-07, + "loss": 0.4699, + "step": 11468 + }, + { + "epoch": 0.9506009117281392, + "grad_norm": 0.22757266461849213, + "learning_rate": 2.474098632407791e-07, + "loss": 0.4483, + "step": 11469 + }, + { + "epoch": 0.9506837961044343, + "grad_norm": 0.25424155592918396, + "learning_rate": 2.469954413593038e-07, + "loss": 0.4892, + "step": 11470 + }, + { + "epoch": 0.9507666804807294, + "grad_norm": 0.22376203536987305, + "learning_rate": 2.4658101947782846e-07, + "loss": 0.433, + "step": 11471 + }, + { + "epoch": 0.9508495648570244, + "grad_norm": 0.23910333216190338, + "learning_rate": 2.461665975963531e-07, + "loss": 0.4968, + "step": 11472 + }, + { + "epoch": 0.9509324492333195, + "grad_norm": 0.24683396518230438, + "learning_rate": 2.4575217571487774e-07, + "loss": 0.5425, + "step": 11473 + }, + { + "epoch": 0.9510153336096145, + "grad_norm": 0.2437259405851364, + "learning_rate": 2.453377538334024e-07, + "loss": 0.5357, + "step": 11474 + }, + { + "epoch": 0.9510982179859097, + "grad_norm": 0.24964791536331177, + "learning_rate": 2.449233319519271e-07, + "loss": 0.5293, + "step": 11475 + }, + { + "epoch": 0.9511811023622048, + "grad_norm": 0.23889335989952087, + "learning_rate": 2.4450891007045173e-07, + "loss": 0.5303, + "step": 11476 + }, + { + "epoch": 0.9512639867384998, + "grad_norm": 0.2193920761346817, + "learning_rate": 2.4409448818897637e-07, + "loss": 0.4535, + "step": 11477 + }, + { + "epoch": 0.9513468711147949, + "grad_norm": 0.24763436615467072, + "learning_rate": 2.4368006630750107e-07, + "loss": 0.4949, + "step": 11478 + }, + { + "epoch": 0.9514297554910899, + "grad_norm": 0.2315860390663147, + "learning_rate": 2.432656444260257e-07, + "loss": 0.4983, + "step": 11479 + }, + { + "epoch": 0.951512639867385, + "grad_norm": 0.23822708427906036, + "learning_rate": 2.428512225445504e-07, + "loss": 0.4738, + "step": 11480 + }, + { + "epoch": 0.9515955242436801, + "grad_norm": 0.24683980643749237, + "learning_rate": 2.4243680066307505e-07, + "loss": 0.5078, + "step": 11481 + }, + { + "epoch": 0.9516784086199751, + "grad_norm": 0.2627587914466858, + "learning_rate": 2.420223787815997e-07, + "loss": 0.498, + "step": 11482 + }, + { + "epoch": 0.9517612929962702, + "grad_norm": 0.24849840998649597, + "learning_rate": 2.4160795690012434e-07, + "loss": 0.4799, + "step": 11483 + }, + { + "epoch": 0.9518441773725653, + "grad_norm": 0.25037574768066406, + "learning_rate": 2.4119353501864903e-07, + "loss": 0.4839, + "step": 11484 + }, + { + "epoch": 0.9519270617488603, + "grad_norm": 0.22121314704418182, + "learning_rate": 2.407791131371737e-07, + "loss": 0.4504, + "step": 11485 + }, + { + "epoch": 0.9520099461251554, + "grad_norm": 0.24444936215877533, + "learning_rate": 2.403646912556983e-07, + "loss": 0.5249, + "step": 11486 + }, + { + "epoch": 0.9520928305014504, + "grad_norm": 0.25873878598213196, + "learning_rate": 2.3995026937422297e-07, + "loss": 0.4916, + "step": 11487 + }, + { + "epoch": 0.9521757148777455, + "grad_norm": 0.2335178703069687, + "learning_rate": 2.395358474927476e-07, + "loss": 0.519, + "step": 11488 + }, + { + "epoch": 0.9522585992540407, + "grad_norm": 0.23868240416049957, + "learning_rate": 2.391214256112723e-07, + "loss": 0.4965, + "step": 11489 + }, + { + "epoch": 0.9523414836303357, + "grad_norm": 0.25725406408309937, + "learning_rate": 2.3870700372979695e-07, + "loss": 0.5198, + "step": 11490 + }, + { + "epoch": 0.9524243680066308, + "grad_norm": 0.230649933218956, + "learning_rate": 2.3829258184832162e-07, + "loss": 0.4942, + "step": 11491 + }, + { + "epoch": 0.9525072523829258, + "grad_norm": 0.239837646484375, + "learning_rate": 2.3787815996684626e-07, + "loss": 0.4814, + "step": 11492 + }, + { + "epoch": 0.9525901367592209, + "grad_norm": 0.22614091634750366, + "learning_rate": 2.374637380853709e-07, + "loss": 0.4842, + "step": 11493 + }, + { + "epoch": 0.952673021135516, + "grad_norm": 0.23631073534488678, + "learning_rate": 2.370493162038956e-07, + "loss": 0.5261, + "step": 11494 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 0.2368180751800537, + "learning_rate": 2.3663489432242025e-07, + "loss": 0.453, + "step": 11495 + }, + { + "epoch": 0.9528387898881061, + "grad_norm": 0.24035529792308807, + "learning_rate": 2.362204724409449e-07, + "loss": 0.5026, + "step": 11496 + }, + { + "epoch": 0.9529216742644011, + "grad_norm": 0.23279960453510284, + "learning_rate": 2.3580605055946956e-07, + "loss": 0.4919, + "step": 11497 + }, + { + "epoch": 0.9530045586406962, + "grad_norm": 0.23398230969905853, + "learning_rate": 2.3539162867799423e-07, + "loss": 0.4869, + "step": 11498 + }, + { + "epoch": 0.9530874430169913, + "grad_norm": 0.2521860897541046, + "learning_rate": 2.3497720679651887e-07, + "loss": 0.5105, + "step": 11499 + }, + { + "epoch": 0.9531703273932863, + "grad_norm": 0.24483869969844818, + "learning_rate": 2.3456278491504354e-07, + "loss": 0.5046, + "step": 11500 + }, + { + "epoch": 0.9532532117695814, + "grad_norm": 0.23811382055282593, + "learning_rate": 2.341483630335682e-07, + "loss": 0.4751, + "step": 11501 + }, + { + "epoch": 0.9533360961458766, + "grad_norm": 0.23791871964931488, + "learning_rate": 2.3373394115209283e-07, + "loss": 0.495, + "step": 11502 + }, + { + "epoch": 0.9534189805221716, + "grad_norm": 0.2494068145751953, + "learning_rate": 2.3331951927061753e-07, + "loss": 0.5028, + "step": 11503 + }, + { + "epoch": 0.9535018648984667, + "grad_norm": 0.25121548771858215, + "learning_rate": 2.3290509738914217e-07, + "loss": 0.5125, + "step": 11504 + }, + { + "epoch": 0.9535847492747617, + "grad_norm": 0.29218968749046326, + "learning_rate": 2.3249067550766681e-07, + "loss": 0.4647, + "step": 11505 + }, + { + "epoch": 0.9536676336510568, + "grad_norm": 0.22791144251823425, + "learning_rate": 2.3207625362619148e-07, + "loss": 0.4765, + "step": 11506 + }, + { + "epoch": 0.9537505180273519, + "grad_norm": 0.23343703150749207, + "learning_rate": 2.3166183174471613e-07, + "loss": 0.407, + "step": 11507 + }, + { + "epoch": 0.9538334024036469, + "grad_norm": 0.2413557469844818, + "learning_rate": 2.3124740986324082e-07, + "loss": 0.5103, + "step": 11508 + }, + { + "epoch": 0.953916286779942, + "grad_norm": 0.2250439077615738, + "learning_rate": 2.3083298798176547e-07, + "loss": 0.4627, + "step": 11509 + }, + { + "epoch": 0.953999171156237, + "grad_norm": 0.24006031453609467, + "learning_rate": 2.304185661002901e-07, + "loss": 0.4903, + "step": 11510 + }, + { + "epoch": 0.9540820555325321, + "grad_norm": 0.2530214190483093, + "learning_rate": 2.3000414421881475e-07, + "loss": 0.572, + "step": 11511 + }, + { + "epoch": 0.9541649399088272, + "grad_norm": 0.23022891581058502, + "learning_rate": 2.2958972233733942e-07, + "loss": 0.4989, + "step": 11512 + }, + { + "epoch": 0.9542478242851222, + "grad_norm": 0.23795513808727264, + "learning_rate": 2.291753004558641e-07, + "loss": 0.4922, + "step": 11513 + }, + { + "epoch": 0.9543307086614173, + "grad_norm": 0.24331046640872955, + "learning_rate": 2.2876087857438874e-07, + "loss": 0.5459, + "step": 11514 + }, + { + "epoch": 0.9544135930377123, + "grad_norm": 0.24965006113052368, + "learning_rate": 2.283464566929134e-07, + "loss": 0.4962, + "step": 11515 + }, + { + "epoch": 0.9544964774140074, + "grad_norm": 0.25401321053504944, + "learning_rate": 2.2793203481143805e-07, + "loss": 0.5234, + "step": 11516 + }, + { + "epoch": 0.9545793617903026, + "grad_norm": 0.23063895106315613, + "learning_rate": 2.2751761292996275e-07, + "loss": 0.4903, + "step": 11517 + }, + { + "epoch": 0.9546622461665976, + "grad_norm": 0.24711798131465912, + "learning_rate": 2.271031910484874e-07, + "loss": 0.5031, + "step": 11518 + }, + { + "epoch": 0.9547451305428927, + "grad_norm": 0.23872782289981842, + "learning_rate": 2.2668876916701204e-07, + "loss": 0.4768, + "step": 11519 + }, + { + "epoch": 0.9548280149191878, + "grad_norm": 0.23470158874988556, + "learning_rate": 2.2627434728553668e-07, + "loss": 0.4724, + "step": 11520 + }, + { + "epoch": 0.9549108992954828, + "grad_norm": 0.23377691209316254, + "learning_rate": 2.2585992540406135e-07, + "loss": 0.4911, + "step": 11521 + }, + { + "epoch": 0.9549937836717779, + "grad_norm": 0.23262900114059448, + "learning_rate": 2.2544550352258602e-07, + "loss": 0.4689, + "step": 11522 + }, + { + "epoch": 0.9550766680480729, + "grad_norm": 0.22857481241226196, + "learning_rate": 2.250310816411107e-07, + "loss": 0.4559, + "step": 11523 + }, + { + "epoch": 0.955159552424368, + "grad_norm": 0.27256226539611816, + "learning_rate": 2.2461665975963533e-07, + "loss": 0.4899, + "step": 11524 + }, + { + "epoch": 0.9552424368006631, + "grad_norm": 0.23273974657058716, + "learning_rate": 2.2420223787815998e-07, + "loss": 0.4904, + "step": 11525 + }, + { + "epoch": 0.9553253211769581, + "grad_norm": 0.2618629038333893, + "learning_rate": 2.2378781599668462e-07, + "loss": 0.4981, + "step": 11526 + }, + { + "epoch": 0.9554082055532532, + "grad_norm": 0.24410027265548706, + "learning_rate": 2.2337339411520932e-07, + "loss": 0.498, + "step": 11527 + }, + { + "epoch": 0.9554910899295482, + "grad_norm": 0.2294190675020218, + "learning_rate": 2.2295897223373396e-07, + "loss": 0.5037, + "step": 11528 + }, + { + "epoch": 0.9555739743058433, + "grad_norm": 0.24854129552841187, + "learning_rate": 2.225445503522586e-07, + "loss": 0.5191, + "step": 11529 + }, + { + "epoch": 0.9556568586821385, + "grad_norm": 0.23091191053390503, + "learning_rate": 2.2213012847078327e-07, + "loss": 0.4761, + "step": 11530 + }, + { + "epoch": 0.9557397430584335, + "grad_norm": 0.23049861192703247, + "learning_rate": 2.2171570658930792e-07, + "loss": 0.4806, + "step": 11531 + }, + { + "epoch": 0.9558226274347286, + "grad_norm": 0.28631240129470825, + "learning_rate": 2.2130128470783261e-07, + "loss": 0.5001, + "step": 11532 + }, + { + "epoch": 0.9559055118110236, + "grad_norm": 0.24992364645004272, + "learning_rate": 2.2088686282635726e-07, + "loss": 0.494, + "step": 11533 + }, + { + "epoch": 0.9559883961873187, + "grad_norm": 0.23707568645477295, + "learning_rate": 2.204724409448819e-07, + "loss": 0.5077, + "step": 11534 + }, + { + "epoch": 0.9560712805636138, + "grad_norm": 0.2294483482837677, + "learning_rate": 2.2005801906340654e-07, + "loss": 0.506, + "step": 11535 + }, + { + "epoch": 0.9561541649399088, + "grad_norm": 0.23395442962646484, + "learning_rate": 2.1964359718193124e-07, + "loss": 0.4992, + "step": 11536 + }, + { + "epoch": 0.9562370493162039, + "grad_norm": 0.28922855854034424, + "learning_rate": 2.1922917530045588e-07, + "loss": 0.522, + "step": 11537 + }, + { + "epoch": 0.956319933692499, + "grad_norm": 0.23162803053855896, + "learning_rate": 2.1881475341898055e-07, + "loss": 0.4992, + "step": 11538 + }, + { + "epoch": 0.956402818068794, + "grad_norm": 0.22779174149036407, + "learning_rate": 2.184003315375052e-07, + "loss": 0.4726, + "step": 11539 + }, + { + "epoch": 0.9564857024450891, + "grad_norm": 0.23975765705108643, + "learning_rate": 2.1798590965602984e-07, + "loss": 0.459, + "step": 11540 + }, + { + "epoch": 0.9565685868213841, + "grad_norm": 0.2500116527080536, + "learning_rate": 2.1757148777455454e-07, + "loss": 0.4844, + "step": 11541 + }, + { + "epoch": 0.9566514711976792, + "grad_norm": 0.2350892424583435, + "learning_rate": 2.1715706589307918e-07, + "loss": 0.5168, + "step": 11542 + }, + { + "epoch": 0.9567343555739743, + "grad_norm": 0.34779059886932373, + "learning_rate": 2.1674264401160382e-07, + "loss": 0.5117, + "step": 11543 + }, + { + "epoch": 0.9568172399502693, + "grad_norm": 0.24453336000442505, + "learning_rate": 2.163282221301285e-07, + "loss": 0.4964, + "step": 11544 + }, + { + "epoch": 0.9569001243265645, + "grad_norm": 0.260163813829422, + "learning_rate": 2.1591380024865314e-07, + "loss": 0.5178, + "step": 11545 + }, + { + "epoch": 0.9569830087028595, + "grad_norm": 0.23483802378177643, + "learning_rate": 2.154993783671778e-07, + "loss": 0.4503, + "step": 11546 + }, + { + "epoch": 0.9570658930791546, + "grad_norm": 0.21458134055137634, + "learning_rate": 2.1508495648570248e-07, + "loss": 0.4267, + "step": 11547 + }, + { + "epoch": 0.9571487774554497, + "grad_norm": 0.3438349962234497, + "learning_rate": 2.1467053460422712e-07, + "loss": 0.4916, + "step": 11548 + }, + { + "epoch": 0.9572316618317447, + "grad_norm": 0.2512207329273224, + "learning_rate": 2.1425611272275177e-07, + "loss": 0.5555, + "step": 11549 + }, + { + "epoch": 0.9573145462080398, + "grad_norm": 0.23165930807590485, + "learning_rate": 2.138416908412764e-07, + "loss": 0.4847, + "step": 11550 + }, + { + "epoch": 0.9573974305843349, + "grad_norm": 0.23015096783638, + "learning_rate": 2.134272689598011e-07, + "loss": 0.5031, + "step": 11551 + }, + { + "epoch": 0.9574803149606299, + "grad_norm": 0.23096735775470734, + "learning_rate": 2.1301284707832575e-07, + "loss": 0.4946, + "step": 11552 + }, + { + "epoch": 0.957563199336925, + "grad_norm": 0.2485814392566681, + "learning_rate": 2.1259842519685042e-07, + "loss": 0.5395, + "step": 11553 + }, + { + "epoch": 0.95764608371322, + "grad_norm": 0.23511578142642975, + "learning_rate": 2.1218400331537506e-07, + "loss": 0.5217, + "step": 11554 + }, + { + "epoch": 0.9577289680895151, + "grad_norm": 0.25680437684059143, + "learning_rate": 2.1176958143389973e-07, + "loss": 0.5458, + "step": 11555 + }, + { + "epoch": 0.9578118524658102, + "grad_norm": 0.23044374585151672, + "learning_rate": 2.113551595524244e-07, + "loss": 0.4716, + "step": 11556 + }, + { + "epoch": 0.9578947368421052, + "grad_norm": 0.2271798998117447, + "learning_rate": 2.1094073767094905e-07, + "loss": 0.4975, + "step": 11557 + }, + { + "epoch": 0.9579776212184004, + "grad_norm": 0.2291669398546219, + "learning_rate": 2.105263157894737e-07, + "loss": 0.4802, + "step": 11558 + }, + { + "epoch": 0.9580605055946954, + "grad_norm": 0.21892346441745758, + "learning_rate": 2.1011189390799836e-07, + "loss": 0.5009, + "step": 11559 + }, + { + "epoch": 0.9581433899709905, + "grad_norm": 0.2392626851797104, + "learning_rate": 2.0969747202652303e-07, + "loss": 0.4788, + "step": 11560 + }, + { + "epoch": 0.9582262743472856, + "grad_norm": 0.2307290881872177, + "learning_rate": 2.0928305014504767e-07, + "loss": 0.4819, + "step": 11561 + }, + { + "epoch": 0.9583091587235806, + "grad_norm": 0.2401381880044937, + "learning_rate": 2.0886862826357234e-07, + "loss": 0.5158, + "step": 11562 + }, + { + "epoch": 0.9583920430998757, + "grad_norm": 0.24652504920959473, + "learning_rate": 2.08454206382097e-07, + "loss": 0.5083, + "step": 11563 + }, + { + "epoch": 0.9584749274761707, + "grad_norm": 0.2285657376050949, + "learning_rate": 2.0803978450062163e-07, + "loss": 0.4625, + "step": 11564 + }, + { + "epoch": 0.9585578118524658, + "grad_norm": 0.22796174883842468, + "learning_rate": 2.0762536261914633e-07, + "loss": 0.4568, + "step": 11565 + }, + { + "epoch": 0.9586406962287609, + "grad_norm": 0.2518675625324249, + "learning_rate": 2.0721094073767097e-07, + "loss": 0.5081, + "step": 11566 + }, + { + "epoch": 0.9587235806050559, + "grad_norm": 0.24449212849140167, + "learning_rate": 2.0679651885619561e-07, + "loss": 0.4923, + "step": 11567 + }, + { + "epoch": 0.958806464981351, + "grad_norm": 0.22354218363761902, + "learning_rate": 2.0638209697472028e-07, + "loss": 0.474, + "step": 11568 + }, + { + "epoch": 0.9588893493576461, + "grad_norm": 0.2479730099439621, + "learning_rate": 2.0596767509324493e-07, + "loss": 0.5084, + "step": 11569 + }, + { + "epoch": 0.9589722337339411, + "grad_norm": 0.23699064552783966, + "learning_rate": 2.055532532117696e-07, + "loss": 0.5258, + "step": 11570 + }, + { + "epoch": 0.9590551181102362, + "grad_norm": 0.2332630306482315, + "learning_rate": 2.0513883133029427e-07, + "loss": 0.5, + "step": 11571 + }, + { + "epoch": 0.9591380024865312, + "grad_norm": 0.22753863036632538, + "learning_rate": 2.047244094488189e-07, + "loss": 0.4882, + "step": 11572 + }, + { + "epoch": 0.9592208868628264, + "grad_norm": 0.23604479432106018, + "learning_rate": 2.0430998756734356e-07, + "loss": 0.47, + "step": 11573 + }, + { + "epoch": 0.9593037712391215, + "grad_norm": 0.2312099188566208, + "learning_rate": 2.0389556568586825e-07, + "loss": 0.4991, + "step": 11574 + }, + { + "epoch": 0.9593866556154165, + "grad_norm": 0.22952425479888916, + "learning_rate": 2.034811438043929e-07, + "loss": 0.4353, + "step": 11575 + }, + { + "epoch": 0.9594695399917116, + "grad_norm": 0.23269960284233093, + "learning_rate": 2.0306672192291754e-07, + "loss": 0.4459, + "step": 11576 + }, + { + "epoch": 0.9595524243680066, + "grad_norm": 0.2479499876499176, + "learning_rate": 2.026523000414422e-07, + "loss": 0.5134, + "step": 11577 + }, + { + "epoch": 0.9596353087443017, + "grad_norm": 0.2259528785943985, + "learning_rate": 2.0223787815996685e-07, + "loss": 0.4825, + "step": 11578 + }, + { + "epoch": 0.9597181931205968, + "grad_norm": 0.23423464596271515, + "learning_rate": 2.0182345627849155e-07, + "loss": 0.4673, + "step": 11579 + }, + { + "epoch": 0.9598010774968918, + "grad_norm": 0.23612290620803833, + "learning_rate": 2.014090343970162e-07, + "loss": 0.4607, + "step": 11580 + }, + { + "epoch": 0.9598839618731869, + "grad_norm": 0.23022866249084473, + "learning_rate": 2.0099461251554084e-07, + "loss": 0.4824, + "step": 11581 + }, + { + "epoch": 0.9599668462494819, + "grad_norm": 0.25907281041145325, + "learning_rate": 2.0058019063406548e-07, + "loss": 0.5088, + "step": 11582 + }, + { + "epoch": 0.960049730625777, + "grad_norm": 0.24690519273281097, + "learning_rate": 2.0016576875259015e-07, + "loss": 0.4984, + "step": 11583 + }, + { + "epoch": 0.9601326150020721, + "grad_norm": 0.26572611927986145, + "learning_rate": 1.9975134687111482e-07, + "loss": 0.4495, + "step": 11584 + }, + { + "epoch": 0.9602154993783671, + "grad_norm": 0.23756366968154907, + "learning_rate": 1.9933692498963946e-07, + "loss": 0.4917, + "step": 11585 + }, + { + "epoch": 0.9602983837546623, + "grad_norm": 0.24758611619472504, + "learning_rate": 1.9892250310816413e-07, + "loss": 0.526, + "step": 11586 + }, + { + "epoch": 0.9603812681309574, + "grad_norm": 0.2469363659620285, + "learning_rate": 1.9850808122668878e-07, + "loss": 0.5222, + "step": 11587 + }, + { + "epoch": 0.9604641525072524, + "grad_norm": 0.24411101639270782, + "learning_rate": 1.9809365934521342e-07, + "loss": 0.5226, + "step": 11588 + }, + { + "epoch": 0.9605470368835475, + "grad_norm": 0.25218936800956726, + "learning_rate": 1.9767923746373812e-07, + "loss": 0.5206, + "step": 11589 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 0.22348524630069733, + "learning_rate": 1.9726481558226276e-07, + "loss": 0.498, + "step": 11590 + }, + { + "epoch": 0.9607128056361376, + "grad_norm": 0.24386516213417053, + "learning_rate": 1.968503937007874e-07, + "loss": 0.5173, + "step": 11591 + }, + { + "epoch": 0.9607956900124327, + "grad_norm": 0.23551739752292633, + "learning_rate": 1.9643597181931207e-07, + "loss": 0.4845, + "step": 11592 + }, + { + "epoch": 0.9608785743887277, + "grad_norm": 0.2320747673511505, + "learning_rate": 1.9602154993783672e-07, + "loss": 0.5055, + "step": 11593 + }, + { + "epoch": 0.9609614587650228, + "grad_norm": 0.22112582623958588, + "learning_rate": 1.9560712805636141e-07, + "loss": 0.4677, + "step": 11594 + }, + { + "epoch": 0.9610443431413178, + "grad_norm": 0.23640844225883484, + "learning_rate": 1.9519270617488606e-07, + "loss": 0.4665, + "step": 11595 + }, + { + "epoch": 0.9611272275176129, + "grad_norm": 0.24216777086257935, + "learning_rate": 1.947782842934107e-07, + "loss": 0.4813, + "step": 11596 + }, + { + "epoch": 0.961210111893908, + "grad_norm": 0.2325787991285324, + "learning_rate": 1.9436386241193534e-07, + "loss": 0.4821, + "step": 11597 + }, + { + "epoch": 0.961292996270203, + "grad_norm": 0.23373226821422577, + "learning_rate": 1.9394944053046004e-07, + "loss": 0.5123, + "step": 11598 + }, + { + "epoch": 0.9613758806464981, + "grad_norm": 0.24051830172538757, + "learning_rate": 1.9353501864898468e-07, + "loss": 0.5303, + "step": 11599 + }, + { + "epoch": 0.9614587650227931, + "grad_norm": 0.22909189760684967, + "learning_rate": 1.9312059676750933e-07, + "loss": 0.4681, + "step": 11600 + }, + { + "epoch": 0.9615416493990883, + "grad_norm": 0.21570426225662231, + "learning_rate": 1.92706174886034e-07, + "loss": 0.4685, + "step": 11601 + }, + { + "epoch": 0.9616245337753834, + "grad_norm": 0.2307077795267105, + "learning_rate": 1.9229175300455864e-07, + "loss": 0.4648, + "step": 11602 + }, + { + "epoch": 0.9617074181516784, + "grad_norm": 0.2341223955154419, + "learning_rate": 1.9187733112308334e-07, + "loss": 0.4799, + "step": 11603 + }, + { + "epoch": 0.9617903025279735, + "grad_norm": 0.2464323192834854, + "learning_rate": 1.9146290924160798e-07, + "loss": 0.4958, + "step": 11604 + }, + { + "epoch": 0.9618731869042686, + "grad_norm": 0.2349671870470047, + "learning_rate": 1.9104848736013263e-07, + "loss": 0.4703, + "step": 11605 + }, + { + "epoch": 0.9619560712805636, + "grad_norm": 0.2386629581451416, + "learning_rate": 1.9063406547865727e-07, + "loss": 0.4903, + "step": 11606 + }, + { + "epoch": 0.9620389556568587, + "grad_norm": 0.22993889451026917, + "learning_rate": 1.9021964359718194e-07, + "loss": 0.4699, + "step": 11607 + }, + { + "epoch": 0.9621218400331537, + "grad_norm": 0.23804034292697906, + "learning_rate": 1.898052217157066e-07, + "loss": 0.5004, + "step": 11608 + }, + { + "epoch": 0.9622047244094488, + "grad_norm": 0.2303001880645752, + "learning_rate": 1.8939079983423128e-07, + "loss": 0.4767, + "step": 11609 + }, + { + "epoch": 0.9622876087857439, + "grad_norm": 0.21901145577430725, + "learning_rate": 1.8897637795275592e-07, + "loss": 0.4698, + "step": 11610 + }, + { + "epoch": 0.9623704931620389, + "grad_norm": 0.24075929820537567, + "learning_rate": 1.8856195607128057e-07, + "loss": 0.4919, + "step": 11611 + }, + { + "epoch": 0.962453377538334, + "grad_norm": 0.22686778008937836, + "learning_rate": 1.881475341898052e-07, + "loss": 0.4563, + "step": 11612 + }, + { + "epoch": 0.962536261914629, + "grad_norm": 0.23063145577907562, + "learning_rate": 1.877331123083299e-07, + "loss": 0.47, + "step": 11613 + }, + { + "epoch": 0.9626191462909242, + "grad_norm": 0.2510869801044464, + "learning_rate": 1.8731869042685455e-07, + "loss": 0.5042, + "step": 11614 + }, + { + "epoch": 0.9627020306672193, + "grad_norm": 0.24981316924095154, + "learning_rate": 1.8690426854537922e-07, + "loss": 0.4569, + "step": 11615 + }, + { + "epoch": 0.9627849150435143, + "grad_norm": 0.2531927227973938, + "learning_rate": 1.8648984666390386e-07, + "loss": 0.5345, + "step": 11616 + }, + { + "epoch": 0.9628677994198094, + "grad_norm": 0.24499429762363434, + "learning_rate": 1.8607542478242853e-07, + "loss": 0.5139, + "step": 11617 + }, + { + "epoch": 0.9629506837961045, + "grad_norm": 0.24915966391563416, + "learning_rate": 1.856610029009532e-07, + "loss": 0.5265, + "step": 11618 + }, + { + "epoch": 0.9630335681723995, + "grad_norm": 0.22571071982383728, + "learning_rate": 1.8524658101947785e-07, + "loss": 0.4552, + "step": 11619 + }, + { + "epoch": 0.9631164525486946, + "grad_norm": 0.2669590413570404, + "learning_rate": 1.848321591380025e-07, + "loss": 0.546, + "step": 11620 + }, + { + "epoch": 0.9631993369249896, + "grad_norm": 0.23659366369247437, + "learning_rate": 1.8441773725652713e-07, + "loss": 0.492, + "step": 11621 + }, + { + "epoch": 0.9632822213012847, + "grad_norm": 0.2415950447320938, + "learning_rate": 1.8400331537505183e-07, + "loss": 0.4968, + "step": 11622 + }, + { + "epoch": 0.9633651056775798, + "grad_norm": 0.22345170378684998, + "learning_rate": 1.8358889349357647e-07, + "loss": 0.4581, + "step": 11623 + }, + { + "epoch": 0.9634479900538748, + "grad_norm": 0.21745528280735016, + "learning_rate": 1.8317447161210114e-07, + "loss": 0.4711, + "step": 11624 + }, + { + "epoch": 0.9635308744301699, + "grad_norm": 0.23609809577465057, + "learning_rate": 1.827600497306258e-07, + "loss": 0.4645, + "step": 11625 + }, + { + "epoch": 0.9636137588064649, + "grad_norm": 0.24616733193397522, + "learning_rate": 1.8234562784915043e-07, + "loss": 0.5228, + "step": 11626 + }, + { + "epoch": 0.96369664318276, + "grad_norm": 0.23023216426372528, + "learning_rate": 1.8193120596767513e-07, + "loss": 0.4388, + "step": 11627 + }, + { + "epoch": 0.9637795275590552, + "grad_norm": 0.25028255581855774, + "learning_rate": 1.8151678408619977e-07, + "loss": 0.5146, + "step": 11628 + }, + { + "epoch": 0.9638624119353502, + "grad_norm": 0.23362818360328674, + "learning_rate": 1.8110236220472441e-07, + "loss": 0.4954, + "step": 11629 + }, + { + "epoch": 0.9639452963116453, + "grad_norm": 0.25039246678352356, + "learning_rate": 1.8068794032324908e-07, + "loss": 0.5322, + "step": 11630 + }, + { + "epoch": 0.9640281806879403, + "grad_norm": 0.23690639436244965, + "learning_rate": 1.8027351844177373e-07, + "loss": 0.4904, + "step": 11631 + }, + { + "epoch": 0.9641110650642354, + "grad_norm": 0.2504991292953491, + "learning_rate": 1.798590965602984e-07, + "loss": 0.5215, + "step": 11632 + }, + { + "epoch": 0.9641939494405305, + "grad_norm": 0.22675436735153198, + "learning_rate": 1.7944467467882307e-07, + "loss": 0.4971, + "step": 11633 + }, + { + "epoch": 0.9642768338168255, + "grad_norm": 0.21506141126155853, + "learning_rate": 1.790302527973477e-07, + "loss": 0.4413, + "step": 11634 + }, + { + "epoch": 0.9643597181931206, + "grad_norm": 0.23777830600738525, + "learning_rate": 1.7861583091587236e-07, + "loss": 0.5256, + "step": 11635 + }, + { + "epoch": 0.9644426025694157, + "grad_norm": 0.24924315512180328, + "learning_rate": 1.7820140903439705e-07, + "loss": 0.4862, + "step": 11636 + }, + { + "epoch": 0.9645254869457107, + "grad_norm": 0.2316451072692871, + "learning_rate": 1.777869871529217e-07, + "loss": 0.5048, + "step": 11637 + }, + { + "epoch": 0.9646083713220058, + "grad_norm": 0.2337750345468521, + "learning_rate": 1.7737256527144634e-07, + "loss": 0.5213, + "step": 11638 + }, + { + "epoch": 0.9646912556983008, + "grad_norm": 0.23372675478458405, + "learning_rate": 1.76958143389971e-07, + "loss": 0.5147, + "step": 11639 + }, + { + "epoch": 0.9647741400745959, + "grad_norm": 0.24473783373832703, + "learning_rate": 1.7654372150849565e-07, + "loss": 0.5102, + "step": 11640 + }, + { + "epoch": 0.964857024450891, + "grad_norm": 0.22098101675510406, + "learning_rate": 1.7612929962702032e-07, + "loss": 0.4793, + "step": 11641 + }, + { + "epoch": 0.964939908827186, + "grad_norm": 0.23614834249019623, + "learning_rate": 1.75714877745545e-07, + "loss": 0.4977, + "step": 11642 + }, + { + "epoch": 0.9650227932034812, + "grad_norm": 0.25280675292015076, + "learning_rate": 1.7530045586406964e-07, + "loss": 0.511, + "step": 11643 + }, + { + "epoch": 0.9651056775797762, + "grad_norm": 0.23735055327415466, + "learning_rate": 1.7488603398259428e-07, + "loss": 0.4562, + "step": 11644 + }, + { + "epoch": 0.9651885619560713, + "grad_norm": 0.2493891417980194, + "learning_rate": 1.7447161210111895e-07, + "loss": 0.4937, + "step": 11645 + }, + { + "epoch": 0.9652714463323664, + "grad_norm": 0.24593788385391235, + "learning_rate": 1.7405719021964362e-07, + "loss": 0.4597, + "step": 11646 + }, + { + "epoch": 0.9653543307086614, + "grad_norm": 0.2506912648677826, + "learning_rate": 1.7364276833816826e-07, + "loss": 0.5225, + "step": 11647 + }, + { + "epoch": 0.9654372150849565, + "grad_norm": 0.24720820784568787, + "learning_rate": 1.7322834645669293e-07, + "loss": 0.4915, + "step": 11648 + }, + { + "epoch": 0.9655200994612515, + "grad_norm": 0.2304706871509552, + "learning_rate": 1.7281392457521758e-07, + "loss": 0.479, + "step": 11649 + }, + { + "epoch": 0.9656029838375466, + "grad_norm": 0.2361813336610794, + "learning_rate": 1.7239950269374222e-07, + "loss": 0.4931, + "step": 11650 + }, + { + "epoch": 0.9656858682138417, + "grad_norm": 0.22386133670806885, + "learning_rate": 1.7198508081226692e-07, + "loss": 0.4477, + "step": 11651 + }, + { + "epoch": 0.9657687525901367, + "grad_norm": 0.23604901134967804, + "learning_rate": 1.7157065893079156e-07, + "loss": 0.4926, + "step": 11652 + }, + { + "epoch": 0.9658516369664318, + "grad_norm": 0.24856916069984436, + "learning_rate": 1.711562370493162e-07, + "loss": 0.5303, + "step": 11653 + }, + { + "epoch": 0.965934521342727, + "grad_norm": 0.259681761264801, + "learning_rate": 1.7074181516784087e-07, + "loss": 0.4978, + "step": 11654 + }, + { + "epoch": 0.966017405719022, + "grad_norm": 0.22540505230426788, + "learning_rate": 1.7032739328636554e-07, + "loss": 0.4525, + "step": 11655 + }, + { + "epoch": 0.9661002900953171, + "grad_norm": 0.24227911233901978, + "learning_rate": 1.699129714048902e-07, + "loss": 0.5228, + "step": 11656 + }, + { + "epoch": 0.9661831744716121, + "grad_norm": 0.23791764676570892, + "learning_rate": 1.6949854952341486e-07, + "loss": 0.5208, + "step": 11657 + }, + { + "epoch": 0.9662660588479072, + "grad_norm": 0.24094922840595245, + "learning_rate": 1.690841276419395e-07, + "loss": 0.4915, + "step": 11658 + }, + { + "epoch": 0.9663489432242023, + "grad_norm": 0.2308809459209442, + "learning_rate": 1.6866970576046414e-07, + "loss": 0.4811, + "step": 11659 + }, + { + "epoch": 0.9664318276004973, + "grad_norm": 0.2231217324733734, + "learning_rate": 1.6825528387898884e-07, + "loss": 0.4649, + "step": 11660 + }, + { + "epoch": 0.9665147119767924, + "grad_norm": 0.24096228182315826, + "learning_rate": 1.6784086199751348e-07, + "loss": 0.4859, + "step": 11661 + }, + { + "epoch": 0.9665975963530874, + "grad_norm": 0.24548421800136566, + "learning_rate": 1.6742644011603813e-07, + "loss": 0.5345, + "step": 11662 + }, + { + "epoch": 0.9666804807293825, + "grad_norm": 0.2601916491985321, + "learning_rate": 1.670120182345628e-07, + "loss": 0.4897, + "step": 11663 + }, + { + "epoch": 0.9667633651056776, + "grad_norm": 0.4145394265651703, + "learning_rate": 1.6659759635308744e-07, + "loss": 0.4737, + "step": 11664 + }, + { + "epoch": 0.9668462494819726, + "grad_norm": 0.23107504844665527, + "learning_rate": 1.6618317447161214e-07, + "loss": 0.4608, + "step": 11665 + }, + { + "epoch": 0.9669291338582677, + "grad_norm": 0.233512744307518, + "learning_rate": 1.6576875259013678e-07, + "loss": 0.4465, + "step": 11666 + }, + { + "epoch": 0.9670120182345627, + "grad_norm": 0.23450098931789398, + "learning_rate": 1.6535433070866143e-07, + "loss": 0.4947, + "step": 11667 + }, + { + "epoch": 0.9670949026108578, + "grad_norm": 0.22688589990139008, + "learning_rate": 1.6493990882718607e-07, + "loss": 0.4829, + "step": 11668 + }, + { + "epoch": 0.967177786987153, + "grad_norm": 0.23399899899959564, + "learning_rate": 1.6452548694571074e-07, + "loss": 0.4667, + "step": 11669 + }, + { + "epoch": 0.967260671363448, + "grad_norm": 0.2491019070148468, + "learning_rate": 1.641110650642354e-07, + "loss": 0.5109, + "step": 11670 + }, + { + "epoch": 0.9673435557397431, + "grad_norm": 0.2536167800426483, + "learning_rate": 1.6369664318276008e-07, + "loss": 0.4832, + "step": 11671 + }, + { + "epoch": 0.9674264401160382, + "grad_norm": 0.23418401181697845, + "learning_rate": 1.6328222130128472e-07, + "loss": 0.4736, + "step": 11672 + }, + { + "epoch": 0.9675093244923332, + "grad_norm": 0.23996701836585999, + "learning_rate": 1.6286779941980937e-07, + "loss": 0.4838, + "step": 11673 + }, + { + "epoch": 0.9675922088686283, + "grad_norm": 0.2355376034975052, + "learning_rate": 1.6245337753833406e-07, + "loss": 0.4993, + "step": 11674 + }, + { + "epoch": 0.9676750932449233, + "grad_norm": 0.25527045130729675, + "learning_rate": 1.620389556568587e-07, + "loss": 0.5192, + "step": 11675 + }, + { + "epoch": 0.9677579776212184, + "grad_norm": 0.25223252177238464, + "learning_rate": 1.6162453377538335e-07, + "loss": 0.499, + "step": 11676 + }, + { + "epoch": 0.9678408619975135, + "grad_norm": 0.22714149951934814, + "learning_rate": 1.61210111893908e-07, + "loss": 0.4662, + "step": 11677 + }, + { + "epoch": 0.9679237463738085, + "grad_norm": 0.227950319647789, + "learning_rate": 1.6079569001243266e-07, + "loss": 0.4879, + "step": 11678 + }, + { + "epoch": 0.9680066307501036, + "grad_norm": 0.2543065845966339, + "learning_rate": 1.6038126813095733e-07, + "loss": 0.5134, + "step": 11679 + }, + { + "epoch": 0.9680895151263986, + "grad_norm": 0.30708324909210205, + "learning_rate": 1.59966846249482e-07, + "loss": 0.4833, + "step": 11680 + }, + { + "epoch": 0.9681723995026937, + "grad_norm": 0.23332397639751434, + "learning_rate": 1.5955242436800665e-07, + "loss": 0.4794, + "step": 11681 + }, + { + "epoch": 0.9682552838789888, + "grad_norm": 0.2302892506122589, + "learning_rate": 1.591380024865313e-07, + "loss": 0.464, + "step": 11682 + }, + { + "epoch": 0.9683381682552838, + "grad_norm": 0.23344974219799042, + "learning_rate": 1.5872358060505593e-07, + "loss": 0.4932, + "step": 11683 + }, + { + "epoch": 0.968421052631579, + "grad_norm": 0.22484828531742096, + "learning_rate": 1.5830915872358063e-07, + "loss": 0.498, + "step": 11684 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 0.24427063763141632, + "learning_rate": 1.5789473684210527e-07, + "loss": 0.4902, + "step": 11685 + }, + { + "epoch": 0.9685868213841691, + "grad_norm": 0.21958975493907928, + "learning_rate": 1.5748031496062994e-07, + "loss": 0.4606, + "step": 11686 + }, + { + "epoch": 0.9686697057604642, + "grad_norm": 0.23497961461544037, + "learning_rate": 1.570658930791546e-07, + "loss": 0.4496, + "step": 11687 + }, + { + "epoch": 0.9687525901367592, + "grad_norm": 0.22901369631290436, + "learning_rate": 1.5665147119767923e-07, + "loss": 0.4435, + "step": 11688 + }, + { + "epoch": 0.9688354745130543, + "grad_norm": 0.21826590597629547, + "learning_rate": 1.562370493162039e-07, + "loss": 0.4729, + "step": 11689 + }, + { + "epoch": 0.9689183588893494, + "grad_norm": 0.2441498339176178, + "learning_rate": 1.5582262743472857e-07, + "loss": 0.4896, + "step": 11690 + }, + { + "epoch": 0.9690012432656444, + "grad_norm": 0.24209052324295044, + "learning_rate": 1.5540820555325321e-07, + "loss": 0.4958, + "step": 11691 + }, + { + "epoch": 0.9690841276419395, + "grad_norm": 0.22992412745952606, + "learning_rate": 1.5499378367177788e-07, + "loss": 0.4969, + "step": 11692 + }, + { + "epoch": 0.9691670120182345, + "grad_norm": 0.24265918135643005, + "learning_rate": 1.5457936179030253e-07, + "loss": 0.5148, + "step": 11693 + }, + { + "epoch": 0.9692498963945296, + "grad_norm": 0.23775020241737366, + "learning_rate": 1.541649399088272e-07, + "loss": 0.4936, + "step": 11694 + }, + { + "epoch": 0.9693327807708247, + "grad_norm": 0.26224419474601746, + "learning_rate": 1.5375051802735187e-07, + "loss": 0.5296, + "step": 11695 + }, + { + "epoch": 0.9694156651471197, + "grad_norm": 0.2430465966463089, + "learning_rate": 1.533360961458765e-07, + "loss": 0.4917, + "step": 11696 + }, + { + "epoch": 0.9694985495234149, + "grad_norm": 0.22634246945381165, + "learning_rate": 1.5292167426440118e-07, + "loss": 0.4727, + "step": 11697 + }, + { + "epoch": 0.9695814338997099, + "grad_norm": 0.2579650282859802, + "learning_rate": 1.5250725238292583e-07, + "loss": 0.5581, + "step": 11698 + }, + { + "epoch": 0.969664318276005, + "grad_norm": 0.24505460262298584, + "learning_rate": 1.520928305014505e-07, + "loss": 0.5024, + "step": 11699 + }, + { + "epoch": 0.9697472026523001, + "grad_norm": 0.2344638556241989, + "learning_rate": 1.5167840861997514e-07, + "loss": 0.4872, + "step": 11700 + }, + { + "epoch": 0.9698300870285951, + "grad_norm": 0.24138784408569336, + "learning_rate": 1.512639867384998e-07, + "loss": 0.4984, + "step": 11701 + }, + { + "epoch": 0.9699129714048902, + "grad_norm": 0.2383626103401184, + "learning_rate": 1.5084956485702445e-07, + "loss": 0.5188, + "step": 11702 + }, + { + "epoch": 0.9699958557811853, + "grad_norm": 0.23715630173683167, + "learning_rate": 1.5043514297554912e-07, + "loss": 0.5097, + "step": 11703 + }, + { + "epoch": 0.9700787401574803, + "grad_norm": 0.23851342499256134, + "learning_rate": 1.500207210940738e-07, + "loss": 0.5239, + "step": 11704 + }, + { + "epoch": 0.9701616245337754, + "grad_norm": 0.23797307908535004, + "learning_rate": 1.4960629921259844e-07, + "loss": 0.4987, + "step": 11705 + }, + { + "epoch": 0.9702445089100704, + "grad_norm": 0.23387955129146576, + "learning_rate": 1.491918773311231e-07, + "loss": 0.4792, + "step": 11706 + }, + { + "epoch": 0.9703273932863655, + "grad_norm": 0.24398402869701385, + "learning_rate": 1.4877745544964775e-07, + "loss": 0.5249, + "step": 11707 + }, + { + "epoch": 0.9704102776626606, + "grad_norm": 0.23988524079322815, + "learning_rate": 1.483630335681724e-07, + "loss": 0.4695, + "step": 11708 + }, + { + "epoch": 0.9704931620389556, + "grad_norm": 0.23017311096191406, + "learning_rate": 1.4794861168669706e-07, + "loss": 0.4982, + "step": 11709 + }, + { + "epoch": 0.9705760464152507, + "grad_norm": 0.24782130122184753, + "learning_rate": 1.4753418980522173e-07, + "loss": 0.4872, + "step": 11710 + }, + { + "epoch": 0.9706589307915457, + "grad_norm": 0.24664202332496643, + "learning_rate": 1.471197679237464e-07, + "loss": 0.5206, + "step": 11711 + }, + { + "epoch": 0.9707418151678409, + "grad_norm": 0.2505713701248169, + "learning_rate": 1.4670534604227105e-07, + "loss": 0.476, + "step": 11712 + }, + { + "epoch": 0.970824699544136, + "grad_norm": 0.23102807998657227, + "learning_rate": 1.462909241607957e-07, + "loss": 0.472, + "step": 11713 + }, + { + "epoch": 0.970907583920431, + "grad_norm": 0.23286588490009308, + "learning_rate": 1.4587650227932036e-07, + "loss": 0.4756, + "step": 11714 + }, + { + "epoch": 0.9709904682967261, + "grad_norm": 0.23830589652061462, + "learning_rate": 1.45462080397845e-07, + "loss": 0.489, + "step": 11715 + }, + { + "epoch": 0.9710733526730211, + "grad_norm": 0.23665091395378113, + "learning_rate": 1.4504765851636967e-07, + "loss": 0.474, + "step": 11716 + }, + { + "epoch": 0.9711562370493162, + "grad_norm": 0.24298688769340515, + "learning_rate": 1.4463323663489434e-07, + "loss": 0.5124, + "step": 11717 + }, + { + "epoch": 0.9712391214256113, + "grad_norm": 0.22405995428562164, + "learning_rate": 1.44218814753419e-07, + "loss": 0.4621, + "step": 11718 + }, + { + "epoch": 0.9713220058019063, + "grad_norm": 0.23787261545658112, + "learning_rate": 1.4380439287194366e-07, + "loss": 0.4936, + "step": 11719 + }, + { + "epoch": 0.9714048901782014, + "grad_norm": 0.26172205805778503, + "learning_rate": 1.433899709904683e-07, + "loss": 0.4997, + "step": 11720 + }, + { + "epoch": 0.9714877745544965, + "grad_norm": 0.2199118584394455, + "learning_rate": 1.4297554910899297e-07, + "loss": 0.4546, + "step": 11721 + }, + { + "epoch": 0.9715706589307915, + "grad_norm": 0.29049453139305115, + "learning_rate": 1.4256112722751761e-07, + "loss": 0.5179, + "step": 11722 + }, + { + "epoch": 0.9716535433070866, + "grad_norm": 0.24185901880264282, + "learning_rate": 1.4214670534604229e-07, + "loss": 0.4697, + "step": 11723 + }, + { + "epoch": 0.9717364276833816, + "grad_norm": 0.25461968779563904, + "learning_rate": 1.4173228346456693e-07, + "loss": 0.54, + "step": 11724 + }, + { + "epoch": 0.9718193120596768, + "grad_norm": 0.22538742423057556, + "learning_rate": 1.413178615830916e-07, + "loss": 0.4886, + "step": 11725 + }, + { + "epoch": 0.9719021964359719, + "grad_norm": 0.23491771519184113, + "learning_rate": 1.4090343970161627e-07, + "loss": 0.4944, + "step": 11726 + }, + { + "epoch": 0.9719850808122669, + "grad_norm": 0.23651108145713806, + "learning_rate": 1.404890178201409e-07, + "loss": 0.4718, + "step": 11727 + }, + { + "epoch": 0.972067965188562, + "grad_norm": 0.5481305122375488, + "learning_rate": 1.4007459593866558e-07, + "loss": 0.4414, + "step": 11728 + }, + { + "epoch": 0.972150849564857, + "grad_norm": 0.2266557216644287, + "learning_rate": 1.3966017405719023e-07, + "loss": 0.4944, + "step": 11729 + }, + { + "epoch": 0.9722337339411521, + "grad_norm": 0.22342516481876373, + "learning_rate": 1.392457521757149e-07, + "loss": 0.4458, + "step": 11730 + }, + { + "epoch": 0.9723166183174472, + "grad_norm": 0.24636584520339966, + "learning_rate": 1.3883133029423954e-07, + "loss": 0.4996, + "step": 11731 + }, + { + "epoch": 0.9723995026937422, + "grad_norm": 0.24615079164505005, + "learning_rate": 1.384169084127642e-07, + "loss": 0.4593, + "step": 11732 + }, + { + "epoch": 0.9724823870700373, + "grad_norm": 0.21660037338733673, + "learning_rate": 1.3800248653128885e-07, + "loss": 0.4402, + "step": 11733 + }, + { + "epoch": 0.9725652714463323, + "grad_norm": 0.31917455792427063, + "learning_rate": 1.3758806464981352e-07, + "loss": 0.4991, + "step": 11734 + }, + { + "epoch": 0.9726481558226274, + "grad_norm": 0.23540067672729492, + "learning_rate": 1.371736427683382e-07, + "loss": 0.4649, + "step": 11735 + }, + { + "epoch": 0.9727310401989225, + "grad_norm": 0.2534312307834625, + "learning_rate": 1.3675922088686284e-07, + "loss": 0.4885, + "step": 11736 + }, + { + "epoch": 0.9728139245752175, + "grad_norm": 0.22527733445167542, + "learning_rate": 1.363447990053875e-07, + "loss": 0.4557, + "step": 11737 + }, + { + "epoch": 0.9728968089515126, + "grad_norm": 0.2349076122045517, + "learning_rate": 1.3593037712391215e-07, + "loss": 0.5027, + "step": 11738 + }, + { + "epoch": 0.9729796933278078, + "grad_norm": 0.23080076277256012, + "learning_rate": 1.355159552424368e-07, + "loss": 0.5044, + "step": 11739 + }, + { + "epoch": 0.9730625777041028, + "grad_norm": 0.2702796757221222, + "learning_rate": 1.3510153336096146e-07, + "loss": 0.4824, + "step": 11740 + }, + { + "epoch": 0.9731454620803979, + "grad_norm": 0.23815755546092987, + "learning_rate": 1.3468711147948613e-07, + "loss": 0.4799, + "step": 11741 + }, + { + "epoch": 0.9732283464566929, + "grad_norm": 0.23453152179718018, + "learning_rate": 1.342726895980108e-07, + "loss": 0.5007, + "step": 11742 + }, + { + "epoch": 0.973311230832988, + "grad_norm": 0.2626645267009735, + "learning_rate": 1.3385826771653545e-07, + "loss": 0.4635, + "step": 11743 + }, + { + "epoch": 0.9733941152092831, + "grad_norm": 0.25828203558921814, + "learning_rate": 1.3344384583506012e-07, + "loss": 0.4919, + "step": 11744 + }, + { + "epoch": 0.9734769995855781, + "grad_norm": 0.2436951845884323, + "learning_rate": 1.3302942395358476e-07, + "loss": 0.4795, + "step": 11745 + }, + { + "epoch": 0.9735598839618732, + "grad_norm": 0.22461315989494324, + "learning_rate": 1.326150020721094e-07, + "loss": 0.5128, + "step": 11746 + }, + { + "epoch": 0.9736427683381682, + "grad_norm": 0.24688681960105896, + "learning_rate": 1.3220058019063407e-07, + "loss": 0.5088, + "step": 11747 + }, + { + "epoch": 0.9737256527144633, + "grad_norm": 0.23668339848518372, + "learning_rate": 1.3178615830915872e-07, + "loss": 0.4992, + "step": 11748 + }, + { + "epoch": 0.9738085370907584, + "grad_norm": 0.22851993143558502, + "learning_rate": 1.313717364276834e-07, + "loss": 0.4639, + "step": 11749 + }, + { + "epoch": 0.9738914214670534, + "grad_norm": 0.22496536374092102, + "learning_rate": 1.3095731454620806e-07, + "loss": 0.478, + "step": 11750 + }, + { + "epoch": 0.9739743058433485, + "grad_norm": 0.23413880169391632, + "learning_rate": 1.305428926647327e-07, + "loss": 0.4958, + "step": 11751 + }, + { + "epoch": 0.9740571902196435, + "grad_norm": 0.237555131316185, + "learning_rate": 1.3012847078325737e-07, + "loss": 0.4718, + "step": 11752 + }, + { + "epoch": 0.9741400745959387, + "grad_norm": 0.2321883887052536, + "learning_rate": 1.2971404890178202e-07, + "loss": 0.4582, + "step": 11753 + }, + { + "epoch": 0.9742229589722338, + "grad_norm": 0.25171712040901184, + "learning_rate": 1.2929962702030669e-07, + "loss": 0.4567, + "step": 11754 + }, + { + "epoch": 0.9743058433485288, + "grad_norm": 0.2547805905342102, + "learning_rate": 1.2888520513883133e-07, + "loss": 0.5053, + "step": 11755 + }, + { + "epoch": 0.9743887277248239, + "grad_norm": 0.23854964971542358, + "learning_rate": 1.28470783257356e-07, + "loss": 0.4839, + "step": 11756 + }, + { + "epoch": 0.974471612101119, + "grad_norm": 0.22962506115436554, + "learning_rate": 1.2805636137588067e-07, + "loss": 0.4783, + "step": 11757 + }, + { + "epoch": 0.974554496477414, + "grad_norm": 0.24587257206439972, + "learning_rate": 1.276419394944053e-07, + "loss": 0.4927, + "step": 11758 + }, + { + "epoch": 0.9746373808537091, + "grad_norm": 0.2377188354730606, + "learning_rate": 1.2722751761292998e-07, + "loss": 0.4894, + "step": 11759 + }, + { + "epoch": 0.9747202652300041, + "grad_norm": 0.23806744813919067, + "learning_rate": 1.2681309573145463e-07, + "loss": 0.5122, + "step": 11760 + }, + { + "epoch": 0.9748031496062992, + "grad_norm": 0.22913971543312073, + "learning_rate": 1.263986738499793e-07, + "loss": 0.4924, + "step": 11761 + }, + { + "epoch": 0.9748860339825943, + "grad_norm": 0.22276706993579865, + "learning_rate": 1.2598425196850394e-07, + "loss": 0.4653, + "step": 11762 + }, + { + "epoch": 0.9749689183588893, + "grad_norm": 0.24303790926933289, + "learning_rate": 1.255698300870286e-07, + "loss": 0.4868, + "step": 11763 + }, + { + "epoch": 0.9750518027351844, + "grad_norm": 0.2564387321472168, + "learning_rate": 1.2515540820555325e-07, + "loss": 0.5139, + "step": 11764 + }, + { + "epoch": 0.9751346871114794, + "grad_norm": 0.22167925536632538, + "learning_rate": 1.2474098632407792e-07, + "loss": 0.4597, + "step": 11765 + }, + { + "epoch": 0.9752175714877745, + "grad_norm": 0.2676071226596832, + "learning_rate": 1.243265644426026e-07, + "loss": 0.5281, + "step": 11766 + }, + { + "epoch": 0.9753004558640697, + "grad_norm": 0.2268039882183075, + "learning_rate": 1.2391214256112724e-07, + "loss": 0.4588, + "step": 11767 + }, + { + "epoch": 0.9753833402403647, + "grad_norm": 0.23486214876174927, + "learning_rate": 1.234977206796519e-07, + "loss": 0.502, + "step": 11768 + }, + { + "epoch": 0.9754662246166598, + "grad_norm": 0.23794502019882202, + "learning_rate": 1.2308329879817655e-07, + "loss": 0.4954, + "step": 11769 + }, + { + "epoch": 0.9755491089929549, + "grad_norm": 0.23502862453460693, + "learning_rate": 1.226688769167012e-07, + "loss": 0.4829, + "step": 11770 + }, + { + "epoch": 0.9756319933692499, + "grad_norm": 0.23881638050079346, + "learning_rate": 1.2225445503522586e-07, + "loss": 0.514, + "step": 11771 + }, + { + "epoch": 0.975714877745545, + "grad_norm": 0.2345491647720337, + "learning_rate": 1.2184003315375053e-07, + "loss": 0.4783, + "step": 11772 + }, + { + "epoch": 0.97579776212184, + "grad_norm": 0.2344938963651657, + "learning_rate": 1.214256112722752e-07, + "loss": 0.4988, + "step": 11773 + }, + { + "epoch": 0.9758806464981351, + "grad_norm": 0.2330857217311859, + "learning_rate": 1.2101118939079985e-07, + "loss": 0.4551, + "step": 11774 + }, + { + "epoch": 0.9759635308744302, + "grad_norm": 0.2315533310174942, + "learning_rate": 1.2059676750932452e-07, + "loss": 0.5013, + "step": 11775 + }, + { + "epoch": 0.9760464152507252, + "grad_norm": 0.2282886505126953, + "learning_rate": 1.2018234562784916e-07, + "loss": 0.4384, + "step": 11776 + }, + { + "epoch": 0.9761292996270203, + "grad_norm": 0.23390866816043854, + "learning_rate": 1.197679237463738e-07, + "loss": 0.5094, + "step": 11777 + }, + { + "epoch": 0.9762121840033153, + "grad_norm": 0.23666155338287354, + "learning_rate": 1.1935350186489847e-07, + "loss": 0.4886, + "step": 11778 + }, + { + "epoch": 0.9762950683796104, + "grad_norm": 0.2412065714597702, + "learning_rate": 1.1893907998342313e-07, + "loss": 0.5066, + "step": 11779 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 0.24465449154376984, + "learning_rate": 1.185246581019478e-07, + "loss": 0.49, + "step": 11780 + }, + { + "epoch": 0.9764608371322006, + "grad_norm": 0.2542954385280609, + "learning_rate": 1.1811023622047244e-07, + "loss": 0.5022, + "step": 11781 + }, + { + "epoch": 0.9765437215084957, + "grad_norm": 0.2582472860813141, + "learning_rate": 1.1769581433899711e-07, + "loss": 0.5335, + "step": 11782 + }, + { + "epoch": 0.9766266058847907, + "grad_norm": 0.22526097297668457, + "learning_rate": 1.1728139245752177e-07, + "loss": 0.4955, + "step": 11783 + }, + { + "epoch": 0.9767094902610858, + "grad_norm": 0.2228958010673523, + "learning_rate": 1.1686697057604642e-07, + "loss": 0.468, + "step": 11784 + }, + { + "epoch": 0.9767923746373809, + "grad_norm": 0.22459550201892853, + "learning_rate": 1.1645254869457109e-07, + "loss": 0.4971, + "step": 11785 + }, + { + "epoch": 0.9768752590136759, + "grad_norm": 0.22185394167900085, + "learning_rate": 1.1603812681309574e-07, + "loss": 0.4588, + "step": 11786 + }, + { + "epoch": 0.976958143389971, + "grad_norm": 0.23653076589107513, + "learning_rate": 1.1562370493162041e-07, + "loss": 0.4866, + "step": 11787 + }, + { + "epoch": 0.9770410277662661, + "grad_norm": 0.2410949319601059, + "learning_rate": 1.1520928305014506e-07, + "loss": 0.5037, + "step": 11788 + }, + { + "epoch": 0.9771239121425611, + "grad_norm": 0.2485639601945877, + "learning_rate": 1.1479486116866971e-07, + "loss": 0.5307, + "step": 11789 + }, + { + "epoch": 0.9772067965188562, + "grad_norm": 0.24131004512310028, + "learning_rate": 1.1438043928719437e-07, + "loss": 0.5156, + "step": 11790 + }, + { + "epoch": 0.9772896808951512, + "grad_norm": 0.23325081169605255, + "learning_rate": 1.1396601740571903e-07, + "loss": 0.4992, + "step": 11791 + }, + { + "epoch": 0.9773725652714463, + "grad_norm": 0.209193155169487, + "learning_rate": 1.135515955242437e-07, + "loss": 0.4215, + "step": 11792 + }, + { + "epoch": 0.9774554496477414, + "grad_norm": 0.2237791121006012, + "learning_rate": 1.1313717364276834e-07, + "loss": 0.4801, + "step": 11793 + }, + { + "epoch": 0.9775383340240364, + "grad_norm": 0.23202930390834808, + "learning_rate": 1.1272275176129301e-07, + "loss": 0.4861, + "step": 11794 + }, + { + "epoch": 0.9776212184003316, + "grad_norm": 0.23205792903900146, + "learning_rate": 1.1230832987981767e-07, + "loss": 0.4881, + "step": 11795 + }, + { + "epoch": 0.9777041027766266, + "grad_norm": 0.2342948466539383, + "learning_rate": 1.1189390799834231e-07, + "loss": 0.4798, + "step": 11796 + }, + { + "epoch": 0.9777869871529217, + "grad_norm": 0.22847341001033783, + "learning_rate": 1.1147948611686698e-07, + "loss": 0.4662, + "step": 11797 + }, + { + "epoch": 0.9778698715292168, + "grad_norm": 0.24070043861865997, + "learning_rate": 1.1106506423539164e-07, + "loss": 0.4891, + "step": 11798 + }, + { + "epoch": 0.9779527559055118, + "grad_norm": 0.23210304975509644, + "learning_rate": 1.1065064235391631e-07, + "loss": 0.5172, + "step": 11799 + }, + { + "epoch": 0.9780356402818069, + "grad_norm": 0.22449378669261932, + "learning_rate": 1.1023622047244095e-07, + "loss": 0.4634, + "step": 11800 + }, + { + "epoch": 0.9781185246581019, + "grad_norm": 0.24791301786899567, + "learning_rate": 1.0982179859096562e-07, + "loss": 0.5071, + "step": 11801 + }, + { + "epoch": 0.978201409034397, + "grad_norm": 0.23964373767375946, + "learning_rate": 1.0940737670949028e-07, + "loss": 0.4988, + "step": 11802 + }, + { + "epoch": 0.9782842934106921, + "grad_norm": 0.2607148587703705, + "learning_rate": 1.0899295482801492e-07, + "loss": 0.5053, + "step": 11803 + }, + { + "epoch": 0.9783671777869871, + "grad_norm": 0.23051497340202332, + "learning_rate": 1.0857853294653959e-07, + "loss": 0.4927, + "step": 11804 + }, + { + "epoch": 0.9784500621632822, + "grad_norm": 0.2400609850883484, + "learning_rate": 1.0816411106506425e-07, + "loss": 0.5027, + "step": 11805 + }, + { + "epoch": 0.9785329465395773, + "grad_norm": 0.2280253916978836, + "learning_rate": 1.077496891835889e-07, + "loss": 0.4711, + "step": 11806 + }, + { + "epoch": 0.9786158309158723, + "grad_norm": 0.24322310090065002, + "learning_rate": 1.0733526730211356e-07, + "loss": 0.5314, + "step": 11807 + }, + { + "epoch": 0.9786987152921675, + "grad_norm": 0.21233293414115906, + "learning_rate": 1.069208454206382e-07, + "loss": 0.4513, + "step": 11808 + }, + { + "epoch": 0.9787815996684625, + "grad_norm": 0.2329973727464676, + "learning_rate": 1.0650642353916287e-07, + "loss": 0.4812, + "step": 11809 + }, + { + "epoch": 0.9788644840447576, + "grad_norm": 0.21965472400188446, + "learning_rate": 1.0609200165768753e-07, + "loss": 0.4394, + "step": 11810 + }, + { + "epoch": 0.9789473684210527, + "grad_norm": 0.23069357872009277, + "learning_rate": 1.056775797762122e-07, + "loss": 0.4674, + "step": 11811 + }, + { + "epoch": 0.9790302527973477, + "grad_norm": 0.24411876499652863, + "learning_rate": 1.0526315789473685e-07, + "loss": 0.4851, + "step": 11812 + }, + { + "epoch": 0.9791131371736428, + "grad_norm": 0.2291504591703415, + "learning_rate": 1.0484873601326152e-07, + "loss": 0.487, + "step": 11813 + }, + { + "epoch": 0.9791960215499378, + "grad_norm": 0.24929113686084747, + "learning_rate": 1.0443431413178617e-07, + "loss": 0.4944, + "step": 11814 + }, + { + "epoch": 0.9792789059262329, + "grad_norm": 0.24389970302581787, + "learning_rate": 1.0401989225031082e-07, + "loss": 0.5119, + "step": 11815 + }, + { + "epoch": 0.979361790302528, + "grad_norm": 0.23593416810035706, + "learning_rate": 1.0360547036883549e-07, + "loss": 0.4863, + "step": 11816 + }, + { + "epoch": 0.979444674678823, + "grad_norm": 0.24846488237380981, + "learning_rate": 1.0319104848736014e-07, + "loss": 0.4838, + "step": 11817 + }, + { + "epoch": 0.9795275590551181, + "grad_norm": 0.22382566332817078, + "learning_rate": 1.027766266058848e-07, + "loss": 0.5078, + "step": 11818 + }, + { + "epoch": 0.9796104434314131, + "grad_norm": 0.2432345151901245, + "learning_rate": 1.0236220472440946e-07, + "loss": 0.5098, + "step": 11819 + }, + { + "epoch": 0.9796933278077082, + "grad_norm": 0.23982611298561096, + "learning_rate": 1.0194778284293413e-07, + "loss": 0.4472, + "step": 11820 + }, + { + "epoch": 0.9797762121840033, + "grad_norm": 0.24780595302581787, + "learning_rate": 1.0153336096145877e-07, + "loss": 0.508, + "step": 11821 + }, + { + "epoch": 0.9798590965602983, + "grad_norm": 0.21929827332496643, + "learning_rate": 1.0111893907998343e-07, + "loss": 0.4702, + "step": 11822 + }, + { + "epoch": 0.9799419809365935, + "grad_norm": 0.24335245788097382, + "learning_rate": 1.007045171985081e-07, + "loss": 0.5318, + "step": 11823 + }, + { + "epoch": 0.9800248653128886, + "grad_norm": 0.23285932838916779, + "learning_rate": 1.0029009531703274e-07, + "loss": 0.4579, + "step": 11824 + }, + { + "epoch": 0.9801077496891836, + "grad_norm": 0.23499077558517456, + "learning_rate": 9.987567343555741e-08, + "loss": 0.4653, + "step": 11825 + }, + { + "epoch": 0.9801906340654787, + "grad_norm": 0.22935618460178375, + "learning_rate": 9.946125155408207e-08, + "loss": 0.4814, + "step": 11826 + }, + { + "epoch": 0.9802735184417737, + "grad_norm": 0.2378963828086853, + "learning_rate": 9.904682967260671e-08, + "loss": 0.5396, + "step": 11827 + }, + { + "epoch": 0.9803564028180688, + "grad_norm": 0.25031784176826477, + "learning_rate": 9.863240779113138e-08, + "loss": 0.5322, + "step": 11828 + }, + { + "epoch": 0.9804392871943639, + "grad_norm": 0.2555597126483917, + "learning_rate": 9.821798590965604e-08, + "loss": 0.5343, + "step": 11829 + }, + { + "epoch": 0.9805221715706589, + "grad_norm": 0.23133452236652374, + "learning_rate": 9.780356402818071e-08, + "loss": 0.493, + "step": 11830 + }, + { + "epoch": 0.980605055946954, + "grad_norm": 0.2306736707687378, + "learning_rate": 9.738914214670535e-08, + "loss": 0.5015, + "step": 11831 + }, + { + "epoch": 0.980687940323249, + "grad_norm": 0.23854543268680573, + "learning_rate": 9.697472026523002e-08, + "loss": 0.4688, + "step": 11832 + }, + { + "epoch": 0.9807708246995441, + "grad_norm": 0.29027870297431946, + "learning_rate": 9.656029838375466e-08, + "loss": 0.477, + "step": 11833 + }, + { + "epoch": 0.9808537090758392, + "grad_norm": 0.24477535486221313, + "learning_rate": 9.614587650227932e-08, + "loss": 0.5511, + "step": 11834 + }, + { + "epoch": 0.9809365934521342, + "grad_norm": 0.248052179813385, + "learning_rate": 9.573145462080399e-08, + "loss": 0.5109, + "step": 11835 + }, + { + "epoch": 0.9810194778284294, + "grad_norm": 0.23871536552906036, + "learning_rate": 9.531703273932863e-08, + "loss": 0.4958, + "step": 11836 + }, + { + "epoch": 0.9811023622047244, + "grad_norm": 0.2360985428094864, + "learning_rate": 9.49026108578533e-08, + "loss": 0.4954, + "step": 11837 + }, + { + "epoch": 0.9811852465810195, + "grad_norm": 0.2580941319465637, + "learning_rate": 9.448818897637796e-08, + "loss": 0.5112, + "step": 11838 + }, + { + "epoch": 0.9812681309573146, + "grad_norm": 0.22538773715496063, + "learning_rate": 9.40737670949026e-08, + "loss": 0.4692, + "step": 11839 + }, + { + "epoch": 0.9813510153336096, + "grad_norm": 0.2319340854883194, + "learning_rate": 9.365934521342727e-08, + "loss": 0.4463, + "step": 11840 + }, + { + "epoch": 0.9814338997099047, + "grad_norm": 0.24273844063282013, + "learning_rate": 9.324492333195193e-08, + "loss": 0.5178, + "step": 11841 + }, + { + "epoch": 0.9815167840861998, + "grad_norm": 0.25397345423698425, + "learning_rate": 9.28305014504766e-08, + "loss": 0.506, + "step": 11842 + }, + { + "epoch": 0.9815996684624948, + "grad_norm": 0.252839595079422, + "learning_rate": 9.241607956900125e-08, + "loss": 0.5132, + "step": 11843 + }, + { + "epoch": 0.9816825528387899, + "grad_norm": 0.2370888590812683, + "learning_rate": 9.200165768752592e-08, + "loss": 0.4935, + "step": 11844 + }, + { + "epoch": 0.9817654372150849, + "grad_norm": 0.23182334005832672, + "learning_rate": 9.158723580605057e-08, + "loss": 0.5178, + "step": 11845 + }, + { + "epoch": 0.98184832159138, + "grad_norm": 0.22561600804328918, + "learning_rate": 9.117281392457522e-08, + "loss": 0.4691, + "step": 11846 + }, + { + "epoch": 0.9819312059676751, + "grad_norm": 0.24554184079170227, + "learning_rate": 9.075839204309989e-08, + "loss": 0.5026, + "step": 11847 + }, + { + "epoch": 0.9820140903439701, + "grad_norm": 0.24844196438789368, + "learning_rate": 9.034397016162454e-08, + "loss": 0.4864, + "step": 11848 + }, + { + "epoch": 0.9820969747202652, + "grad_norm": 0.24892954528331757, + "learning_rate": 8.99295482801492e-08, + "loss": 0.5253, + "step": 11849 + }, + { + "epoch": 0.9821798590965602, + "grad_norm": 0.233580082654953, + "learning_rate": 8.951512639867386e-08, + "loss": 0.5079, + "step": 11850 + }, + { + "epoch": 0.9822627434728554, + "grad_norm": 0.21836219727993011, + "learning_rate": 8.910070451719853e-08, + "loss": 0.4616, + "step": 11851 + }, + { + "epoch": 0.9823456278491505, + "grad_norm": 0.33230355381965637, + "learning_rate": 8.868628263572317e-08, + "loss": 0.5048, + "step": 11852 + }, + { + "epoch": 0.9824285122254455, + "grad_norm": 0.24131882190704346, + "learning_rate": 8.827186075424783e-08, + "loss": 0.5341, + "step": 11853 + }, + { + "epoch": 0.9825113966017406, + "grad_norm": 0.22750020027160645, + "learning_rate": 8.78574388727725e-08, + "loss": 0.4802, + "step": 11854 + }, + { + "epoch": 0.9825942809780357, + "grad_norm": 0.21188703179359436, + "learning_rate": 8.744301699129714e-08, + "loss": 0.4716, + "step": 11855 + }, + { + "epoch": 0.9826771653543307, + "grad_norm": 0.2633380889892578, + "learning_rate": 8.702859510982181e-08, + "loss": 0.5232, + "step": 11856 + }, + { + "epoch": 0.9827600497306258, + "grad_norm": 0.2334357649087906, + "learning_rate": 8.661417322834647e-08, + "loss": 0.5021, + "step": 11857 + }, + { + "epoch": 0.9828429341069208, + "grad_norm": 0.2343648374080658, + "learning_rate": 8.619975134687111e-08, + "loss": 0.4819, + "step": 11858 + }, + { + "epoch": 0.9829258184832159, + "grad_norm": 0.24575845897197723, + "learning_rate": 8.578532946539578e-08, + "loss": 0.4621, + "step": 11859 + }, + { + "epoch": 0.983008702859511, + "grad_norm": 0.2643239498138428, + "learning_rate": 8.537090758392044e-08, + "loss": 0.4734, + "step": 11860 + }, + { + "epoch": 0.983091587235806, + "grad_norm": 0.23404446244239807, + "learning_rate": 8.49564857024451e-08, + "loss": 0.4665, + "step": 11861 + }, + { + "epoch": 0.9831744716121011, + "grad_norm": 0.24067482352256775, + "learning_rate": 8.454206382096975e-08, + "loss": 0.5129, + "step": 11862 + }, + { + "epoch": 0.9832573559883961, + "grad_norm": 0.23089559376239777, + "learning_rate": 8.412764193949442e-08, + "loss": 0.4899, + "step": 11863 + }, + { + "epoch": 0.9833402403646913, + "grad_norm": 0.2319847196340561, + "learning_rate": 8.371322005801906e-08, + "loss": 0.4744, + "step": 11864 + }, + { + "epoch": 0.9834231247409864, + "grad_norm": 0.23708324134349823, + "learning_rate": 8.329879817654372e-08, + "loss": 0.4949, + "step": 11865 + }, + { + "epoch": 0.9835060091172814, + "grad_norm": 0.2182731032371521, + "learning_rate": 8.288437629506839e-08, + "loss": 0.4474, + "step": 11866 + }, + { + "epoch": 0.9835888934935765, + "grad_norm": 0.233004629611969, + "learning_rate": 8.246995441359303e-08, + "loss": 0.4695, + "step": 11867 + }, + { + "epoch": 0.9836717778698715, + "grad_norm": 0.245550274848938, + "learning_rate": 8.20555325321177e-08, + "loss": 0.5033, + "step": 11868 + }, + { + "epoch": 0.9837546622461666, + "grad_norm": 0.23801569640636444, + "learning_rate": 8.164111065064236e-08, + "loss": 0.5053, + "step": 11869 + }, + { + "epoch": 0.9838375466224617, + "grad_norm": 0.2312181442975998, + "learning_rate": 8.122668876916703e-08, + "loss": 0.4424, + "step": 11870 + }, + { + "epoch": 0.9839204309987567, + "grad_norm": 0.22994934022426605, + "learning_rate": 8.081226688769167e-08, + "loss": 0.486, + "step": 11871 + }, + { + "epoch": 0.9840033153750518, + "grad_norm": 0.23256976902484894, + "learning_rate": 8.039784500621633e-08, + "loss": 0.4882, + "step": 11872 + }, + { + "epoch": 0.9840861997513469, + "grad_norm": 0.25140446424484253, + "learning_rate": 7.9983423124741e-08, + "loss": 0.5068, + "step": 11873 + }, + { + "epoch": 0.9841690841276419, + "grad_norm": 0.24937766790390015, + "learning_rate": 7.956900124326565e-08, + "loss": 0.4796, + "step": 11874 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 0.23924224078655243, + "learning_rate": 7.915457936179032e-08, + "loss": 0.4934, + "step": 11875 + }, + { + "epoch": 0.984334852880232, + "grad_norm": 0.24474428594112396, + "learning_rate": 7.874015748031497e-08, + "loss": 0.4478, + "step": 11876 + }, + { + "epoch": 0.9844177372565271, + "grad_norm": 0.2355259507894516, + "learning_rate": 7.832573559883962e-08, + "loss": 0.4456, + "step": 11877 + }, + { + "epoch": 0.9845006216328223, + "grad_norm": 0.21741284430027008, + "learning_rate": 7.791131371736429e-08, + "loss": 0.4478, + "step": 11878 + }, + { + "epoch": 0.9845835060091173, + "grad_norm": 0.22226358950138092, + "learning_rate": 7.749689183588894e-08, + "loss": 0.4675, + "step": 11879 + }, + { + "epoch": 0.9846663903854124, + "grad_norm": 0.2428506314754486, + "learning_rate": 7.70824699544136e-08, + "loss": 0.4613, + "step": 11880 + }, + { + "epoch": 0.9847492747617074, + "grad_norm": 0.24963834881782532, + "learning_rate": 7.666804807293826e-08, + "loss": 0.5153, + "step": 11881 + }, + { + "epoch": 0.9848321591380025, + "grad_norm": 0.2206316590309143, + "learning_rate": 7.625362619146291e-08, + "loss": 0.4674, + "step": 11882 + }, + { + "epoch": 0.9849150435142976, + "grad_norm": 0.2320261299610138, + "learning_rate": 7.583920430998757e-08, + "loss": 0.4875, + "step": 11883 + }, + { + "epoch": 0.9849979278905926, + "grad_norm": 0.24462468922138214, + "learning_rate": 7.542478242851223e-08, + "loss": 0.5167, + "step": 11884 + }, + { + "epoch": 0.9850808122668877, + "grad_norm": 0.23995983600616455, + "learning_rate": 7.50103605470369e-08, + "loss": 0.489, + "step": 11885 + }, + { + "epoch": 0.9851636966431827, + "grad_norm": 0.2169325053691864, + "learning_rate": 7.459593866556155e-08, + "loss": 0.447, + "step": 11886 + }, + { + "epoch": 0.9852465810194778, + "grad_norm": 0.25540024042129517, + "learning_rate": 7.41815167840862e-08, + "loss": 0.515, + "step": 11887 + }, + { + "epoch": 0.9853294653957729, + "grad_norm": 0.23668022453784943, + "learning_rate": 7.376709490261087e-08, + "loss": 0.4326, + "step": 11888 + }, + { + "epoch": 0.9854123497720679, + "grad_norm": 0.2646447718143463, + "learning_rate": 7.335267302113552e-08, + "loss": 0.5222, + "step": 11889 + }, + { + "epoch": 0.985495234148363, + "grad_norm": 0.23025104403495789, + "learning_rate": 7.293825113966018e-08, + "loss": 0.4632, + "step": 11890 + }, + { + "epoch": 0.9855781185246582, + "grad_norm": 0.2159164398908615, + "learning_rate": 7.252382925818484e-08, + "loss": 0.463, + "step": 11891 + }, + { + "epoch": 0.9856610029009532, + "grad_norm": 0.2576632797718048, + "learning_rate": 7.21094073767095e-08, + "loss": 0.4828, + "step": 11892 + }, + { + "epoch": 0.9857438872772483, + "grad_norm": 0.2347642034292221, + "learning_rate": 7.169498549523415e-08, + "loss": 0.4645, + "step": 11893 + }, + { + "epoch": 0.9858267716535433, + "grad_norm": 0.24141916632652283, + "learning_rate": 7.128056361375881e-08, + "loss": 0.4868, + "step": 11894 + }, + { + "epoch": 0.9859096560298384, + "grad_norm": 0.23681232333183289, + "learning_rate": 7.086614173228346e-08, + "loss": 0.4716, + "step": 11895 + }, + { + "epoch": 0.9859925404061335, + "grad_norm": 0.23111100494861603, + "learning_rate": 7.045171985080813e-08, + "loss": 0.4584, + "step": 11896 + }, + { + "epoch": 0.9860754247824285, + "grad_norm": 0.23563224077224731, + "learning_rate": 7.003729796933279e-08, + "loss": 0.4997, + "step": 11897 + }, + { + "epoch": 0.9861583091587236, + "grad_norm": 0.23613525927066803, + "learning_rate": 6.962287608785745e-08, + "loss": 0.4865, + "step": 11898 + }, + { + "epoch": 0.9862411935350186, + "grad_norm": 0.23636846244335175, + "learning_rate": 6.92084542063821e-08, + "loss": 0.5117, + "step": 11899 + }, + { + "epoch": 0.9863240779113137, + "grad_norm": 0.23982609808444977, + "learning_rate": 6.879403232490676e-08, + "loss": 0.5059, + "step": 11900 + }, + { + "epoch": 0.9864069622876088, + "grad_norm": 0.26182761788368225, + "learning_rate": 6.837961044343142e-08, + "loss": 0.5766, + "step": 11901 + }, + { + "epoch": 0.9864898466639038, + "grad_norm": 0.23941224813461304, + "learning_rate": 6.796518856195608e-08, + "loss": 0.5065, + "step": 11902 + }, + { + "epoch": 0.9865727310401989, + "grad_norm": 0.21527402102947235, + "learning_rate": 6.755076668048073e-08, + "loss": 0.4757, + "step": 11903 + }, + { + "epoch": 0.9866556154164939, + "grad_norm": 0.24356994032859802, + "learning_rate": 6.71363447990054e-08, + "loss": 0.4763, + "step": 11904 + }, + { + "epoch": 0.986738499792789, + "grad_norm": 0.23533159494400024, + "learning_rate": 6.672192291753006e-08, + "loss": 0.4972, + "step": 11905 + }, + { + "epoch": 0.9868213841690842, + "grad_norm": 0.24300925433635712, + "learning_rate": 6.63075010360547e-08, + "loss": 0.4536, + "step": 11906 + }, + { + "epoch": 0.9869042685453792, + "grad_norm": 0.23382534086704254, + "learning_rate": 6.589307915457936e-08, + "loss": 0.5101, + "step": 11907 + }, + { + "epoch": 0.9869871529216743, + "grad_norm": 0.25081637501716614, + "learning_rate": 6.547865727310403e-08, + "loss": 0.5325, + "step": 11908 + }, + { + "epoch": 0.9870700372979694, + "grad_norm": 0.24301600456237793, + "learning_rate": 6.506423539162869e-08, + "loss": 0.4901, + "step": 11909 + }, + { + "epoch": 0.9871529216742644, + "grad_norm": 0.22034436464309692, + "learning_rate": 6.464981351015334e-08, + "loss": 0.4505, + "step": 11910 + }, + { + "epoch": 0.9872358060505595, + "grad_norm": 0.23030133545398712, + "learning_rate": 6.4235391628678e-08, + "loss": 0.488, + "step": 11911 + }, + { + "epoch": 0.9873186904268545, + "grad_norm": 0.2509356439113617, + "learning_rate": 6.382096974720266e-08, + "loss": 0.5275, + "step": 11912 + }, + { + "epoch": 0.9874015748031496, + "grad_norm": 0.2885178327560425, + "learning_rate": 6.340654786572731e-08, + "loss": 0.5761, + "step": 11913 + }, + { + "epoch": 0.9874844591794447, + "grad_norm": 0.23797667026519775, + "learning_rate": 6.299212598425197e-08, + "loss": 0.5004, + "step": 11914 + }, + { + "epoch": 0.9875673435557397, + "grad_norm": 0.24039199948310852, + "learning_rate": 6.257770410277663e-08, + "loss": 0.4896, + "step": 11915 + }, + { + "epoch": 0.9876502279320348, + "grad_norm": 0.23154106736183167, + "learning_rate": 6.21632822213013e-08, + "loss": 0.4944, + "step": 11916 + }, + { + "epoch": 0.9877331123083298, + "grad_norm": 0.2512812316417694, + "learning_rate": 6.174886033982595e-08, + "loss": 0.5152, + "step": 11917 + }, + { + "epoch": 0.9878159966846249, + "grad_norm": 0.22832129895687103, + "learning_rate": 6.13344384583506e-08, + "loss": 0.5291, + "step": 11918 + }, + { + "epoch": 0.98789888106092, + "grad_norm": 0.22945675253868103, + "learning_rate": 6.092001657687527e-08, + "loss": 0.4668, + "step": 11919 + }, + { + "epoch": 0.987981765437215, + "grad_norm": 0.2275460958480835, + "learning_rate": 6.050559469539992e-08, + "loss": 0.4424, + "step": 11920 + }, + { + "epoch": 0.9880646498135102, + "grad_norm": 0.2540155053138733, + "learning_rate": 6.009117281392458e-08, + "loss": 0.5266, + "step": 11921 + }, + { + "epoch": 0.9881475341898053, + "grad_norm": 0.22664789855480194, + "learning_rate": 5.967675093244924e-08, + "loss": 0.4733, + "step": 11922 + }, + { + "epoch": 0.9882304185661003, + "grad_norm": 0.2391948103904724, + "learning_rate": 5.92623290509739e-08, + "loss": 0.5156, + "step": 11923 + }, + { + "epoch": 0.9883133029423954, + "grad_norm": 0.23422402143478394, + "learning_rate": 5.884790716949856e-08, + "loss": 0.5122, + "step": 11924 + }, + { + "epoch": 0.9883961873186904, + "grad_norm": 0.2413766235113144, + "learning_rate": 5.843348528802321e-08, + "loss": 0.489, + "step": 11925 + }, + { + "epoch": 0.9884790716949855, + "grad_norm": 0.2245371788740158, + "learning_rate": 5.801906340654787e-08, + "loss": 0.4522, + "step": 11926 + }, + { + "epoch": 0.9885619560712806, + "grad_norm": 0.22544023394584656, + "learning_rate": 5.760464152507253e-08, + "loss": 0.4655, + "step": 11927 + }, + { + "epoch": 0.9886448404475756, + "grad_norm": 0.286283940076828, + "learning_rate": 5.7190219643597185e-08, + "loss": 0.4428, + "step": 11928 + }, + { + "epoch": 0.9887277248238707, + "grad_norm": 0.22778591513633728, + "learning_rate": 5.677579776212185e-08, + "loss": 0.4676, + "step": 11929 + }, + { + "epoch": 0.9888106092001657, + "grad_norm": 0.22700420022010803, + "learning_rate": 5.6361375880646505e-08, + "loss": 0.4496, + "step": 11930 + }, + { + "epoch": 0.9888934935764608, + "grad_norm": 0.23475316166877747, + "learning_rate": 5.5946953999171155e-08, + "loss": 0.5013, + "step": 11931 + }, + { + "epoch": 0.988976377952756, + "grad_norm": 0.23371654748916626, + "learning_rate": 5.553253211769582e-08, + "loss": 0.5105, + "step": 11932 + }, + { + "epoch": 0.989059262329051, + "grad_norm": 0.22765368223190308, + "learning_rate": 5.5118110236220475e-08, + "loss": 0.4542, + "step": 11933 + }, + { + "epoch": 0.9891421467053461, + "grad_norm": 0.23157592117786407, + "learning_rate": 5.470368835474514e-08, + "loss": 0.4552, + "step": 11934 + }, + { + "epoch": 0.9892250310816411, + "grad_norm": 0.23502469062805176, + "learning_rate": 5.4289266473269795e-08, + "loss": 0.5115, + "step": 11935 + }, + { + "epoch": 0.9893079154579362, + "grad_norm": 0.23723039031028748, + "learning_rate": 5.387484459179445e-08, + "loss": 0.5018, + "step": 11936 + }, + { + "epoch": 0.9893907998342313, + "grad_norm": 0.23719067871570587, + "learning_rate": 5.34604227103191e-08, + "loss": 0.4922, + "step": 11937 + }, + { + "epoch": 0.9894736842105263, + "grad_norm": 0.22289016842842102, + "learning_rate": 5.3046000828843766e-08, + "loss": 0.4372, + "step": 11938 + }, + { + "epoch": 0.9895565685868214, + "grad_norm": 0.26834636926651, + "learning_rate": 5.263157894736842e-08, + "loss": 0.5355, + "step": 11939 + }, + { + "epoch": 0.9896394529631165, + "grad_norm": 0.22611293196678162, + "learning_rate": 5.2217157065893086e-08, + "loss": 0.4987, + "step": 11940 + }, + { + "epoch": 0.9897223373394115, + "grad_norm": 0.23650966584682465, + "learning_rate": 5.180273518441774e-08, + "loss": 0.474, + "step": 11941 + }, + { + "epoch": 0.9898052217157066, + "grad_norm": 0.2375989854335785, + "learning_rate": 5.13883133029424e-08, + "loss": 0.5146, + "step": 11942 + }, + { + "epoch": 0.9898881060920016, + "grad_norm": 0.24325630068778992, + "learning_rate": 5.097389142146706e-08, + "loss": 0.4777, + "step": 11943 + }, + { + "epoch": 0.9899709904682967, + "grad_norm": 0.22153566777706146, + "learning_rate": 5.055946953999171e-08, + "loss": 0.4623, + "step": 11944 + }, + { + "epoch": 0.9900538748445918, + "grad_norm": 0.218144491314888, + "learning_rate": 5.014504765851637e-08, + "loss": 0.4978, + "step": 11945 + }, + { + "epoch": 0.9901367592208868, + "grad_norm": 0.23251193761825562, + "learning_rate": 4.973062577704103e-08, + "loss": 0.4618, + "step": 11946 + }, + { + "epoch": 0.990219643597182, + "grad_norm": 0.22755560278892517, + "learning_rate": 4.931620389556569e-08, + "loss": 0.4762, + "step": 11947 + }, + { + "epoch": 0.990302527973477, + "grad_norm": 0.2331409454345703, + "learning_rate": 4.8901782014090353e-08, + "loss": 0.4995, + "step": 11948 + }, + { + "epoch": 0.9903854123497721, + "grad_norm": 0.23117515444755554, + "learning_rate": 4.848736013261501e-08, + "loss": 0.4807, + "step": 11949 + }, + { + "epoch": 0.9904682967260672, + "grad_norm": 0.2386603057384491, + "learning_rate": 4.807293825113966e-08, + "loss": 0.4779, + "step": 11950 + }, + { + "epoch": 0.9905511811023622, + "grad_norm": 0.25577667355537415, + "learning_rate": 4.765851636966432e-08, + "loss": 0.4883, + "step": 11951 + }, + { + "epoch": 0.9906340654786573, + "grad_norm": 0.25080516934394836, + "learning_rate": 4.724409448818898e-08, + "loss": 0.4608, + "step": 11952 + }, + { + "epoch": 0.9907169498549523, + "grad_norm": 0.23709852993488312, + "learning_rate": 4.682967260671364e-08, + "loss": 0.5106, + "step": 11953 + }, + { + "epoch": 0.9907998342312474, + "grad_norm": 0.2171367108821869, + "learning_rate": 4.64152507252383e-08, + "loss": 0.4503, + "step": 11954 + }, + { + "epoch": 0.9908827186075425, + "grad_norm": 0.25950828194618225, + "learning_rate": 4.600082884376296e-08, + "loss": 0.5145, + "step": 11955 + }, + { + "epoch": 0.9909656029838375, + "grad_norm": 0.23632265627384186, + "learning_rate": 4.558640696228761e-08, + "loss": 0.4962, + "step": 11956 + }, + { + "epoch": 0.9910484873601326, + "grad_norm": 0.22492541372776031, + "learning_rate": 4.517198508081227e-08, + "loss": 0.4734, + "step": 11957 + }, + { + "epoch": 0.9911313717364277, + "grad_norm": 0.2316717803478241, + "learning_rate": 4.475756319933693e-08, + "loss": 0.4916, + "step": 11958 + }, + { + "epoch": 0.9912142561127227, + "grad_norm": 0.23104631900787354, + "learning_rate": 4.4343141317861585e-08, + "loss": 0.4988, + "step": 11959 + }, + { + "epoch": 0.9912971404890178, + "grad_norm": 0.23332098126411438, + "learning_rate": 4.392871943638625e-08, + "loss": 0.4756, + "step": 11960 + }, + { + "epoch": 0.9913800248653128, + "grad_norm": 0.2626270651817322, + "learning_rate": 4.3514297554910905e-08, + "loss": 0.4689, + "step": 11961 + }, + { + "epoch": 0.991462909241608, + "grad_norm": 0.247817263007164, + "learning_rate": 4.3099875673435555e-08, + "loss": 0.5228, + "step": 11962 + }, + { + "epoch": 0.9915457936179031, + "grad_norm": 0.22212739288806915, + "learning_rate": 4.268545379196022e-08, + "loss": 0.4247, + "step": 11963 + }, + { + "epoch": 0.9916286779941981, + "grad_norm": 0.23963312804698944, + "learning_rate": 4.2271031910484875e-08, + "loss": 0.4958, + "step": 11964 + }, + { + "epoch": 0.9917115623704932, + "grad_norm": 0.2311537116765976, + "learning_rate": 4.185661002900953e-08, + "loss": 0.4966, + "step": 11965 + }, + { + "epoch": 0.9917944467467882, + "grad_norm": 0.22074271738529205, + "learning_rate": 4.1442188147534195e-08, + "loss": 0.4625, + "step": 11966 + }, + { + "epoch": 0.9918773311230833, + "grad_norm": 0.21827974915504456, + "learning_rate": 4.102776626605885e-08, + "loss": 0.4301, + "step": 11967 + }, + { + "epoch": 0.9919602154993784, + "grad_norm": 0.23314930498600006, + "learning_rate": 4.0613344384583516e-08, + "loss": 0.5338, + "step": 11968 + }, + { + "epoch": 0.9920430998756734, + "grad_norm": 0.23467302322387695, + "learning_rate": 4.0198922503108166e-08, + "loss": 0.509, + "step": 11969 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 0.23083645105361938, + "learning_rate": 3.978450062163282e-08, + "loss": 0.4611, + "step": 11970 + }, + { + "epoch": 0.9922088686282635, + "grad_norm": 0.2542565166950226, + "learning_rate": 3.9370078740157486e-08, + "loss": 0.5081, + "step": 11971 + }, + { + "epoch": 0.9922917530045586, + "grad_norm": 0.24121272563934326, + "learning_rate": 3.895565685868214e-08, + "loss": 0.4873, + "step": 11972 + }, + { + "epoch": 0.9923746373808537, + "grad_norm": 0.24228845536708832, + "learning_rate": 3.85412349772068e-08, + "loss": 0.4804, + "step": 11973 + }, + { + "epoch": 0.9924575217571487, + "grad_norm": 0.3105582594871521, + "learning_rate": 3.8126813095731456e-08, + "loss": 0.5319, + "step": 11974 + }, + { + "epoch": 0.9925404061334439, + "grad_norm": 0.23462159931659698, + "learning_rate": 3.771239121425611e-08, + "loss": 0.468, + "step": 11975 + }, + { + "epoch": 0.992623290509739, + "grad_norm": 0.23684440553188324, + "learning_rate": 3.7297969332780777e-08, + "loss": 0.471, + "step": 11976 + }, + { + "epoch": 0.992706174886034, + "grad_norm": 0.26398345828056335, + "learning_rate": 3.6883547451305433e-08, + "loss": 0.4663, + "step": 11977 + }, + { + "epoch": 0.9927890592623291, + "grad_norm": 0.24819479882717133, + "learning_rate": 3.646912556983009e-08, + "loss": 0.5084, + "step": 11978 + }, + { + "epoch": 0.9928719436386241, + "grad_norm": 0.22807623445987701, + "learning_rate": 3.605470368835475e-08, + "loss": 0.4661, + "step": 11979 + }, + { + "epoch": 0.9929548280149192, + "grad_norm": 0.21865898370742798, + "learning_rate": 3.5640281806879404e-08, + "loss": 0.4957, + "step": 11980 + }, + { + "epoch": 0.9930377123912143, + "grad_norm": 0.23125533759593964, + "learning_rate": 3.522585992540407e-08, + "loss": 0.494, + "step": 11981 + }, + { + "epoch": 0.9931205967675093, + "grad_norm": 0.246164008975029, + "learning_rate": 3.4811438043928724e-08, + "loss": 0.5175, + "step": 11982 + }, + { + "epoch": 0.9932034811438044, + "grad_norm": 0.24465960264205933, + "learning_rate": 3.439701616245338e-08, + "loss": 0.5155, + "step": 11983 + }, + { + "epoch": 0.9932863655200994, + "grad_norm": 0.24551472067832947, + "learning_rate": 3.398259428097804e-08, + "loss": 0.4953, + "step": 11984 + }, + { + "epoch": 0.9933692498963945, + "grad_norm": 0.23648928105831146, + "learning_rate": 3.35681723995027e-08, + "loss": 0.4919, + "step": 11985 + }, + { + "epoch": 0.9934521342726896, + "grad_norm": 0.25939568877220154, + "learning_rate": 3.315375051802735e-08, + "loss": 0.4954, + "step": 11986 + }, + { + "epoch": 0.9935350186489846, + "grad_norm": 0.22458592057228088, + "learning_rate": 3.2739328636552014e-08, + "loss": 0.4642, + "step": 11987 + }, + { + "epoch": 0.9936179030252797, + "grad_norm": 0.24559809267520905, + "learning_rate": 3.232490675507667e-08, + "loss": 0.4897, + "step": 11988 + }, + { + "epoch": 0.9937007874015747, + "grad_norm": 0.22790616750717163, + "learning_rate": 3.191048487360133e-08, + "loss": 0.5078, + "step": 11989 + }, + { + "epoch": 0.9937836717778699, + "grad_norm": 0.24239851534366608, + "learning_rate": 3.1496062992125985e-08, + "loss": 0.4833, + "step": 11990 + }, + { + "epoch": 0.993866556154165, + "grad_norm": 0.2240189015865326, + "learning_rate": 3.108164111065065e-08, + "loss": 0.4785, + "step": 11991 + }, + { + "epoch": 0.99394944053046, + "grad_norm": 0.23581857979297638, + "learning_rate": 3.06672192291753e-08, + "loss": 0.5255, + "step": 11992 + }, + { + "epoch": 0.9940323249067551, + "grad_norm": 0.2444993406534195, + "learning_rate": 3.025279734769996e-08, + "loss": 0.5135, + "step": 11993 + }, + { + "epoch": 0.9941152092830502, + "grad_norm": 0.22983190417289734, + "learning_rate": 2.983837546622462e-08, + "loss": 0.4984, + "step": 11994 + }, + { + "epoch": 0.9941980936593452, + "grad_norm": 0.21865390241146088, + "learning_rate": 2.942395358474928e-08, + "loss": 0.4476, + "step": 11995 + }, + { + "epoch": 0.9942809780356403, + "grad_norm": 0.24026194214820862, + "learning_rate": 2.9009531703273936e-08, + "loss": 0.4889, + "step": 11996 + }, + { + "epoch": 0.9943638624119353, + "grad_norm": 0.22295883297920227, + "learning_rate": 2.8595109821798592e-08, + "loss": 0.4417, + "step": 11997 + }, + { + "epoch": 0.9944467467882304, + "grad_norm": 0.23055024445056915, + "learning_rate": 2.8180687940323252e-08, + "loss": 0.471, + "step": 11998 + }, + { + "epoch": 0.9945296311645255, + "grad_norm": 0.22983194887638092, + "learning_rate": 2.776626605884791e-08, + "loss": 0.4692, + "step": 11999 + }, + { + "epoch": 0.9946125155408205, + "grad_norm": 0.24024836719036102, + "learning_rate": 2.735184417737257e-08, + "loss": 0.5164, + "step": 12000 + }, + { + "epoch": 0.9946953999171156, + "grad_norm": 0.2442254275083542, + "learning_rate": 2.6937422295897226e-08, + "loss": 0.4774, + "step": 12001 + }, + { + "epoch": 0.9947782842934106, + "grad_norm": 0.24018704891204834, + "learning_rate": 2.6523000414421883e-08, + "loss": 0.5192, + "step": 12002 + }, + { + "epoch": 0.9948611686697058, + "grad_norm": 0.2414112240076065, + "learning_rate": 2.6108578532946543e-08, + "loss": 0.4967, + "step": 12003 + }, + { + "epoch": 0.9949440530460009, + "grad_norm": 0.2431427389383316, + "learning_rate": 2.56941566514712e-08, + "loss": 0.5216, + "step": 12004 + }, + { + "epoch": 0.9950269374222959, + "grad_norm": 0.2354496270418167, + "learning_rate": 2.5279734769995857e-08, + "loss": 0.4443, + "step": 12005 + }, + { + "epoch": 0.995109821798591, + "grad_norm": 0.24053913354873657, + "learning_rate": 2.4865312888520517e-08, + "loss": 0.4637, + "step": 12006 + }, + { + "epoch": 0.9951927061748861, + "grad_norm": 0.2304239273071289, + "learning_rate": 2.4450891007045177e-08, + "loss": 0.4811, + "step": 12007 + }, + { + "epoch": 0.9952755905511811, + "grad_norm": 0.22476941347122192, + "learning_rate": 2.403646912556983e-08, + "loss": 0.4704, + "step": 12008 + }, + { + "epoch": 0.9953584749274762, + "grad_norm": 0.24249713122844696, + "learning_rate": 2.362204724409449e-08, + "loss": 0.4768, + "step": 12009 + }, + { + "epoch": 0.9954413593037712, + "grad_norm": 0.23730440437793732, + "learning_rate": 2.320762536261915e-08, + "loss": 0.4766, + "step": 12010 + }, + { + "epoch": 0.9955242436800663, + "grad_norm": 0.22052134573459625, + "learning_rate": 2.2793203481143804e-08, + "loss": 0.4891, + "step": 12011 + }, + { + "epoch": 0.9956071280563614, + "grad_norm": 0.24449223279953003, + "learning_rate": 2.2378781599668464e-08, + "loss": 0.4931, + "step": 12012 + }, + { + "epoch": 0.9956900124326564, + "grad_norm": 0.23828630149364471, + "learning_rate": 2.1964359718193124e-08, + "loss": 0.4734, + "step": 12013 + }, + { + "epoch": 0.9957728968089515, + "grad_norm": 0.2595207691192627, + "learning_rate": 2.1549937836717778e-08, + "loss": 0.5193, + "step": 12014 + }, + { + "epoch": 0.9958557811852465, + "grad_norm": 0.2495441883802414, + "learning_rate": 2.1135515955242438e-08, + "loss": 0.4961, + "step": 12015 + }, + { + "epoch": 0.9959386655615416, + "grad_norm": 0.2310778945684433, + "learning_rate": 2.0721094073767098e-08, + "loss": 0.4583, + "step": 12016 + }, + { + "epoch": 0.9960215499378368, + "grad_norm": 0.2557829022407532, + "learning_rate": 2.0306672192291758e-08, + "loss": 0.5148, + "step": 12017 + }, + { + "epoch": 0.9961044343141318, + "grad_norm": 0.23183347284793854, + "learning_rate": 1.989225031081641e-08, + "loss": 0.4973, + "step": 12018 + }, + { + "epoch": 0.9961873186904269, + "grad_norm": 0.24386180937290192, + "learning_rate": 1.947782842934107e-08, + "loss": 0.5177, + "step": 12019 + }, + { + "epoch": 0.9962702030667219, + "grad_norm": 0.22114747762680054, + "learning_rate": 1.9063406547865728e-08, + "loss": 0.4773, + "step": 12020 + }, + { + "epoch": 0.996353087443017, + "grad_norm": 0.23613232374191284, + "learning_rate": 1.8648984666390388e-08, + "loss": 0.4798, + "step": 12021 + }, + { + "epoch": 0.9964359718193121, + "grad_norm": 0.23915687203407288, + "learning_rate": 1.8234562784915045e-08, + "loss": 0.4892, + "step": 12022 + }, + { + "epoch": 0.9965188561956071, + "grad_norm": 0.23713932931423187, + "learning_rate": 1.7820140903439702e-08, + "loss": 0.5089, + "step": 12023 + }, + { + "epoch": 0.9966017405719022, + "grad_norm": 0.2399298995733261, + "learning_rate": 1.7405719021964362e-08, + "loss": 0.4795, + "step": 12024 + }, + { + "epoch": 0.9966846249481973, + "grad_norm": 0.2414558082818985, + "learning_rate": 1.699129714048902e-08, + "loss": 0.4871, + "step": 12025 + }, + { + "epoch": 0.9967675093244923, + "grad_norm": 0.23296226561069489, + "learning_rate": 1.6576875259013676e-08, + "loss": 0.4547, + "step": 12026 + }, + { + "epoch": 0.9968503937007874, + "grad_norm": 0.25012198090553284, + "learning_rate": 1.6162453377538336e-08, + "loss": 0.4953, + "step": 12027 + }, + { + "epoch": 0.9969332780770824, + "grad_norm": 0.239393949508667, + "learning_rate": 1.5748031496062992e-08, + "loss": 0.4988, + "step": 12028 + }, + { + "epoch": 0.9970161624533775, + "grad_norm": 0.22165410220623016, + "learning_rate": 1.533360961458765e-08, + "loss": 0.4624, + "step": 12029 + }, + { + "epoch": 0.9970990468296727, + "grad_norm": 0.23059017956256866, + "learning_rate": 1.491918773311231e-08, + "loss": 0.5243, + "step": 12030 + }, + { + "epoch": 0.9971819312059677, + "grad_norm": 0.2299015074968338, + "learning_rate": 1.4504765851636968e-08, + "loss": 0.513, + "step": 12031 + }, + { + "epoch": 0.9972648155822628, + "grad_norm": 0.23083621263504028, + "learning_rate": 1.4090343970161626e-08, + "loss": 0.4867, + "step": 12032 + }, + { + "epoch": 0.9973476999585578, + "grad_norm": 0.24810348451137543, + "learning_rate": 1.3675922088686285e-08, + "loss": 0.4923, + "step": 12033 + }, + { + "epoch": 0.9974305843348529, + "grad_norm": 0.238129124045372, + "learning_rate": 1.3261500207210941e-08, + "loss": 0.4802, + "step": 12034 + }, + { + "epoch": 0.997513468711148, + "grad_norm": 0.2280401736497879, + "learning_rate": 1.28470783257356e-08, + "loss": 0.4409, + "step": 12035 + }, + { + "epoch": 0.997596353087443, + "grad_norm": 0.23373138904571533, + "learning_rate": 1.2432656444260258e-08, + "loss": 0.4719, + "step": 12036 + }, + { + "epoch": 0.9976792374637381, + "grad_norm": 0.23595111072063446, + "learning_rate": 1.2018234562784915e-08, + "loss": 0.5209, + "step": 12037 + }, + { + "epoch": 0.9977621218400331, + "grad_norm": 0.22071516513824463, + "learning_rate": 1.1603812681309575e-08, + "loss": 0.4726, + "step": 12038 + }, + { + "epoch": 0.9978450062163282, + "grad_norm": 0.21958477795124054, + "learning_rate": 1.1189390799834232e-08, + "loss": 0.4466, + "step": 12039 + }, + { + "epoch": 0.9979278905926233, + "grad_norm": 0.2587796449661255, + "learning_rate": 1.0774968918358889e-08, + "loss": 0.5064, + "step": 12040 + }, + { + "epoch": 0.9980107749689183, + "grad_norm": 0.24406568706035614, + "learning_rate": 1.0360547036883549e-08, + "loss": 0.51, + "step": 12041 + }, + { + "epoch": 0.9980936593452134, + "grad_norm": 0.24379894137382507, + "learning_rate": 9.946125155408206e-09, + "loss": 0.4818, + "step": 12042 + }, + { + "epoch": 0.9981765437215085, + "grad_norm": 0.2333594411611557, + "learning_rate": 9.531703273932864e-09, + "loss": 0.4693, + "step": 12043 + }, + { + "epoch": 0.9982594280978035, + "grad_norm": 0.2185812145471573, + "learning_rate": 9.117281392457523e-09, + "loss": 0.4679, + "step": 12044 + }, + { + "epoch": 0.9983423124740987, + "grad_norm": 0.22691938281059265, + "learning_rate": 8.702859510982181e-09, + "loss": 0.4835, + "step": 12045 + }, + { + "epoch": 0.9984251968503937, + "grad_norm": 0.23906050622463226, + "learning_rate": 8.288437629506838e-09, + "loss": 0.4722, + "step": 12046 + }, + { + "epoch": 0.9985080812266888, + "grad_norm": 0.22063198685646057, + "learning_rate": 7.874015748031496e-09, + "loss": 0.4264, + "step": 12047 + }, + { + "epoch": 0.9985909656029839, + "grad_norm": 0.2429054230451584, + "learning_rate": 7.459593866556155e-09, + "loss": 0.5222, + "step": 12048 + }, + { + "epoch": 0.9986738499792789, + "grad_norm": 0.23870894312858582, + "learning_rate": 7.045171985080813e-09, + "loss": 0.4786, + "step": 12049 + }, + { + "epoch": 0.998756734355574, + "grad_norm": 0.2468215525150299, + "learning_rate": 6.630750103605471e-09, + "loss": 0.4888, + "step": 12050 + }, + { + "epoch": 0.998839618731869, + "grad_norm": 0.23533962666988373, + "learning_rate": 6.216328222130129e-09, + "loss": 0.4972, + "step": 12051 + }, + { + "epoch": 0.9989225031081641, + "grad_norm": 0.22867600619792938, + "learning_rate": 5.801906340654788e-09, + "loss": 0.5204, + "step": 12052 + }, + { + "epoch": 0.9990053874844592, + "grad_norm": 0.2282409369945526, + "learning_rate": 5.387484459179444e-09, + "loss": 0.4767, + "step": 12053 + }, + { + "epoch": 0.9990882718607542, + "grad_norm": 0.24481633305549622, + "learning_rate": 4.973062577704103e-09, + "loss": 0.5168, + "step": 12054 + }, + { + "epoch": 0.9991711562370493, + "grad_norm": 0.22726504504680634, + "learning_rate": 4.558640696228761e-09, + "loss": 0.45, + "step": 12055 + }, + { + "epoch": 0.9992540406133443, + "grad_norm": 0.24039137363433838, + "learning_rate": 4.144218814753419e-09, + "loss": 0.4976, + "step": 12056 + }, + { + "epoch": 0.9993369249896394, + "grad_norm": 0.23129616677761078, + "learning_rate": 3.729796933278077e-09, + "loss": 0.4881, + "step": 12057 + }, + { + "epoch": 0.9994198093659346, + "grad_norm": 0.23496733605861664, + "learning_rate": 3.3153750518027354e-09, + "loss": 0.5327, + "step": 12058 + }, + { + "epoch": 0.9995026937422296, + "grad_norm": 0.23547039926052094, + "learning_rate": 2.900953170327394e-09, + "loss": 0.4695, + "step": 12059 + }, + { + "epoch": 0.9995855781185247, + "grad_norm": 0.22152018547058105, + "learning_rate": 2.4865312888520514e-09, + "loss": 0.4556, + "step": 12060 + }, + { + "epoch": 0.9996684624948198, + "grad_norm": 0.2281203418970108, + "learning_rate": 2.0721094073767094e-09, + "loss": 0.5014, + "step": 12061 + }, + { + "epoch": 0.9997513468711148, + "grad_norm": 0.23514430224895477, + "learning_rate": 1.6576875259013677e-09, + "loss": 0.4738, + "step": 12062 + }, + { + "epoch": 0.9998342312474099, + "grad_norm": 0.22735436260700226, + "learning_rate": 1.2432656444260257e-09, + "loss": 0.4778, + "step": 12063 + }, + { + "epoch": 0.9999171156237049, + "grad_norm": 0.5213358998298645, + "learning_rate": 8.288437629506838e-10, + "loss": 0.447, + "step": 12064 + }, + { + "epoch": 1.0, + "grad_norm": 0.21175578236579895, + "learning_rate": 4.144218814753419e-10, + "loss": 0.4294, + "step": 12065 + } + ], + "logging_steps": 1.0, + "max_steps": 12065, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 4021, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.2951623956147103e+20, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}