Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoint-2500/added_tokens.json +3 -0
- checkpoint-2500/chat_template.jinja +47 -0
- checkpoint-2500/config.json +110 -0
- checkpoint-2500/generation_config.json +13 -0
- checkpoint-2500/model-00001-of-00005.safetensors +3 -0
- checkpoint-2500/model-00002-of-00005.safetensors +3 -0
- checkpoint-2500/model-00003-of-00005.safetensors +3 -0
- checkpoint-2500/model-00004-of-00005.safetensors +3 -0
- checkpoint-2500/model-00005-of-00005.safetensors +3 -0
- checkpoint-2500/model.safetensors.index.json +0 -0
- checkpoint-2500/optimizer.pt +3 -0
- checkpoint-2500/rng_state.pth +3 -0
- checkpoint-2500/scheduler.pt +3 -0
- checkpoint-2500/special_tokens_map.json +33 -0
- checkpoint-2500/tokenizer.json +3 -0
- checkpoint-2500/tokenizer.model +3 -0
- checkpoint-2500/tokenizer_config.json +0 -0
- checkpoint-2500/trainer_state.json +3084 -0
- checkpoint-2500/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -37,3 +37,4 @@ checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
| 37 |
checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
checkpoint-2000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 37 |
checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
checkpoint-2000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
checkpoint-2500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
checkpoint-2500/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<image_soft_token>": 262144
|
| 3 |
+
}
|
checkpoint-2500/chat_template.jinja
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{{ bos_token }}
|
| 2 |
+
{%- if messages[0]['role'] == 'system' -%}
|
| 3 |
+
{%- if messages[0]['content'] is string -%}
|
| 4 |
+
{%- set first_user_prefix = messages[0]['content'] + '
|
| 5 |
+
|
| 6 |
+
' -%}
|
| 7 |
+
{%- else -%}
|
| 8 |
+
{%- set first_user_prefix = messages[0]['content'][0]['text'] + '
|
| 9 |
+
|
| 10 |
+
' -%}
|
| 11 |
+
{%- endif -%}
|
| 12 |
+
{%- set loop_messages = messages[1:] -%}
|
| 13 |
+
{%- else -%}
|
| 14 |
+
{%- set first_user_prefix = "" -%}
|
| 15 |
+
{%- set loop_messages = messages -%}
|
| 16 |
+
{%- endif -%}
|
| 17 |
+
{%- for message in loop_messages -%}
|
| 18 |
+
{%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
|
| 19 |
+
{{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
|
| 20 |
+
{%- endif -%}
|
| 21 |
+
{%- if (message['role'] == 'assistant') -%}
|
| 22 |
+
{%- set role = "model" -%}
|
| 23 |
+
{%- else -%}
|
| 24 |
+
{%- set role = message['role'] -%}
|
| 25 |
+
{%- endif -%}
|
| 26 |
+
{{ '<start_of_turn>' + role + '
|
| 27 |
+
' + (first_user_prefix if loop.first else "") }}
|
| 28 |
+
{%- if message['content'] is string -%}
|
| 29 |
+
{{ message['content'] | trim }}
|
| 30 |
+
{%- elif message['content'] is iterable -%}
|
| 31 |
+
{%- for item in message['content'] -%}
|
| 32 |
+
{%- if item['type'] == 'image' -%}
|
| 33 |
+
{{ '<start_of_image>' }}
|
| 34 |
+
{%- elif item['type'] == 'text' -%}
|
| 35 |
+
{{ item['text'] | trim }}
|
| 36 |
+
{%- endif -%}
|
| 37 |
+
{%- endfor -%}
|
| 38 |
+
{%- else -%}
|
| 39 |
+
{{ raise_exception("Invalid content type") }}
|
| 40 |
+
{%- endif -%}
|
| 41 |
+
{{ '<end_of_turn>
|
| 42 |
+
' }}
|
| 43 |
+
{%- endfor -%}
|
| 44 |
+
{%- if add_generation_prompt -%}
|
| 45 |
+
{{'<start_of_turn>model
|
| 46 |
+
'}}
|
| 47 |
+
{%- endif -%}
|
checkpoint-2500/config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Gemma3ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"boi_token_index": 255999,
|
| 6 |
+
"bos_token_id": 2,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eoi_token_index": 256000,
|
| 9 |
+
"eos_token_id": 1,
|
| 10 |
+
"image_token_index": 262144,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"mm_tokens_per_image": 256,
|
| 13 |
+
"model_type": "gemma3",
|
| 14 |
+
"pad_token_id": 0,
|
| 15 |
+
"text_config": {
|
| 16 |
+
"_sliding_window_pattern": 6,
|
| 17 |
+
"attention_bias": false,
|
| 18 |
+
"attention_dropout": 0.0,
|
| 19 |
+
"attn_logit_softcapping": null,
|
| 20 |
+
"dtype": "bfloat16",
|
| 21 |
+
"final_logit_softcapping": null,
|
| 22 |
+
"head_dim": 256,
|
| 23 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
| 24 |
+
"hidden_size": 3840,
|
| 25 |
+
"initializer_range": 0.02,
|
| 26 |
+
"intermediate_size": 15360,
|
| 27 |
+
"layer_types": [
|
| 28 |
+
"sliding_attention",
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"sliding_attention",
|
| 31 |
+
"sliding_attention",
|
| 32 |
+
"sliding_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"sliding_attention",
|
| 35 |
+
"sliding_attention",
|
| 36 |
+
"sliding_attention",
|
| 37 |
+
"sliding_attention",
|
| 38 |
+
"sliding_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"sliding_attention",
|
| 41 |
+
"sliding_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"sliding_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"sliding_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"sliding_attention",
|
| 51 |
+
"full_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"sliding_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"sliding_attention",
|
| 57 |
+
"full_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"sliding_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"sliding_attention",
|
| 63 |
+
"full_attention",
|
| 64 |
+
"sliding_attention",
|
| 65 |
+
"sliding_attention",
|
| 66 |
+
"sliding_attention",
|
| 67 |
+
"sliding_attention",
|
| 68 |
+
"sliding_attention",
|
| 69 |
+
"full_attention",
|
| 70 |
+
"sliding_attention",
|
| 71 |
+
"sliding_attention",
|
| 72 |
+
"sliding_attention",
|
| 73 |
+
"sliding_attention",
|
| 74 |
+
"sliding_attention",
|
| 75 |
+
"full_attention"
|
| 76 |
+
],
|
| 77 |
+
"max_position_embeddings": 131072,
|
| 78 |
+
"model_type": "gemma3_text",
|
| 79 |
+
"num_attention_heads": 16,
|
| 80 |
+
"num_hidden_layers": 48,
|
| 81 |
+
"num_key_value_heads": 8,
|
| 82 |
+
"query_pre_attn_scalar": 256,
|
| 83 |
+
"rms_norm_eps": 1e-06,
|
| 84 |
+
"rope_local_base_freq": 10000.0,
|
| 85 |
+
"rope_scaling": {
|
| 86 |
+
"factor": 8.0,
|
| 87 |
+
"rope_type": "linear"
|
| 88 |
+
},
|
| 89 |
+
"rope_theta": 1000000.0,
|
| 90 |
+
"sliding_window": 1024,
|
| 91 |
+
"use_cache": true,
|
| 92 |
+
"vocab_size": 262208
|
| 93 |
+
},
|
| 94 |
+
"transformers_version": "4.56.1",
|
| 95 |
+
"vision_config": {
|
| 96 |
+
"attention_dropout": 0.0,
|
| 97 |
+
"dtype": "bfloat16",
|
| 98 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 99 |
+
"hidden_size": 1152,
|
| 100 |
+
"image_size": 896,
|
| 101 |
+
"intermediate_size": 4304,
|
| 102 |
+
"layer_norm_eps": 1e-06,
|
| 103 |
+
"model_type": "siglip_vision_model",
|
| 104 |
+
"num_attention_heads": 16,
|
| 105 |
+
"num_channels": 3,
|
| 106 |
+
"num_hidden_layers": 27,
|
| 107 |
+
"patch_size": 14,
|
| 108 |
+
"vision_use_head": false
|
| 109 |
+
}
|
| 110 |
+
}
|
checkpoint-2500/generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 2,
|
| 3 |
+
"do_sample": true,
|
| 4 |
+
"eos_token_id": [
|
| 5 |
+
1,
|
| 6 |
+
1,
|
| 7 |
+
106
|
| 8 |
+
],
|
| 9 |
+
"pad_token_id": 0,
|
| 10 |
+
"top_k": 64,
|
| 11 |
+
"top_p": 0.95,
|
| 12 |
+
"transformers_version": "4.56.1"
|
| 13 |
+
}
|
checkpoint-2500/model-00001-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4847447e92599833e8dbaa3067cd201c3bb5c052efa91f11ba891e43234f7832
|
| 3 |
+
size 4979902192
|
checkpoint-2500/model-00002-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:891bd54eed03cba9ee1e705533a02a8217fcc29f356e4a1f53e5fd0d178883ad
|
| 3 |
+
size 4931296592
|
checkpoint-2500/model-00003-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7cee411d9d57324e50ce064a192cc5a858276d508611b12fc599e0c9767112e0
|
| 3 |
+
size 4931296656
|
checkpoint-2500/model-00004-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a8401ec76fe154b3be135809cee8e1f5e7703baae99299a6eaf1c2f03209ecd
|
| 3 |
+
size 4931296656
|
checkpoint-2500/model-00005-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cd450cc7eaa8b7136a339b81177e049a0ea8c320d4fb048d33697613deb0398
|
| 3 |
+
size 4601000928
|
checkpoint-2500/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22f91130e44a00d17a640fcbc7f4a1feef98d004e570e25a5ed5355f4c687fc4
|
| 3 |
+
size 7286558019
|
checkpoint-2500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85d3e81a34152130965a77104dd133038beafd6611ff4da82b3b4229eac123a2
|
| 3 |
+
size 14645
|
checkpoint-2500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd200143f139327b4d41bcb99ffeaa13b2acc11ded5b3dbab181a6ff2ce855d7
|
| 3 |
+
size 1465
|
checkpoint-2500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"boi_token": "<start_of_image>",
|
| 3 |
+
"bos_token": {
|
| 4 |
+
"content": "<bos>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
"eoi_token": "<end_of_image>",
|
| 11 |
+
"eos_token": {
|
| 12 |
+
"content": "<eos>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false
|
| 17 |
+
},
|
| 18 |
+
"image_token": "<image_soft_token>",
|
| 19 |
+
"pad_token": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false
|
| 25 |
+
},
|
| 26 |
+
"unk_token": {
|
| 27 |
+
"content": "<unk>",
|
| 28 |
+
"lstrip": false,
|
| 29 |
+
"normalized": false,
|
| 30 |
+
"rstrip": false,
|
| 31 |
+
"single_word": false
|
| 32 |
+
}
|
| 33 |
+
}
|
checkpoint-2500/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
|
| 3 |
+
size 33384568
|
checkpoint-2500/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
|
| 3 |
+
size 4689074
|
checkpoint-2500/tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2500/trainer_state.json
ADDED
|
@@ -0,0 +1,3084 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.7581225659698247,
|
| 14 |
+
"epoch": 0.004,
|
| 15 |
+
"grad_norm": 101.5,
|
| 16 |
+
"learning_rate": 1.8e-07,
|
| 17 |
+
"loss": 18.1141,
|
| 18 |
+
"mean_token_accuracy": 0.7393306585028767,
|
| 19 |
+
"num_tokens": 72542.0,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.7520309612154961,
|
| 24 |
+
"epoch": 0.008,
|
| 25 |
+
"grad_norm": 162.0,
|
| 26 |
+
"learning_rate": 3.8e-07,
|
| 27 |
+
"loss": 18.2441,
|
| 28 |
+
"mean_token_accuracy": 0.7392390985041857,
|
| 29 |
+
"num_tokens": 138146.0,
|
| 30 |
+
"step": 20
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.741816989146173,
|
| 34 |
+
"epoch": 0.012,
|
| 35 |
+
"grad_norm": 92.5,
|
| 36 |
+
"learning_rate": 5.800000000000001e-07,
|
| 37 |
+
"loss": 17.8539,
|
| 38 |
+
"mean_token_accuracy": 0.7405363619327545,
|
| 39 |
+
"num_tokens": 208409.0,
|
| 40 |
+
"step": 30
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.7636421315371991,
|
| 44 |
+
"epoch": 0.016,
|
| 45 |
+
"grad_norm": 88.0,
|
| 46 |
+
"learning_rate": 7.8e-07,
|
| 47 |
+
"loss": 18.142,
|
| 48 |
+
"mean_token_accuracy": 0.7400334309786558,
|
| 49 |
+
"num_tokens": 280941.0,
|
| 50 |
+
"step": 40
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"entropy": 0.7905357209965587,
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"grad_norm": 108.0,
|
| 56 |
+
"learning_rate": 9.800000000000001e-07,
|
| 57 |
+
"loss": 18.1041,
|
| 58 |
+
"mean_token_accuracy": 0.7354144401848316,
|
| 59 |
+
"num_tokens": 349731.0,
|
| 60 |
+
"step": 50
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"entropy": 0.8223027169704438,
|
| 64 |
+
"epoch": 0.024,
|
| 65 |
+
"grad_norm": 112.5,
|
| 66 |
+
"learning_rate": 1.1800000000000001e-06,
|
| 67 |
+
"loss": 18.4599,
|
| 68 |
+
"mean_token_accuracy": 0.731158659607172,
|
| 69 |
+
"num_tokens": 419127.0,
|
| 70 |
+
"step": 60
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"entropy": 0.7908559931442142,
|
| 74 |
+
"epoch": 0.028,
|
| 75 |
+
"grad_norm": 106.0,
|
| 76 |
+
"learning_rate": 1.3800000000000001e-06,
|
| 77 |
+
"loss": 17.5053,
|
| 78 |
+
"mean_token_accuracy": 0.7440328601747751,
|
| 79 |
+
"num_tokens": 489016.0,
|
| 80 |
+
"step": 70
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"entropy": 0.8229548260569572,
|
| 84 |
+
"epoch": 0.032,
|
| 85 |
+
"grad_norm": 86.0,
|
| 86 |
+
"learning_rate": 1.5800000000000001e-06,
|
| 87 |
+
"loss": 17.3714,
|
| 88 |
+
"mean_token_accuracy": 0.7438380591571331,
|
| 89 |
+
"num_tokens": 558056.0,
|
| 90 |
+
"step": 80
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"entropy": 0.84512098710984,
|
| 94 |
+
"epoch": 0.036,
|
| 95 |
+
"grad_norm": 100.5,
|
| 96 |
+
"learning_rate": 1.7800000000000001e-06,
|
| 97 |
+
"loss": 17.2456,
|
| 98 |
+
"mean_token_accuracy": 0.7478047309443354,
|
| 99 |
+
"num_tokens": 627804.0,
|
| 100 |
+
"step": 90
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"entropy": 0.8738276792690158,
|
| 104 |
+
"epoch": 0.04,
|
| 105 |
+
"grad_norm": 67.0,
|
| 106 |
+
"learning_rate": 1.98e-06,
|
| 107 |
+
"loss": 16.6896,
|
| 108 |
+
"mean_token_accuracy": 0.748936639726162,
|
| 109 |
+
"num_tokens": 697556.0,
|
| 110 |
+
"step": 100
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.04,
|
| 114 |
+
"eval_biology_entropy": 1.1379984815716744,
|
| 115 |
+
"eval_biology_loss": 1.2692387104034424,
|
| 116 |
+
"eval_biology_mean_token_accuracy": 0.6880569371581078,
|
| 117 |
+
"eval_biology_num_tokens": 697556.0,
|
| 118 |
+
"eval_biology_runtime": 40.5573,
|
| 119 |
+
"eval_biology_samples_per_second": 12.328,
|
| 120 |
+
"eval_biology_steps_per_second": 12.328,
|
| 121 |
+
"step": 100
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04,
|
| 125 |
+
"eval_math_entropy": 0.9095822716951371,
|
| 126 |
+
"eval_math_loss": 1.0899367332458496,
|
| 127 |
+
"eval_math_mean_token_accuracy": 0.7395683622956276,
|
| 128 |
+
"eval_math_num_tokens": 697556.0,
|
| 129 |
+
"eval_math_runtime": 48.0541,
|
| 130 |
+
"eval_math_samples_per_second": 10.405,
|
| 131 |
+
"eval_math_steps_per_second": 10.405,
|
| 132 |
+
"step": 100
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.9193563124164938,
|
| 136 |
+
"epoch": 0.044,
|
| 137 |
+
"grad_norm": 88.5,
|
| 138 |
+
"learning_rate": 2.1800000000000003e-06,
|
| 139 |
+
"loss": 17.3763,
|
| 140 |
+
"mean_token_accuracy": 0.7420263335108757,
|
| 141 |
+
"num_tokens": 765262.0,
|
| 142 |
+
"step": 110
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.8906641118228436,
|
| 146 |
+
"epoch": 0.048,
|
| 147 |
+
"grad_norm": 78.0,
|
| 148 |
+
"learning_rate": 2.38e-06,
|
| 149 |
+
"loss": 16.2142,
|
| 150 |
+
"mean_token_accuracy": 0.7552934620529413,
|
| 151 |
+
"num_tokens": 836753.0,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 0.9178624864667654,
|
| 156 |
+
"epoch": 0.052,
|
| 157 |
+
"grad_norm": 70.0,
|
| 158 |
+
"learning_rate": 2.5800000000000003e-06,
|
| 159 |
+
"loss": 16.3594,
|
| 160 |
+
"mean_token_accuracy": 0.7507667759433388,
|
| 161 |
+
"num_tokens": 909947.0,
|
| 162 |
+
"step": 130
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 0.9556385111063719,
|
| 166 |
+
"epoch": 0.056,
|
| 167 |
+
"grad_norm": 70.0,
|
| 168 |
+
"learning_rate": 2.7800000000000005e-06,
|
| 169 |
+
"loss": 16.6438,
|
| 170 |
+
"mean_token_accuracy": 0.748611630871892,
|
| 171 |
+
"num_tokens": 981959.0,
|
| 172 |
+
"step": 140
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"entropy": 0.9541467605158687,
|
| 176 |
+
"epoch": 0.06,
|
| 177 |
+
"grad_norm": 86.0,
|
| 178 |
+
"learning_rate": 2.9800000000000003e-06,
|
| 179 |
+
"loss": 16.5304,
|
| 180 |
+
"mean_token_accuracy": 0.7486465960741043,
|
| 181 |
+
"num_tokens": 1053624.0,
|
| 182 |
+
"step": 150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"entropy": 0.9960285406559706,
|
| 186 |
+
"epoch": 0.064,
|
| 187 |
+
"grad_norm": 78.0,
|
| 188 |
+
"learning_rate": 3.1800000000000005e-06,
|
| 189 |
+
"loss": 16.9435,
|
| 190 |
+
"mean_token_accuracy": 0.7448480188846588,
|
| 191 |
+
"num_tokens": 1126446.0,
|
| 192 |
+
"step": 160
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"entropy": 0.9677659103646874,
|
| 196 |
+
"epoch": 0.068,
|
| 197 |
+
"grad_norm": 106.5,
|
| 198 |
+
"learning_rate": 3.3800000000000007e-06,
|
| 199 |
+
"loss": 16.1013,
|
| 200 |
+
"mean_token_accuracy": 0.7561902595683933,
|
| 201 |
+
"num_tokens": 1196286.0,
|
| 202 |
+
"step": 170
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"entropy": 0.9270429328083992,
|
| 206 |
+
"epoch": 0.072,
|
| 207 |
+
"grad_norm": 62.0,
|
| 208 |
+
"learning_rate": 3.58e-06,
|
| 209 |
+
"loss": 15.3741,
|
| 210 |
+
"mean_token_accuracy": 0.763100459985435,
|
| 211 |
+
"num_tokens": 1273905.0,
|
| 212 |
+
"step": 180
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"entropy": 0.9678281666710973,
|
| 216 |
+
"epoch": 0.076,
|
| 217 |
+
"grad_norm": 51.0,
|
| 218 |
+
"learning_rate": 3.7800000000000002e-06,
|
| 219 |
+
"loss": 15.8618,
|
| 220 |
+
"mean_token_accuracy": 0.7534618094563484,
|
| 221 |
+
"num_tokens": 1346454.0,
|
| 222 |
+
"step": 190
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"entropy": 0.9695806819945574,
|
| 226 |
+
"epoch": 0.08,
|
| 227 |
+
"grad_norm": 61.25,
|
| 228 |
+
"learning_rate": 3.980000000000001e-06,
|
| 229 |
+
"loss": 15.7874,
|
| 230 |
+
"mean_token_accuracy": 0.7560878146439791,
|
| 231 |
+
"num_tokens": 1413596.0,
|
| 232 |
+
"step": 200
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.08,
|
| 236 |
+
"eval_biology_entropy": 1.214613477051258,
|
| 237 |
+
"eval_biology_loss": 1.2630934715270996,
|
| 238 |
+
"eval_biology_mean_token_accuracy": 0.6885634814500808,
|
| 239 |
+
"eval_biology_num_tokens": 1413596.0,
|
| 240 |
+
"eval_biology_runtime": 40.4294,
|
| 241 |
+
"eval_biology_samples_per_second": 12.367,
|
| 242 |
+
"eval_biology_steps_per_second": 12.367,
|
| 243 |
+
"step": 200
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08,
|
| 247 |
+
"eval_math_entropy": 0.9574190213680267,
|
| 248 |
+
"eval_math_loss": 0.9871246218681335,
|
| 249 |
+
"eval_math_mean_token_accuracy": 0.7557044946551323,
|
| 250 |
+
"eval_math_num_tokens": 1413596.0,
|
| 251 |
+
"eval_math_runtime": 48.0895,
|
| 252 |
+
"eval_math_samples_per_second": 10.397,
|
| 253 |
+
"eval_math_steps_per_second": 10.397,
|
| 254 |
+
"step": 200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 0.8888338362798095,
|
| 258 |
+
"epoch": 0.084,
|
| 259 |
+
"grad_norm": 61.25,
|
| 260 |
+
"learning_rate": 4.18e-06,
|
| 261 |
+
"loss": 14.623,
|
| 262 |
+
"mean_token_accuracy": 0.7706882227212191,
|
| 263 |
+
"num_tokens": 1485147.0,
|
| 264 |
+
"step": 210
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 0.906145141646266,
|
| 268 |
+
"epoch": 0.088,
|
| 269 |
+
"grad_norm": 53.5,
|
| 270 |
+
"learning_rate": 4.38e-06,
|
| 271 |
+
"loss": 14.8128,
|
| 272 |
+
"mean_token_accuracy": 0.7661576565355063,
|
| 273 |
+
"num_tokens": 1559450.0,
|
| 274 |
+
"step": 220
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 0.95110107883811,
|
| 278 |
+
"epoch": 0.092,
|
| 279 |
+
"grad_norm": 63.25,
|
| 280 |
+
"learning_rate": 4.58e-06,
|
| 281 |
+
"loss": 15.2457,
|
| 282 |
+
"mean_token_accuracy": 0.761219710484147,
|
| 283 |
+
"num_tokens": 1631535.0,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 0.9411580273881555,
|
| 288 |
+
"epoch": 0.096,
|
| 289 |
+
"grad_norm": 51.0,
|
| 290 |
+
"learning_rate": 4.78e-06,
|
| 291 |
+
"loss": 14.938,
|
| 292 |
+
"mean_token_accuracy": 0.7646324578672647,
|
| 293 |
+
"num_tokens": 1701193.0,
|
| 294 |
+
"step": 240
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"entropy": 0.9254088591784239,
|
| 298 |
+
"epoch": 0.1,
|
| 299 |
+
"grad_norm": 56.25,
|
| 300 |
+
"learning_rate": 4.980000000000001e-06,
|
| 301 |
+
"loss": 14.6029,
|
| 302 |
+
"mean_token_accuracy": 0.7678503680974245,
|
| 303 |
+
"num_tokens": 1773596.0,
|
| 304 |
+
"step": 250
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"entropy": 0.9214777417480946,
|
| 308 |
+
"epoch": 0.104,
|
| 309 |
+
"grad_norm": 54.0,
|
| 310 |
+
"learning_rate": 5.18e-06,
|
| 311 |
+
"loss": 15.043,
|
| 312 |
+
"mean_token_accuracy": 0.7618352189660073,
|
| 313 |
+
"num_tokens": 1840165.0,
|
| 314 |
+
"step": 260
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"entropy": 0.9363527970388532,
|
| 318 |
+
"epoch": 0.108,
|
| 319 |
+
"grad_norm": 111.5,
|
| 320 |
+
"learning_rate": 5.380000000000001e-06,
|
| 321 |
+
"loss": 15.6052,
|
| 322 |
+
"mean_token_accuracy": 0.7564919672906398,
|
| 323 |
+
"num_tokens": 1908149.0,
|
| 324 |
+
"step": 270
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"entropy": 0.9113753678277134,
|
| 328 |
+
"epoch": 0.112,
|
| 329 |
+
"grad_norm": 53.0,
|
| 330 |
+
"learning_rate": 5.580000000000001e-06,
|
| 331 |
+
"loss": 14.4645,
|
| 332 |
+
"mean_token_accuracy": 0.7736385434865951,
|
| 333 |
+
"num_tokens": 1979232.0,
|
| 334 |
+
"step": 280
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"entropy": 0.916975767724216,
|
| 338 |
+
"epoch": 0.116,
|
| 339 |
+
"grad_norm": 54.5,
|
| 340 |
+
"learning_rate": 5.78e-06,
|
| 341 |
+
"loss": 15.0318,
|
| 342 |
+
"mean_token_accuracy": 0.7656373519450426,
|
| 343 |
+
"num_tokens": 2052039.0,
|
| 344 |
+
"step": 290
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"entropy": 0.915022067911923,
|
| 348 |
+
"epoch": 0.12,
|
| 349 |
+
"grad_norm": 46.25,
|
| 350 |
+
"learning_rate": 5.98e-06,
|
| 351 |
+
"loss": 14.7236,
|
| 352 |
+
"mean_token_accuracy": 0.7691222932189703,
|
| 353 |
+
"num_tokens": 2125815.0,
|
| 354 |
+
"step": 300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.12,
|
| 358 |
+
"eval_biology_entropy": 1.2286294117569923,
|
| 359 |
+
"eval_biology_loss": 1.268429160118103,
|
| 360 |
+
"eval_biology_mean_token_accuracy": 0.6861296361088752,
|
| 361 |
+
"eval_biology_num_tokens": 2125815.0,
|
| 362 |
+
"eval_biology_runtime": 40.4544,
|
| 363 |
+
"eval_biology_samples_per_second": 12.36,
|
| 364 |
+
"eval_biology_steps_per_second": 12.36,
|
| 365 |
+
"step": 300
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.12,
|
| 369 |
+
"eval_math_entropy": 0.9302992832064628,
|
| 370 |
+
"eval_math_loss": 0.9273415803909302,
|
| 371 |
+
"eval_math_mean_token_accuracy": 0.7664831226468086,
|
| 372 |
+
"eval_math_num_tokens": 2125815.0,
|
| 373 |
+
"eval_math_runtime": 48.047,
|
| 374 |
+
"eval_math_samples_per_second": 10.406,
|
| 375 |
+
"eval_math_steps_per_second": 10.406,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 0.9336849415674806,
|
| 380 |
+
"epoch": 0.124,
|
| 381 |
+
"grad_norm": 55.25,
|
| 382 |
+
"learning_rate": 6.18e-06,
|
| 383 |
+
"loss": 14.9674,
|
| 384 |
+
"mean_token_accuracy": 0.7661335166543722,
|
| 385 |
+
"num_tokens": 2201991.0,
|
| 386 |
+
"step": 310
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 0.8756070390343667,
|
| 390 |
+
"epoch": 0.128,
|
| 391 |
+
"grad_norm": 52.25,
|
| 392 |
+
"learning_rate": 6.380000000000001e-06,
|
| 393 |
+
"loss": 14.0,
|
| 394 |
+
"mean_token_accuracy": 0.7786776568740607,
|
| 395 |
+
"num_tokens": 2274177.0,
|
| 396 |
+
"step": 320
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 0.892756593041122,
|
| 400 |
+
"epoch": 0.132,
|
| 401 |
+
"grad_norm": 45.75,
|
| 402 |
+
"learning_rate": 6.5800000000000005e-06,
|
| 403 |
+
"loss": 14.1896,
|
| 404 |
+
"mean_token_accuracy": 0.7731969062238931,
|
| 405 |
+
"num_tokens": 2342173.0,
|
| 406 |
+
"step": 330
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 0.8899674592539668,
|
| 410 |
+
"epoch": 0.136,
|
| 411 |
+
"grad_norm": 56.5,
|
| 412 |
+
"learning_rate": 6.780000000000001e-06,
|
| 413 |
+
"loss": 14.2167,
|
| 414 |
+
"mean_token_accuracy": 0.7683363553136587,
|
| 415 |
+
"num_tokens": 2412435.0,
|
| 416 |
+
"step": 340
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"entropy": 0.9237651886418462,
|
| 420 |
+
"epoch": 0.14,
|
| 421 |
+
"grad_norm": 49.5,
|
| 422 |
+
"learning_rate": 6.98e-06,
|
| 423 |
+
"loss": 14.7203,
|
| 424 |
+
"mean_token_accuracy": 0.7684664275497198,
|
| 425 |
+
"num_tokens": 2481105.0,
|
| 426 |
+
"step": 350
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"entropy": 0.8494639024138451,
|
| 430 |
+
"epoch": 0.144,
|
| 431 |
+
"grad_norm": 49.0,
|
| 432 |
+
"learning_rate": 7.180000000000001e-06,
|
| 433 |
+
"loss": 13.6956,
|
| 434 |
+
"mean_token_accuracy": 0.7817885402590037,
|
| 435 |
+
"num_tokens": 2548445.0,
|
| 436 |
+
"step": 360
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"entropy": 0.8341110894456506,
|
| 440 |
+
"epoch": 0.148,
|
| 441 |
+
"grad_norm": 44.0,
|
| 442 |
+
"learning_rate": 7.3800000000000005e-06,
|
| 443 |
+
"loss": 13.2223,
|
| 444 |
+
"mean_token_accuracy": 0.789648124948144,
|
| 445 |
+
"num_tokens": 2621364.0,
|
| 446 |
+
"step": 370
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"entropy": 0.8608193326741457,
|
| 450 |
+
"epoch": 0.152,
|
| 451 |
+
"grad_norm": 52.25,
|
| 452 |
+
"learning_rate": 7.58e-06,
|
| 453 |
+
"loss": 13.8861,
|
| 454 |
+
"mean_token_accuracy": 0.7754677496850491,
|
| 455 |
+
"num_tokens": 2686462.0,
|
| 456 |
+
"step": 380
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"entropy": 0.8447565982118249,
|
| 460 |
+
"epoch": 0.156,
|
| 461 |
+
"grad_norm": 50.0,
|
| 462 |
+
"learning_rate": 7.78e-06,
|
| 463 |
+
"loss": 13.5199,
|
| 464 |
+
"mean_token_accuracy": 0.7812942665070295,
|
| 465 |
+
"num_tokens": 2754241.0,
|
| 466 |
+
"step": 390
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"entropy": 0.8711243279278278,
|
| 470 |
+
"epoch": 0.16,
|
| 471 |
+
"grad_norm": 79.0,
|
| 472 |
+
"learning_rate": 7.980000000000002e-06,
|
| 473 |
+
"loss": 14.1276,
|
| 474 |
+
"mean_token_accuracy": 0.7742310840636492,
|
| 475 |
+
"num_tokens": 2824418.0,
|
| 476 |
+
"step": 400
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.16,
|
| 480 |
+
"eval_biology_entropy": 1.22758097666502,
|
| 481 |
+
"eval_biology_loss": 1.2762199640274048,
|
| 482 |
+
"eval_biology_mean_token_accuracy": 0.6836535354852676,
|
| 483 |
+
"eval_biology_num_tokens": 2824418.0,
|
| 484 |
+
"eval_biology_runtime": 40.7281,
|
| 485 |
+
"eval_biology_samples_per_second": 12.277,
|
| 486 |
+
"eval_biology_steps_per_second": 12.277,
|
| 487 |
+
"step": 400
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.16,
|
| 491 |
+
"eval_math_entropy": 0.9047691844701767,
|
| 492 |
+
"eval_math_loss": 0.8841754198074341,
|
| 493 |
+
"eval_math_mean_token_accuracy": 0.775034920334816,
|
| 494 |
+
"eval_math_num_tokens": 2824418.0,
|
| 495 |
+
"eval_math_runtime": 47.9197,
|
| 496 |
+
"eval_math_samples_per_second": 10.434,
|
| 497 |
+
"eval_math_steps_per_second": 10.434,
|
| 498 |
+
"step": 400
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"entropy": 0.8859184484928846,
|
| 502 |
+
"epoch": 0.164,
|
| 503 |
+
"grad_norm": 127.0,
|
| 504 |
+
"learning_rate": 8.18e-06,
|
| 505 |
+
"loss": 14.142,
|
| 506 |
+
"mean_token_accuracy": 0.7765266574919224,
|
| 507 |
+
"num_tokens": 2895667.0,
|
| 508 |
+
"step": 410
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"entropy": 0.8501209424808621,
|
| 512 |
+
"epoch": 0.168,
|
| 513 |
+
"grad_norm": 40.0,
|
| 514 |
+
"learning_rate": 8.380000000000001e-06,
|
| 515 |
+
"loss": 13.8435,
|
| 516 |
+
"mean_token_accuracy": 0.7774271417409182,
|
| 517 |
+
"num_tokens": 2966031.0,
|
| 518 |
+
"step": 420
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"entropy": 0.8510567891411484,
|
| 522 |
+
"epoch": 0.172,
|
| 523 |
+
"grad_norm": 45.75,
|
| 524 |
+
"learning_rate": 8.580000000000001e-06,
|
| 525 |
+
"loss": 13.3673,
|
| 526 |
+
"mean_token_accuracy": 0.7858892437070608,
|
| 527 |
+
"num_tokens": 3039780.0,
|
| 528 |
+
"step": 430
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"entropy": 0.8355857279151678,
|
| 532 |
+
"epoch": 0.176,
|
| 533 |
+
"grad_norm": 44.5,
|
| 534 |
+
"learning_rate": 8.78e-06,
|
| 535 |
+
"loss": 13.7383,
|
| 536 |
+
"mean_token_accuracy": 0.7784004233777523,
|
| 537 |
+
"num_tokens": 3110313.0,
|
| 538 |
+
"step": 440
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"entropy": 0.8662491450086236,
|
| 542 |
+
"epoch": 0.18,
|
| 543 |
+
"grad_norm": 43.75,
|
| 544 |
+
"learning_rate": 8.98e-06,
|
| 545 |
+
"loss": 13.6485,
|
| 546 |
+
"mean_token_accuracy": 0.7806661173701286,
|
| 547 |
+
"num_tokens": 3179963.0,
|
| 548 |
+
"step": 450
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"entropy": 0.8596813324838877,
|
| 552 |
+
"epoch": 0.184,
|
| 553 |
+
"grad_norm": 46.0,
|
| 554 |
+
"learning_rate": 9.180000000000002e-06,
|
| 555 |
+
"loss": 13.7717,
|
| 556 |
+
"mean_token_accuracy": 0.7796813258901238,
|
| 557 |
+
"num_tokens": 3251139.0,
|
| 558 |
+
"step": 460
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"entropy": 0.8454847941175103,
|
| 562 |
+
"epoch": 0.188,
|
| 563 |
+
"grad_norm": 56.25,
|
| 564 |
+
"learning_rate": 9.38e-06,
|
| 565 |
+
"loss": 13.4682,
|
| 566 |
+
"mean_token_accuracy": 0.7803339377045632,
|
| 567 |
+
"num_tokens": 3323519.0,
|
| 568 |
+
"step": 470
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"entropy": 0.8228314617648721,
|
| 572 |
+
"epoch": 0.192,
|
| 573 |
+
"grad_norm": 49.75,
|
| 574 |
+
"learning_rate": 9.58e-06,
|
| 575 |
+
"loss": 13.3842,
|
| 576 |
+
"mean_token_accuracy": 0.7827629748731851,
|
| 577 |
+
"num_tokens": 3394170.0,
|
| 578 |
+
"step": 480
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"entropy": 0.8164506537839771,
|
| 582 |
+
"epoch": 0.196,
|
| 583 |
+
"grad_norm": 54.25,
|
| 584 |
+
"learning_rate": 9.780000000000001e-06,
|
| 585 |
+
"loss": 13.1965,
|
| 586 |
+
"mean_token_accuracy": 0.7878578577190638,
|
| 587 |
+
"num_tokens": 3465341.0,
|
| 588 |
+
"step": 490
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"entropy": 0.8838530456647277,
|
| 592 |
+
"epoch": 0.2,
|
| 593 |
+
"grad_norm": 51.0,
|
| 594 |
+
"learning_rate": 9.980000000000001e-06,
|
| 595 |
+
"loss": 13.9771,
|
| 596 |
+
"mean_token_accuracy": 0.7734297584742308,
|
| 597 |
+
"num_tokens": 3533571.0,
|
| 598 |
+
"step": 500
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.2,
|
| 602 |
+
"eval_biology_entropy": 1.2344867270588875,
|
| 603 |
+
"eval_biology_loss": 1.2817195653915405,
|
| 604 |
+
"eval_biology_mean_token_accuracy": 0.6832386125922203,
|
| 605 |
+
"eval_biology_num_tokens": 3533571.0,
|
| 606 |
+
"eval_biology_runtime": 40.4092,
|
| 607 |
+
"eval_biology_samples_per_second": 12.373,
|
| 608 |
+
"eval_biology_steps_per_second": 12.373,
|
| 609 |
+
"step": 500
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.2,
|
| 613 |
+
"eval_math_entropy": 0.8546719763875008,
|
| 614 |
+
"eval_math_loss": 0.8547734618186951,
|
| 615 |
+
"eval_math_mean_token_accuracy": 0.7804626071453095,
|
| 616 |
+
"eval_math_num_tokens": 3533571.0,
|
| 617 |
+
"eval_math_runtime": 48.0961,
|
| 618 |
+
"eval_math_samples_per_second": 10.396,
|
| 619 |
+
"eval_math_steps_per_second": 10.396,
|
| 620 |
+
"step": 500
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"entropy": 0.8578155474737287,
|
| 624 |
+
"epoch": 0.204,
|
| 625 |
+
"grad_norm": 45.75,
|
| 626 |
+
"learning_rate": 1.018e-05,
|
| 627 |
+
"loss": 13.8097,
|
| 628 |
+
"mean_token_accuracy": 0.776990968361497,
|
| 629 |
+
"num_tokens": 3602983.0,
|
| 630 |
+
"step": 510
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"entropy": 0.8306718476116657,
|
| 634 |
+
"epoch": 0.208,
|
| 635 |
+
"grad_norm": 50.25,
|
| 636 |
+
"learning_rate": 1.038e-05,
|
| 637 |
+
"loss": 13.4186,
|
| 638 |
+
"mean_token_accuracy": 0.7863588180392981,
|
| 639 |
+
"num_tokens": 3673600.0,
|
| 640 |
+
"step": 520
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"entropy": 0.8772016573697329,
|
| 644 |
+
"epoch": 0.212,
|
| 645 |
+
"grad_norm": 45.75,
|
| 646 |
+
"learning_rate": 1.0580000000000002e-05,
|
| 647 |
+
"loss": 13.9353,
|
| 648 |
+
"mean_token_accuracy": 0.7783012766391039,
|
| 649 |
+
"num_tokens": 3743641.0,
|
| 650 |
+
"step": 530
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"entropy": 0.8009312467649579,
|
| 654 |
+
"epoch": 0.216,
|
| 655 |
+
"grad_norm": 55.25,
|
| 656 |
+
"learning_rate": 1.0780000000000002e-05,
|
| 657 |
+
"loss": 12.9356,
|
| 658 |
+
"mean_token_accuracy": 0.7893661856651306,
|
| 659 |
+
"num_tokens": 3817397.0,
|
| 660 |
+
"step": 540
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"entropy": 0.8427618410438299,
|
| 664 |
+
"epoch": 0.22,
|
| 665 |
+
"grad_norm": 51.0,
|
| 666 |
+
"learning_rate": 1.0980000000000002e-05,
|
| 667 |
+
"loss": 13.3831,
|
| 668 |
+
"mean_token_accuracy": 0.7832117900252342,
|
| 669 |
+
"num_tokens": 3886866.0,
|
| 670 |
+
"step": 550
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"entropy": 0.8758323887363076,
|
| 674 |
+
"epoch": 0.224,
|
| 675 |
+
"grad_norm": 41.5,
|
| 676 |
+
"learning_rate": 1.1180000000000001e-05,
|
| 677 |
+
"loss": 13.8448,
|
| 678 |
+
"mean_token_accuracy": 0.7745687130838632,
|
| 679 |
+
"num_tokens": 3953732.0,
|
| 680 |
+
"step": 560
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"entropy": 0.8352954359725118,
|
| 684 |
+
"epoch": 0.228,
|
| 685 |
+
"grad_norm": 41.75,
|
| 686 |
+
"learning_rate": 1.138e-05,
|
| 687 |
+
"loss": 13.4861,
|
| 688 |
+
"mean_token_accuracy": 0.7828061260282994,
|
| 689 |
+
"num_tokens": 4025795.0,
|
| 690 |
+
"step": 570
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"entropy": 0.7947553901001811,
|
| 694 |
+
"epoch": 0.232,
|
| 695 |
+
"grad_norm": 43.25,
|
| 696 |
+
"learning_rate": 1.1580000000000001e-05,
|
| 697 |
+
"loss": 12.8366,
|
| 698 |
+
"mean_token_accuracy": 0.7880385994911194,
|
| 699 |
+
"num_tokens": 4098989.0,
|
| 700 |
+
"step": 580
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"entropy": 0.8285046070814133,
|
| 704 |
+
"epoch": 0.236,
|
| 705 |
+
"grad_norm": 39.25,
|
| 706 |
+
"learning_rate": 1.178e-05,
|
| 707 |
+
"loss": 13.0789,
|
| 708 |
+
"mean_token_accuracy": 0.7872951343655586,
|
| 709 |
+
"num_tokens": 4170878.0,
|
| 710 |
+
"step": 590
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"entropy": 0.7863880429416895,
|
| 714 |
+
"epoch": 0.24,
|
| 715 |
+
"grad_norm": 45.0,
|
| 716 |
+
"learning_rate": 1.198e-05,
|
| 717 |
+
"loss": 12.7479,
|
| 718 |
+
"mean_token_accuracy": 0.7928395148366689,
|
| 719 |
+
"num_tokens": 4243655.0,
|
| 720 |
+
"step": 600
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.24,
|
| 724 |
+
"eval_biology_entropy": 1.2514844969511032,
|
| 725 |
+
"eval_biology_loss": 1.2855682373046875,
|
| 726 |
+
"eval_biology_mean_token_accuracy": 0.6817903207540512,
|
| 727 |
+
"eval_biology_num_tokens": 4243655.0,
|
| 728 |
+
"eval_biology_runtime": 40.7639,
|
| 729 |
+
"eval_biology_samples_per_second": 12.266,
|
| 730 |
+
"eval_biology_steps_per_second": 12.266,
|
| 731 |
+
"step": 600
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.24,
|
| 735 |
+
"eval_math_entropy": 0.843876126229763,
|
| 736 |
+
"eval_math_loss": 0.8332411646842957,
|
| 737 |
+
"eval_math_mean_token_accuracy": 0.783966186761856,
|
| 738 |
+
"eval_math_num_tokens": 4243655.0,
|
| 739 |
+
"eval_math_runtime": 48.1965,
|
| 740 |
+
"eval_math_samples_per_second": 10.374,
|
| 741 |
+
"eval_math_steps_per_second": 10.374,
|
| 742 |
+
"step": 600
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"entropy": 0.7838549023494125,
|
| 746 |
+
"epoch": 0.244,
|
| 747 |
+
"grad_norm": 41.75,
|
| 748 |
+
"learning_rate": 1.218e-05,
|
| 749 |
+
"loss": 12.4042,
|
| 750 |
+
"mean_token_accuracy": 0.7971927735954523,
|
| 751 |
+
"num_tokens": 4317321.0,
|
| 752 |
+
"step": 610
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"entropy": 0.7797296568751335,
|
| 756 |
+
"epoch": 0.248,
|
| 757 |
+
"grad_norm": 51.75,
|
| 758 |
+
"learning_rate": 1.2380000000000002e-05,
|
| 759 |
+
"loss": 12.5895,
|
| 760 |
+
"mean_token_accuracy": 0.7916548330336809,
|
| 761 |
+
"num_tokens": 4387651.0,
|
| 762 |
+
"step": 620
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"entropy": 0.8094759196043014,
|
| 766 |
+
"epoch": 0.252,
|
| 767 |
+
"grad_norm": 84.5,
|
| 768 |
+
"learning_rate": 1.2580000000000002e-05,
|
| 769 |
+
"loss": 12.9446,
|
| 770 |
+
"mean_token_accuracy": 0.7892590902745724,
|
| 771 |
+
"num_tokens": 4461909.0,
|
| 772 |
+
"step": 630
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"entropy": 0.7797345662489533,
|
| 776 |
+
"epoch": 0.256,
|
| 777 |
+
"grad_norm": 62.0,
|
| 778 |
+
"learning_rate": 1.2780000000000001e-05,
|
| 779 |
+
"loss": 12.462,
|
| 780 |
+
"mean_token_accuracy": 0.7935512393712998,
|
| 781 |
+
"num_tokens": 4531471.0,
|
| 782 |
+
"step": 640
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"entropy": 0.7911159221082926,
|
| 786 |
+
"epoch": 0.26,
|
| 787 |
+
"grad_norm": 40.0,
|
| 788 |
+
"learning_rate": 1.2980000000000001e-05,
|
| 789 |
+
"loss": 12.8515,
|
| 790 |
+
"mean_token_accuracy": 0.7928617469966411,
|
| 791 |
+
"num_tokens": 4600757.0,
|
| 792 |
+
"step": 650
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"entropy": 0.8169831104576588,
|
| 796 |
+
"epoch": 0.264,
|
| 797 |
+
"grad_norm": 49.5,
|
| 798 |
+
"learning_rate": 1.3180000000000001e-05,
|
| 799 |
+
"loss": 13.117,
|
| 800 |
+
"mean_token_accuracy": 0.7893466159701348,
|
| 801 |
+
"num_tokens": 4669389.0,
|
| 802 |
+
"step": 660
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"entropy": 0.8287884982302784,
|
| 806 |
+
"epoch": 0.268,
|
| 807 |
+
"grad_norm": 55.25,
|
| 808 |
+
"learning_rate": 1.3380000000000002e-05,
|
| 809 |
+
"loss": 13.3618,
|
| 810 |
+
"mean_token_accuracy": 0.7853043630719185,
|
| 811 |
+
"num_tokens": 4740492.0,
|
| 812 |
+
"step": 670
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"entropy": 0.7938818972557783,
|
| 816 |
+
"epoch": 0.272,
|
| 817 |
+
"grad_norm": 42.5,
|
| 818 |
+
"learning_rate": 1.3580000000000002e-05,
|
| 819 |
+
"loss": 12.6092,
|
| 820 |
+
"mean_token_accuracy": 0.7941738195717335,
|
| 821 |
+
"num_tokens": 4810284.0,
|
| 822 |
+
"step": 680
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"entropy": 0.7964617364108563,
|
| 826 |
+
"epoch": 0.276,
|
| 827 |
+
"grad_norm": 37.25,
|
| 828 |
+
"learning_rate": 1.378e-05,
|
| 829 |
+
"loss": 12.6495,
|
| 830 |
+
"mean_token_accuracy": 0.7921258114278317,
|
| 831 |
+
"num_tokens": 4885701.0,
|
| 832 |
+
"step": 690
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"entropy": 0.7932526333257556,
|
| 836 |
+
"epoch": 0.28,
|
| 837 |
+
"grad_norm": 48.5,
|
| 838 |
+
"learning_rate": 1.398e-05,
|
| 839 |
+
"loss": 12.8488,
|
| 840 |
+
"mean_token_accuracy": 0.7890512771904469,
|
| 841 |
+
"num_tokens": 4955375.0,
|
| 842 |
+
"step": 700
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.28,
|
| 846 |
+
"eval_biology_entropy": 1.2677810794115068,
|
| 847 |
+
"eval_biology_loss": 1.2936662435531616,
|
| 848 |
+
"eval_biology_mean_token_accuracy": 0.6814599596261978,
|
| 849 |
+
"eval_biology_num_tokens": 4955375.0,
|
| 850 |
+
"eval_biology_runtime": 40.9028,
|
| 851 |
+
"eval_biology_samples_per_second": 12.224,
|
| 852 |
+
"eval_biology_steps_per_second": 12.224,
|
| 853 |
+
"step": 700
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 0.28,
|
| 857 |
+
"eval_math_entropy": 0.8274374225139618,
|
| 858 |
+
"eval_math_loss": 0.8181362748146057,
|
| 859 |
+
"eval_math_mean_token_accuracy": 0.7868153600692749,
|
| 860 |
+
"eval_math_num_tokens": 4955375.0,
|
| 861 |
+
"eval_math_runtime": 48.2074,
|
| 862 |
+
"eval_math_samples_per_second": 10.372,
|
| 863 |
+
"eval_math_steps_per_second": 10.372,
|
| 864 |
+
"step": 700
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"entropy": 0.7877025598660111,
|
| 868 |
+
"epoch": 0.284,
|
| 869 |
+
"grad_norm": 39.25,
|
| 870 |
+
"learning_rate": 1.418e-05,
|
| 871 |
+
"loss": 12.6293,
|
| 872 |
+
"mean_token_accuracy": 0.7944098394364119,
|
| 873 |
+
"num_tokens": 5025638.0,
|
| 874 |
+
"step": 710
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"entropy": 0.7564179511740804,
|
| 878 |
+
"epoch": 0.288,
|
| 879 |
+
"grad_norm": 38.75,
|
| 880 |
+
"learning_rate": 1.4380000000000001e-05,
|
| 881 |
+
"loss": 12.0294,
|
| 882 |
+
"mean_token_accuracy": 0.8003660906106234,
|
| 883 |
+
"num_tokens": 5095104.0,
|
| 884 |
+
"step": 720
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"entropy": 0.7647898152470589,
|
| 888 |
+
"epoch": 0.292,
|
| 889 |
+
"grad_norm": 56.0,
|
| 890 |
+
"learning_rate": 1.4580000000000001e-05,
|
| 891 |
+
"loss": 12.4713,
|
| 892 |
+
"mean_token_accuracy": 0.7947716873139143,
|
| 893 |
+
"num_tokens": 5167049.0,
|
| 894 |
+
"step": 730
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"entropy": 0.8119000235572458,
|
| 898 |
+
"epoch": 0.296,
|
| 899 |
+
"grad_norm": 36.75,
|
| 900 |
+
"learning_rate": 1.478e-05,
|
| 901 |
+
"loss": 13.0034,
|
| 902 |
+
"mean_token_accuracy": 0.7876484204083681,
|
| 903 |
+
"num_tokens": 5240988.0,
|
| 904 |
+
"step": 740
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"entropy": 0.7802535065449774,
|
| 908 |
+
"epoch": 0.3,
|
| 909 |
+
"grad_norm": 40.5,
|
| 910 |
+
"learning_rate": 1.498e-05,
|
| 911 |
+
"loss": 12.4371,
|
| 912 |
+
"mean_token_accuracy": 0.795463678240776,
|
| 913 |
+
"num_tokens": 5312258.0,
|
| 914 |
+
"step": 750
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"entropy": 0.8087782407179475,
|
| 918 |
+
"epoch": 0.304,
|
| 919 |
+
"grad_norm": 44.75,
|
| 920 |
+
"learning_rate": 1.5180000000000002e-05,
|
| 921 |
+
"loss": 12.9661,
|
| 922 |
+
"mean_token_accuracy": 0.7857579983770847,
|
| 923 |
+
"num_tokens": 5383732.0,
|
| 924 |
+
"step": 760
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"entropy": 0.8243673056364059,
|
| 928 |
+
"epoch": 0.308,
|
| 929 |
+
"grad_norm": 39.5,
|
| 930 |
+
"learning_rate": 1.5380000000000002e-05,
|
| 931 |
+
"loss": 13.1334,
|
| 932 |
+
"mean_token_accuracy": 0.7867084331810474,
|
| 933 |
+
"num_tokens": 5456650.0,
|
| 934 |
+
"step": 770
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"entropy": 0.7901528097689152,
|
| 938 |
+
"epoch": 0.312,
|
| 939 |
+
"grad_norm": 40.0,
|
| 940 |
+
"learning_rate": 1.5580000000000003e-05,
|
| 941 |
+
"loss": 12.6608,
|
| 942 |
+
"mean_token_accuracy": 0.7905277730897069,
|
| 943 |
+
"num_tokens": 5530666.0,
|
| 944 |
+
"step": 780
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"entropy": 0.7515985878184438,
|
| 948 |
+
"epoch": 0.316,
|
| 949 |
+
"grad_norm": 44.75,
|
| 950 |
+
"learning_rate": 1.578e-05,
|
| 951 |
+
"loss": 11.9655,
|
| 952 |
+
"mean_token_accuracy": 0.8011779848486185,
|
| 953 |
+
"num_tokens": 5605269.0,
|
| 954 |
+
"step": 790
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"entropy": 0.7314818985760212,
|
| 958 |
+
"epoch": 0.32,
|
| 959 |
+
"grad_norm": 42.5,
|
| 960 |
+
"learning_rate": 1.5980000000000003e-05,
|
| 961 |
+
"loss": 11.7244,
|
| 962 |
+
"mean_token_accuracy": 0.8059452049434185,
|
| 963 |
+
"num_tokens": 5676334.0,
|
| 964 |
+
"step": 800
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.32,
|
| 968 |
+
"eval_biology_entropy": 1.2663928886651994,
|
| 969 |
+
"eval_biology_loss": 1.2994052171707153,
|
| 970 |
+
"eval_biology_mean_token_accuracy": 0.6803199350833893,
|
| 971 |
+
"eval_biology_num_tokens": 5676334.0,
|
| 972 |
+
"eval_biology_runtime": 40.7632,
|
| 973 |
+
"eval_biology_samples_per_second": 12.266,
|
| 974 |
+
"eval_biology_steps_per_second": 12.266,
|
| 975 |
+
"step": 800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.32,
|
| 979 |
+
"eval_math_entropy": 0.7805578476190567,
|
| 980 |
+
"eval_math_loss": 0.8004927635192871,
|
| 981 |
+
"eval_math_mean_token_accuracy": 0.7903167794942856,
|
| 982 |
+
"eval_math_num_tokens": 5676334.0,
|
| 983 |
+
"eval_math_runtime": 48.1783,
|
| 984 |
+
"eval_math_samples_per_second": 10.378,
|
| 985 |
+
"eval_math_steps_per_second": 10.378,
|
| 986 |
+
"step": 800
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"entropy": 0.7661854576319456,
|
| 990 |
+
"epoch": 0.324,
|
| 991 |
+
"grad_norm": 40.5,
|
| 992 |
+
"learning_rate": 1.618e-05,
|
| 993 |
+
"loss": 12.5767,
|
| 994 |
+
"mean_token_accuracy": 0.7944763712584972,
|
| 995 |
+
"num_tokens": 5747779.0,
|
| 996 |
+
"step": 810
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"entropy": 0.803387301787734,
|
| 1000 |
+
"epoch": 0.328,
|
| 1001 |
+
"grad_norm": 42.0,
|
| 1002 |
+
"learning_rate": 1.638e-05,
|
| 1003 |
+
"loss": 12.825,
|
| 1004 |
+
"mean_token_accuracy": 0.7907922245562077,
|
| 1005 |
+
"num_tokens": 5816488.0,
|
| 1006 |
+
"step": 820
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"entropy": 0.8049221463501454,
|
| 1010 |
+
"epoch": 0.332,
|
| 1011 |
+
"grad_norm": 44.0,
|
| 1012 |
+
"learning_rate": 1.658e-05,
|
| 1013 |
+
"loss": 12.9669,
|
| 1014 |
+
"mean_token_accuracy": 0.7854703675955534,
|
| 1015 |
+
"num_tokens": 5887180.0,
|
| 1016 |
+
"step": 830
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"entropy": 0.8102014064788818,
|
| 1020 |
+
"epoch": 0.336,
|
| 1021 |
+
"grad_norm": 38.5,
|
| 1022 |
+
"learning_rate": 1.6780000000000002e-05,
|
| 1023 |
+
"loss": 12.9814,
|
| 1024 |
+
"mean_token_accuracy": 0.7886964596807957,
|
| 1025 |
+
"num_tokens": 5958480.0,
|
| 1026 |
+
"step": 840
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"entropy": 0.7344874950125814,
|
| 1030 |
+
"epoch": 0.34,
|
| 1031 |
+
"grad_norm": 50.25,
|
| 1032 |
+
"learning_rate": 1.698e-05,
|
| 1033 |
+
"loss": 11.7609,
|
| 1034 |
+
"mean_token_accuracy": 0.8062219381332397,
|
| 1035 |
+
"num_tokens": 6031020.0,
|
| 1036 |
+
"step": 850
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"entropy": 0.802255941182375,
|
| 1040 |
+
"epoch": 0.344,
|
| 1041 |
+
"grad_norm": 39.0,
|
| 1042 |
+
"learning_rate": 1.718e-05,
|
| 1043 |
+
"loss": 12.8439,
|
| 1044 |
+
"mean_token_accuracy": 0.7877463910728693,
|
| 1045 |
+
"num_tokens": 6102081.0,
|
| 1046 |
+
"step": 860
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"entropy": 0.7980229431763292,
|
| 1050 |
+
"epoch": 0.348,
|
| 1051 |
+
"grad_norm": 36.25,
|
| 1052 |
+
"learning_rate": 1.7380000000000003e-05,
|
| 1053 |
+
"loss": 12.7774,
|
| 1054 |
+
"mean_token_accuracy": 0.7900870595127344,
|
| 1055 |
+
"num_tokens": 6172871.0,
|
| 1056 |
+
"step": 870
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"entropy": 0.7401213280856609,
|
| 1060 |
+
"epoch": 0.352,
|
| 1061 |
+
"grad_norm": 40.75,
|
| 1062 |
+
"learning_rate": 1.758e-05,
|
| 1063 |
+
"loss": 11.9409,
|
| 1064 |
+
"mean_token_accuracy": 0.801040057092905,
|
| 1065 |
+
"num_tokens": 6242161.0,
|
| 1066 |
+
"step": 880
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"entropy": 0.8095649138092995,
|
| 1070 |
+
"epoch": 0.356,
|
| 1071 |
+
"grad_norm": 46.5,
|
| 1072 |
+
"learning_rate": 1.7780000000000003e-05,
|
| 1073 |
+
"loss": 12.9134,
|
| 1074 |
+
"mean_token_accuracy": 0.7872339155524969,
|
| 1075 |
+
"num_tokens": 6312187.0,
|
| 1076 |
+
"step": 890
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"entropy": 0.7853369496762752,
|
| 1080 |
+
"epoch": 0.36,
|
| 1081 |
+
"grad_norm": 37.75,
|
| 1082 |
+
"learning_rate": 1.798e-05,
|
| 1083 |
+
"loss": 12.4062,
|
| 1084 |
+
"mean_token_accuracy": 0.7960268270224333,
|
| 1085 |
+
"num_tokens": 6377327.0,
|
| 1086 |
+
"step": 900
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.36,
|
| 1090 |
+
"eval_biology_entropy": 1.250610340833664,
|
| 1091 |
+
"eval_biology_loss": 1.3047173023223877,
|
| 1092 |
+
"eval_biology_mean_token_accuracy": 0.6794998321533203,
|
| 1093 |
+
"eval_biology_num_tokens": 6377327.0,
|
| 1094 |
+
"eval_biology_runtime": 40.5105,
|
| 1095 |
+
"eval_biology_samples_per_second": 12.342,
|
| 1096 |
+
"eval_biology_steps_per_second": 12.342,
|
| 1097 |
+
"step": 900
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.36,
|
| 1101 |
+
"eval_math_entropy": 0.773322255730629,
|
| 1102 |
+
"eval_math_loss": 0.789590060710907,
|
| 1103 |
+
"eval_math_mean_token_accuracy": 0.7914780566692352,
|
| 1104 |
+
"eval_math_num_tokens": 6377327.0,
|
| 1105 |
+
"eval_math_runtime": 48.04,
|
| 1106 |
+
"eval_math_samples_per_second": 10.408,
|
| 1107 |
+
"eval_math_steps_per_second": 10.408,
|
| 1108 |
+
"step": 900
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"entropy": 0.763472201116383,
|
| 1112 |
+
"epoch": 0.364,
|
| 1113 |
+
"grad_norm": 41.5,
|
| 1114 |
+
"learning_rate": 1.8180000000000002e-05,
|
| 1115 |
+
"loss": 12.3288,
|
| 1116 |
+
"mean_token_accuracy": 0.7959023587405681,
|
| 1117 |
+
"num_tokens": 6453254.0,
|
| 1118 |
+
"step": 910
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"entropy": 0.7712945869192481,
|
| 1122 |
+
"epoch": 0.368,
|
| 1123 |
+
"grad_norm": 40.25,
|
| 1124 |
+
"learning_rate": 1.8380000000000004e-05,
|
| 1125 |
+
"loss": 12.2854,
|
| 1126 |
+
"mean_token_accuracy": 0.795868669450283,
|
| 1127 |
+
"num_tokens": 6523679.0,
|
| 1128 |
+
"step": 920
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"entropy": 0.7641221083700657,
|
| 1132 |
+
"epoch": 0.372,
|
| 1133 |
+
"grad_norm": 37.25,
|
| 1134 |
+
"learning_rate": 1.858e-05,
|
| 1135 |
+
"loss": 12.2867,
|
| 1136 |
+
"mean_token_accuracy": 0.7949323236942292,
|
| 1137 |
+
"num_tokens": 6594651.0,
|
| 1138 |
+
"step": 930
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"entropy": 0.7518373979255557,
|
| 1142 |
+
"epoch": 0.376,
|
| 1143 |
+
"grad_norm": 42.0,
|
| 1144 |
+
"learning_rate": 1.878e-05,
|
| 1145 |
+
"loss": 12.1179,
|
| 1146 |
+
"mean_token_accuracy": 0.798585458099842,
|
| 1147 |
+
"num_tokens": 6669278.0,
|
| 1148 |
+
"step": 940
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"entropy": 0.7543134400621057,
|
| 1152 |
+
"epoch": 0.38,
|
| 1153 |
+
"grad_norm": 33.75,
|
| 1154 |
+
"learning_rate": 1.898e-05,
|
| 1155 |
+
"loss": 12.023,
|
| 1156 |
+
"mean_token_accuracy": 0.7991646058857441,
|
| 1157 |
+
"num_tokens": 6740491.0,
|
| 1158 |
+
"step": 950
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"entropy": 0.7203706834465265,
|
| 1162 |
+
"epoch": 0.384,
|
| 1163 |
+
"grad_norm": 35.5,
|
| 1164 |
+
"learning_rate": 1.918e-05,
|
| 1165 |
+
"loss": 11.6634,
|
| 1166 |
+
"mean_token_accuracy": 0.8036353968083858,
|
| 1167 |
+
"num_tokens": 6810978.0,
|
| 1168 |
+
"step": 960
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"entropy": 0.7761434537358582,
|
| 1172 |
+
"epoch": 0.388,
|
| 1173 |
+
"grad_norm": 39.5,
|
| 1174 |
+
"learning_rate": 1.938e-05,
|
| 1175 |
+
"loss": 12.4387,
|
| 1176 |
+
"mean_token_accuracy": 0.7944561909884215,
|
| 1177 |
+
"num_tokens": 6882077.0,
|
| 1178 |
+
"step": 970
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"entropy": 0.7785217678174376,
|
| 1182 |
+
"epoch": 0.392,
|
| 1183 |
+
"grad_norm": 33.75,
|
| 1184 |
+
"learning_rate": 1.9580000000000002e-05,
|
| 1185 |
+
"loss": 12.607,
|
| 1186 |
+
"mean_token_accuracy": 0.7944069147109986,
|
| 1187 |
+
"num_tokens": 6953929.0,
|
| 1188 |
+
"step": 980
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"entropy": 0.8195465696975589,
|
| 1192 |
+
"epoch": 0.396,
|
| 1193 |
+
"grad_norm": 65.5,
|
| 1194 |
+
"learning_rate": 1.978e-05,
|
| 1195 |
+
"loss": 13.2202,
|
| 1196 |
+
"mean_token_accuracy": 0.7838506288826466,
|
| 1197 |
+
"num_tokens": 7023835.0,
|
| 1198 |
+
"step": 990
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"entropy": 0.8217529645189643,
|
| 1202 |
+
"epoch": 0.4,
|
| 1203 |
+
"grad_norm": 39.75,
|
| 1204 |
+
"learning_rate": 1.9980000000000002e-05,
|
| 1205 |
+
"loss": 13.1698,
|
| 1206 |
+
"mean_token_accuracy": 0.7848756689578295,
|
| 1207 |
+
"num_tokens": 7096903.0,
|
| 1208 |
+
"step": 1000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.4,
|
| 1212 |
+
"eval_biology_entropy": 1.266352553486824,
|
| 1213 |
+
"eval_biology_loss": 1.312301516532898,
|
| 1214 |
+
"eval_biology_mean_token_accuracy": 0.6775026069879532,
|
| 1215 |
+
"eval_biology_num_tokens": 7096903.0,
|
| 1216 |
+
"eval_biology_runtime": 40.7343,
|
| 1217 |
+
"eval_biology_samples_per_second": 12.275,
|
| 1218 |
+
"eval_biology_steps_per_second": 12.275,
|
| 1219 |
+
"step": 1000
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.4,
|
| 1223 |
+
"eval_math_entropy": 0.7719556384682655,
|
| 1224 |
+
"eval_math_loss": 0.7821868062019348,
|
| 1225 |
+
"eval_math_mean_token_accuracy": 0.7932658619880676,
|
| 1226 |
+
"eval_math_num_tokens": 7096903.0,
|
| 1227 |
+
"eval_math_runtime": 47.8946,
|
| 1228 |
+
"eval_math_samples_per_second": 10.44,
|
| 1229 |
+
"eval_math_steps_per_second": 10.44,
|
| 1230 |
+
"step": 1000
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"entropy": 0.7681734091602266,
|
| 1234 |
+
"epoch": 0.404,
|
| 1235 |
+
"grad_norm": 50.25,
|
| 1236 |
+
"learning_rate": 1.9980000000000002e-05,
|
| 1237 |
+
"loss": 12.3128,
|
| 1238 |
+
"mean_token_accuracy": 0.7956426542252302,
|
| 1239 |
+
"num_tokens": 7165970.0,
|
| 1240 |
+
"step": 1010
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"entropy": 0.7375578137114644,
|
| 1244 |
+
"epoch": 0.408,
|
| 1245 |
+
"grad_norm": 39.5,
|
| 1246 |
+
"learning_rate": 1.995777777777778e-05,
|
| 1247 |
+
"loss": 12.0172,
|
| 1248 |
+
"mean_token_accuracy": 0.7997171815484763,
|
| 1249 |
+
"num_tokens": 7239071.0,
|
| 1250 |
+
"step": 1020
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"entropy": 0.7803538812324404,
|
| 1254 |
+
"epoch": 0.412,
|
| 1255 |
+
"grad_norm": 35.75,
|
| 1256 |
+
"learning_rate": 1.9935555555555557e-05,
|
| 1257 |
+
"loss": 12.3805,
|
| 1258 |
+
"mean_token_accuracy": 0.7939646884799003,
|
| 1259 |
+
"num_tokens": 7308430.0,
|
| 1260 |
+
"step": 1030
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"entropy": 0.7741465013474226,
|
| 1264 |
+
"epoch": 0.416,
|
| 1265 |
+
"grad_norm": 35.5,
|
| 1266 |
+
"learning_rate": 1.9913333333333335e-05,
|
| 1267 |
+
"loss": 12.3671,
|
| 1268 |
+
"mean_token_accuracy": 0.7934583045542241,
|
| 1269 |
+
"num_tokens": 7377181.0,
|
| 1270 |
+
"step": 1040
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"entropy": 0.7638113724067808,
|
| 1274 |
+
"epoch": 0.42,
|
| 1275 |
+
"grad_norm": 90.5,
|
| 1276 |
+
"learning_rate": 1.9891111111111112e-05,
|
| 1277 |
+
"loss": 12.3505,
|
| 1278 |
+
"mean_token_accuracy": 0.7913536671549082,
|
| 1279 |
+
"num_tokens": 7443045.0,
|
| 1280 |
+
"step": 1050
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"entropy": 0.763846804574132,
|
| 1284 |
+
"epoch": 0.424,
|
| 1285 |
+
"grad_norm": 30.5,
|
| 1286 |
+
"learning_rate": 1.986888888888889e-05,
|
| 1287 |
+
"loss": 12.2045,
|
| 1288 |
+
"mean_token_accuracy": 0.7992881424725056,
|
| 1289 |
+
"num_tokens": 7513035.0,
|
| 1290 |
+
"step": 1060
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"entropy": 0.7383330492302775,
|
| 1294 |
+
"epoch": 0.428,
|
| 1295 |
+
"grad_norm": 34.0,
|
| 1296 |
+
"learning_rate": 1.9846666666666668e-05,
|
| 1297 |
+
"loss": 11.9274,
|
| 1298 |
+
"mean_token_accuracy": 0.80244723521173,
|
| 1299 |
+
"num_tokens": 7580676.0,
|
| 1300 |
+
"step": 1070
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"entropy": 0.7183827603235841,
|
| 1304 |
+
"epoch": 0.432,
|
| 1305 |
+
"grad_norm": 37.25,
|
| 1306 |
+
"learning_rate": 1.9824444444444445e-05,
|
| 1307 |
+
"loss": 11.5217,
|
| 1308 |
+
"mean_token_accuracy": 0.8065667234361171,
|
| 1309 |
+
"num_tokens": 7650523.0,
|
| 1310 |
+
"step": 1080
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"entropy": 0.784936741553247,
|
| 1314 |
+
"epoch": 0.436,
|
| 1315 |
+
"grad_norm": 60.5,
|
| 1316 |
+
"learning_rate": 1.9802222222222226e-05,
|
| 1317 |
+
"loss": 12.7959,
|
| 1318 |
+
"mean_token_accuracy": 0.7918864544481039,
|
| 1319 |
+
"num_tokens": 7725087.0,
|
| 1320 |
+
"step": 1090
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"entropy": 0.7581901378929615,
|
| 1324 |
+
"epoch": 0.44,
|
| 1325 |
+
"grad_norm": 35.75,
|
| 1326 |
+
"learning_rate": 1.978e-05,
|
| 1327 |
+
"loss": 12.0708,
|
| 1328 |
+
"mean_token_accuracy": 0.7991419725120068,
|
| 1329 |
+
"num_tokens": 7794006.0,
|
| 1330 |
+
"step": 1100
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 0.44,
|
| 1334 |
+
"eval_biology_entropy": 1.3096306604146957,
|
| 1335 |
+
"eval_biology_loss": 1.3199082612991333,
|
| 1336 |
+
"eval_biology_mean_token_accuracy": 0.6769449328184127,
|
| 1337 |
+
"eval_biology_num_tokens": 7794006.0,
|
| 1338 |
+
"eval_biology_runtime": 41.8088,
|
| 1339 |
+
"eval_biology_samples_per_second": 11.959,
|
| 1340 |
+
"eval_biology_steps_per_second": 11.959,
|
| 1341 |
+
"step": 1100
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.44,
|
| 1345 |
+
"eval_math_entropy": 0.7593404217660427,
|
| 1346 |
+
"eval_math_loss": 0.7711546421051025,
|
| 1347 |
+
"eval_math_mean_token_accuracy": 0.7947946254014969,
|
| 1348 |
+
"eval_math_num_tokens": 7794006.0,
|
| 1349 |
+
"eval_math_runtime": 48.5633,
|
| 1350 |
+
"eval_math_samples_per_second": 10.296,
|
| 1351 |
+
"eval_math_steps_per_second": 10.296,
|
| 1352 |
+
"step": 1100
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"entropy": 0.7571122448891401,
|
| 1356 |
+
"epoch": 0.444,
|
| 1357 |
+
"grad_norm": 35.0,
|
| 1358 |
+
"learning_rate": 1.975777777777778e-05,
|
| 1359 |
+
"loss": 12.1206,
|
| 1360 |
+
"mean_token_accuracy": 0.7960348557680845,
|
| 1361 |
+
"num_tokens": 7869588.0,
|
| 1362 |
+
"step": 1110
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"entropy": 0.7115281330421567,
|
| 1366 |
+
"epoch": 0.448,
|
| 1367 |
+
"grad_norm": 39.75,
|
| 1368 |
+
"learning_rate": 1.9735555555555556e-05,
|
| 1369 |
+
"loss": 11.603,
|
| 1370 |
+
"mean_token_accuracy": 0.8045209631323814,
|
| 1371 |
+
"num_tokens": 7936788.0,
|
| 1372 |
+
"step": 1120
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"entropy": 0.7751916225999593,
|
| 1376 |
+
"epoch": 0.452,
|
| 1377 |
+
"grad_norm": 44.5,
|
| 1378 |
+
"learning_rate": 1.9713333333333337e-05,
|
| 1379 |
+
"loss": 12.2553,
|
| 1380 |
+
"mean_token_accuracy": 0.7963129185140133,
|
| 1381 |
+
"num_tokens": 8007062.0,
|
| 1382 |
+
"step": 1130
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"entropy": 0.7358326924964785,
|
| 1386 |
+
"epoch": 0.456,
|
| 1387 |
+
"grad_norm": 35.75,
|
| 1388 |
+
"learning_rate": 1.969111111111111e-05,
|
| 1389 |
+
"loss": 11.8682,
|
| 1390 |
+
"mean_token_accuracy": 0.8016852129250764,
|
| 1391 |
+
"num_tokens": 8081035.0,
|
| 1392 |
+
"step": 1140
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"entropy": 0.748681915551424,
|
| 1396 |
+
"epoch": 0.46,
|
| 1397 |
+
"grad_norm": 38.75,
|
| 1398 |
+
"learning_rate": 1.9668888888888892e-05,
|
| 1399 |
+
"loss": 12.0865,
|
| 1400 |
+
"mean_token_accuracy": 0.8016125675290823,
|
| 1401 |
+
"num_tokens": 8152005.0,
|
| 1402 |
+
"step": 1150
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"entropy": 0.7732457166537643,
|
| 1406 |
+
"epoch": 0.464,
|
| 1407 |
+
"grad_norm": 30.625,
|
| 1408 |
+
"learning_rate": 1.9646666666666666e-05,
|
| 1409 |
+
"loss": 12.3834,
|
| 1410 |
+
"mean_token_accuracy": 0.7960082966834306,
|
| 1411 |
+
"num_tokens": 8223147.0,
|
| 1412 |
+
"step": 1160
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"entropy": 0.7652292996644974,
|
| 1416 |
+
"epoch": 0.468,
|
| 1417 |
+
"grad_norm": 47.0,
|
| 1418 |
+
"learning_rate": 1.9624444444444447e-05,
|
| 1419 |
+
"loss": 12.5645,
|
| 1420 |
+
"mean_token_accuracy": 0.7940644320100546,
|
| 1421 |
+
"num_tokens": 8296261.0,
|
| 1422 |
+
"step": 1170
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"entropy": 0.7421515263617039,
|
| 1426 |
+
"epoch": 0.472,
|
| 1427 |
+
"grad_norm": 45.75,
|
| 1428 |
+
"learning_rate": 1.9602222222222225e-05,
|
| 1429 |
+
"loss": 11.76,
|
| 1430 |
+
"mean_token_accuracy": 0.8024631194770336,
|
| 1431 |
+
"num_tokens": 8367380.0,
|
| 1432 |
+
"step": 1180
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"entropy": 0.74504639133811,
|
| 1436 |
+
"epoch": 0.476,
|
| 1437 |
+
"grad_norm": 40.25,
|
| 1438 |
+
"learning_rate": 1.9580000000000002e-05,
|
| 1439 |
+
"loss": 11.8789,
|
| 1440 |
+
"mean_token_accuracy": 0.798084694892168,
|
| 1441 |
+
"num_tokens": 8438675.0,
|
| 1442 |
+
"step": 1190
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"entropy": 0.7464414104819298,
|
| 1446 |
+
"epoch": 0.48,
|
| 1447 |
+
"grad_norm": 34.75,
|
| 1448 |
+
"learning_rate": 1.955777777777778e-05,
|
| 1449 |
+
"loss": 12.21,
|
| 1450 |
+
"mean_token_accuracy": 0.7978132009506226,
|
| 1451 |
+
"num_tokens": 8506030.0,
|
| 1452 |
+
"step": 1200
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.48,
|
| 1456 |
+
"eval_biology_entropy": 1.3110252878069877,
|
| 1457 |
+
"eval_biology_loss": 1.324716567993164,
|
| 1458 |
+
"eval_biology_mean_token_accuracy": 0.6763759871721268,
|
| 1459 |
+
"eval_biology_num_tokens": 8506030.0,
|
| 1460 |
+
"eval_biology_runtime": 67.8224,
|
| 1461 |
+
"eval_biology_samples_per_second": 7.372,
|
| 1462 |
+
"eval_biology_steps_per_second": 7.372,
|
| 1463 |
+
"step": 1200
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.48,
|
| 1467 |
+
"eval_math_entropy": 0.7597530280947685,
|
| 1468 |
+
"eval_math_loss": 0.7603512406349182,
|
| 1469 |
+
"eval_math_mean_token_accuracy": 0.7970666073560715,
|
| 1470 |
+
"eval_math_num_tokens": 8506030.0,
|
| 1471 |
+
"eval_math_runtime": 49.7375,
|
| 1472 |
+
"eval_math_samples_per_second": 10.053,
|
| 1473 |
+
"eval_math_steps_per_second": 10.053,
|
| 1474 |
+
"step": 1200
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"entropy": 0.7492084631696343,
|
| 1478 |
+
"epoch": 0.484,
|
| 1479 |
+
"grad_norm": 35.25,
|
| 1480 |
+
"learning_rate": 1.9535555555555557e-05,
|
| 1481 |
+
"loss": 11.91,
|
| 1482 |
+
"mean_token_accuracy": 0.8018037732690573,
|
| 1483 |
+
"num_tokens": 8576915.0,
|
| 1484 |
+
"step": 1210
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"entropy": 0.7045143865048885,
|
| 1488 |
+
"epoch": 0.488,
|
| 1489 |
+
"grad_norm": 33.5,
|
| 1490 |
+
"learning_rate": 1.9513333333333335e-05,
|
| 1491 |
+
"loss": 11.3318,
|
| 1492 |
+
"mean_token_accuracy": 0.8072403989732265,
|
| 1493 |
+
"num_tokens": 8646384.0,
|
| 1494 |
+
"step": 1220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"entropy": 0.7107147205621004,
|
| 1498 |
+
"epoch": 0.492,
|
| 1499 |
+
"grad_norm": 34.5,
|
| 1500 |
+
"learning_rate": 1.9491111111111113e-05,
|
| 1501 |
+
"loss": 11.4898,
|
| 1502 |
+
"mean_token_accuracy": 0.8078423041850329,
|
| 1503 |
+
"num_tokens": 8718547.0,
|
| 1504 |
+
"step": 1230
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"entropy": 0.7382461177185178,
|
| 1508 |
+
"epoch": 0.496,
|
| 1509 |
+
"grad_norm": 34.0,
|
| 1510 |
+
"learning_rate": 1.946888888888889e-05,
|
| 1511 |
+
"loss": 11.8211,
|
| 1512 |
+
"mean_token_accuracy": 0.8013716965913773,
|
| 1513 |
+
"num_tokens": 8788726.0,
|
| 1514 |
+
"step": 1240
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"entropy": 0.7180610377341509,
|
| 1518 |
+
"epoch": 0.5,
|
| 1519 |
+
"grad_norm": 32.0,
|
| 1520 |
+
"learning_rate": 1.9446666666666668e-05,
|
| 1521 |
+
"loss": 11.6359,
|
| 1522 |
+
"mean_token_accuracy": 0.8073719006031752,
|
| 1523 |
+
"num_tokens": 8861341.0,
|
| 1524 |
+
"step": 1250
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"entropy": 0.7594908589497209,
|
| 1528 |
+
"epoch": 0.504,
|
| 1529 |
+
"grad_norm": 33.25,
|
| 1530 |
+
"learning_rate": 1.9424444444444446e-05,
|
| 1531 |
+
"loss": 12.1494,
|
| 1532 |
+
"mean_token_accuracy": 0.7983963489532471,
|
| 1533 |
+
"num_tokens": 8934959.0,
|
| 1534 |
+
"step": 1260
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"entropy": 0.6865442413836718,
|
| 1538 |
+
"epoch": 0.508,
|
| 1539 |
+
"grad_norm": 47.0,
|
| 1540 |
+
"learning_rate": 1.9402222222222223e-05,
|
| 1541 |
+
"loss": 11.0336,
|
| 1542 |
+
"mean_token_accuracy": 0.8123888235539198,
|
| 1543 |
+
"num_tokens": 9008794.0,
|
| 1544 |
+
"step": 1270
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"entropy": 0.736488389596343,
|
| 1548 |
+
"epoch": 0.512,
|
| 1549 |
+
"grad_norm": 36.0,
|
| 1550 |
+
"learning_rate": 1.938e-05,
|
| 1551 |
+
"loss": 11.8236,
|
| 1552 |
+
"mean_token_accuracy": 0.80280033480376,
|
| 1553 |
+
"num_tokens": 9078039.0,
|
| 1554 |
+
"step": 1280
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"entropy": 0.7221866660751403,
|
| 1558 |
+
"epoch": 0.516,
|
| 1559 |
+
"grad_norm": 38.5,
|
| 1560 |
+
"learning_rate": 1.935777777777778e-05,
|
| 1561 |
+
"loss": 11.5549,
|
| 1562 |
+
"mean_token_accuracy": 0.8055313553661108,
|
| 1563 |
+
"num_tokens": 9148271.0,
|
| 1564 |
+
"step": 1290
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"entropy": 0.7004529038444162,
|
| 1568 |
+
"epoch": 0.52,
|
| 1569 |
+
"grad_norm": 35.5,
|
| 1570 |
+
"learning_rate": 1.9335555555555556e-05,
|
| 1571 |
+
"loss": 11.4288,
|
| 1572 |
+
"mean_token_accuracy": 0.8069275438785553,
|
| 1573 |
+
"num_tokens": 9220298.0,
|
| 1574 |
+
"step": 1300
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.52,
|
| 1578 |
+
"eval_biology_entropy": 1.2943904809951783,
|
| 1579 |
+
"eval_biology_loss": 1.3324984312057495,
|
| 1580 |
+
"eval_biology_mean_token_accuracy": 0.67458989828825,
|
| 1581 |
+
"eval_biology_num_tokens": 9220298.0,
|
| 1582 |
+
"eval_biology_runtime": 40.8416,
|
| 1583 |
+
"eval_biology_samples_per_second": 12.242,
|
| 1584 |
+
"eval_biology_steps_per_second": 12.242,
|
| 1585 |
+
"step": 1300
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.52,
|
| 1589 |
+
"eval_math_entropy": 0.7494015108048916,
|
| 1590 |
+
"eval_math_loss": 0.7516281604766846,
|
| 1591 |
+
"eval_math_mean_token_accuracy": 0.799598550081253,
|
| 1592 |
+
"eval_math_num_tokens": 9220298.0,
|
| 1593 |
+
"eval_math_runtime": 48.3706,
|
| 1594 |
+
"eval_math_samples_per_second": 10.337,
|
| 1595 |
+
"eval_math_steps_per_second": 10.337,
|
| 1596 |
+
"step": 1300
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"entropy": 0.7174371790140868,
|
| 1600 |
+
"epoch": 0.524,
|
| 1601 |
+
"grad_norm": 28.375,
|
| 1602 |
+
"learning_rate": 1.9313333333333334e-05,
|
| 1603 |
+
"loss": 11.5,
|
| 1604 |
+
"mean_token_accuracy": 0.8083903137594461,
|
| 1605 |
+
"num_tokens": 9299112.0,
|
| 1606 |
+
"step": 1310
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"entropy": 0.7080350153148174,
|
| 1610 |
+
"epoch": 0.528,
|
| 1611 |
+
"grad_norm": 30.75,
|
| 1612 |
+
"learning_rate": 1.9291111111111115e-05,
|
| 1613 |
+
"loss": 11.26,
|
| 1614 |
+
"mean_token_accuracy": 0.8113009247928857,
|
| 1615 |
+
"num_tokens": 9373860.0,
|
| 1616 |
+
"step": 1320
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"entropy": 0.7306012082844973,
|
| 1620 |
+
"epoch": 0.532,
|
| 1621 |
+
"grad_norm": 44.5,
|
| 1622 |
+
"learning_rate": 1.926888888888889e-05,
|
| 1623 |
+
"loss": 11.8475,
|
| 1624 |
+
"mean_token_accuracy": 0.8006449308246374,
|
| 1625 |
+
"num_tokens": 9446235.0,
|
| 1626 |
+
"step": 1330
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"entropy": 0.7229157039895654,
|
| 1630 |
+
"epoch": 0.536,
|
| 1631 |
+
"grad_norm": 37.25,
|
| 1632 |
+
"learning_rate": 1.924666666666667e-05,
|
| 1633 |
+
"loss": 11.6284,
|
| 1634 |
+
"mean_token_accuracy": 0.8080755949020386,
|
| 1635 |
+
"num_tokens": 9515962.0,
|
| 1636 |
+
"step": 1340
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"entropy": 0.7307992668822407,
|
| 1640 |
+
"epoch": 0.54,
|
| 1641 |
+
"grad_norm": 41.25,
|
| 1642 |
+
"learning_rate": 1.9224444444444444e-05,
|
| 1643 |
+
"loss": 11.6765,
|
| 1644 |
+
"mean_token_accuracy": 0.8042573977261782,
|
| 1645 |
+
"num_tokens": 9587715.0,
|
| 1646 |
+
"step": 1350
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"entropy": 0.7034009215421975,
|
| 1650 |
+
"epoch": 0.544,
|
| 1651 |
+
"grad_norm": 35.75,
|
| 1652 |
+
"learning_rate": 1.9202222222222225e-05,
|
| 1653 |
+
"loss": 11.7562,
|
| 1654 |
+
"mean_token_accuracy": 0.8090484531596303,
|
| 1655 |
+
"num_tokens": 9660940.0,
|
| 1656 |
+
"step": 1360
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"entropy": 0.7848182378336788,
|
| 1660 |
+
"epoch": 0.548,
|
| 1661 |
+
"grad_norm": 36.25,
|
| 1662 |
+
"learning_rate": 1.918e-05,
|
| 1663 |
+
"loss": 12.6939,
|
| 1664 |
+
"mean_token_accuracy": 0.7890455313026905,
|
| 1665 |
+
"num_tokens": 9727806.0,
|
| 1666 |
+
"step": 1370
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"entropy": 0.7604333622381091,
|
| 1670 |
+
"epoch": 0.552,
|
| 1671 |
+
"grad_norm": 39.75,
|
| 1672 |
+
"learning_rate": 1.915777777777778e-05,
|
| 1673 |
+
"loss": 12.152,
|
| 1674 |
+
"mean_token_accuracy": 0.7959740109741688,
|
| 1675 |
+
"num_tokens": 9793170.0,
|
| 1676 |
+
"step": 1380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"entropy": 0.7220884849317372,
|
| 1680 |
+
"epoch": 0.556,
|
| 1681 |
+
"grad_norm": 37.0,
|
| 1682 |
+
"learning_rate": 1.9135555555555555e-05,
|
| 1683 |
+
"loss": 11.6033,
|
| 1684 |
+
"mean_token_accuracy": 0.8036061096936464,
|
| 1685 |
+
"num_tokens": 9860484.0,
|
| 1686 |
+
"step": 1390
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"entropy": 0.704200503975153,
|
| 1690 |
+
"epoch": 0.56,
|
| 1691 |
+
"grad_norm": 29.375,
|
| 1692 |
+
"learning_rate": 1.9113333333333336e-05,
|
| 1693 |
+
"loss": 11.1518,
|
| 1694 |
+
"mean_token_accuracy": 0.8113750666379929,
|
| 1695 |
+
"num_tokens": 9932302.0,
|
| 1696 |
+
"step": 1400
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.56,
|
| 1700 |
+
"eval_biology_entropy": 1.2828473488092422,
|
| 1701 |
+
"eval_biology_loss": 1.3331190347671509,
|
| 1702 |
+
"eval_biology_mean_token_accuracy": 0.6747415412068367,
|
| 1703 |
+
"eval_biology_num_tokens": 9932302.0,
|
| 1704 |
+
"eval_biology_runtime": 45.9485,
|
| 1705 |
+
"eval_biology_samples_per_second": 10.882,
|
| 1706 |
+
"eval_biology_steps_per_second": 10.882,
|
| 1707 |
+
"step": 1400
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.56,
|
| 1711 |
+
"eval_math_entropy": 0.7451034058630467,
|
| 1712 |
+
"eval_math_loss": 0.7451240420341492,
|
| 1713 |
+
"eval_math_mean_token_accuracy": 0.8007836003303528,
|
| 1714 |
+
"eval_math_num_tokens": 9932302.0,
|
| 1715 |
+
"eval_math_runtime": 56.801,
|
| 1716 |
+
"eval_math_samples_per_second": 8.803,
|
| 1717 |
+
"eval_math_steps_per_second": 8.803,
|
| 1718 |
+
"step": 1400
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"entropy": 0.7115965198725462,
|
| 1722 |
+
"epoch": 0.564,
|
| 1723 |
+
"grad_norm": 34.25,
|
| 1724 |
+
"learning_rate": 1.9091111111111113e-05,
|
| 1725 |
+
"loss": 11.5991,
|
| 1726 |
+
"mean_token_accuracy": 0.8047812730073929,
|
| 1727 |
+
"num_tokens": 10001923.0,
|
| 1728 |
+
"step": 1410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"entropy": 0.7081295600160956,
|
| 1732 |
+
"epoch": 0.568,
|
| 1733 |
+
"grad_norm": 32.75,
|
| 1734 |
+
"learning_rate": 1.906888888888889e-05,
|
| 1735 |
+
"loss": 11.2206,
|
| 1736 |
+
"mean_token_accuracy": 0.8095470562577247,
|
| 1737 |
+
"num_tokens": 10072362.0,
|
| 1738 |
+
"step": 1420
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"entropy": 0.7141612969338894,
|
| 1742 |
+
"epoch": 0.572,
|
| 1743 |
+
"grad_norm": 30.375,
|
| 1744 |
+
"learning_rate": 1.904666666666667e-05,
|
| 1745 |
+
"loss": 11.6816,
|
| 1746 |
+
"mean_token_accuracy": 0.8019938107579947,
|
| 1747 |
+
"num_tokens": 10140236.0,
|
| 1748 |
+
"step": 1430
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"entropy": 0.7410996645689011,
|
| 1752 |
+
"epoch": 0.576,
|
| 1753 |
+
"grad_norm": 32.5,
|
| 1754 |
+
"learning_rate": 1.9024444444444446e-05,
|
| 1755 |
+
"loss": 11.7841,
|
| 1756 |
+
"mean_token_accuracy": 0.8034846507012844,
|
| 1757 |
+
"num_tokens": 10215462.0,
|
| 1758 |
+
"step": 1440
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"entropy": 0.7485123924911022,
|
| 1762 |
+
"epoch": 0.58,
|
| 1763 |
+
"grad_norm": 43.25,
|
| 1764 |
+
"learning_rate": 1.9002222222222224e-05,
|
| 1765 |
+
"loss": 11.9787,
|
| 1766 |
+
"mean_token_accuracy": 0.798300775885582,
|
| 1767 |
+
"num_tokens": 10286386.0,
|
| 1768 |
+
"step": 1450
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"entropy": 0.7154224492609501,
|
| 1772 |
+
"epoch": 0.584,
|
| 1773 |
+
"grad_norm": 32.5,
|
| 1774 |
+
"learning_rate": 1.898e-05,
|
| 1775 |
+
"loss": 11.5298,
|
| 1776 |
+
"mean_token_accuracy": 0.8057492181658745,
|
| 1777 |
+
"num_tokens": 10354952.0,
|
| 1778 |
+
"step": 1460
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"entropy": 0.7272289376705885,
|
| 1782 |
+
"epoch": 0.588,
|
| 1783 |
+
"grad_norm": 40.25,
|
| 1784 |
+
"learning_rate": 1.895777777777778e-05,
|
| 1785 |
+
"loss": 11.652,
|
| 1786 |
+
"mean_token_accuracy": 0.8025094602257014,
|
| 1787 |
+
"num_tokens": 10425215.0,
|
| 1788 |
+
"step": 1470
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"entropy": 0.7166844455525279,
|
| 1792 |
+
"epoch": 0.592,
|
| 1793 |
+
"grad_norm": 30.625,
|
| 1794 |
+
"learning_rate": 1.8935555555555556e-05,
|
| 1795 |
+
"loss": 11.4116,
|
| 1796 |
+
"mean_token_accuracy": 0.8082372698932886,
|
| 1797 |
+
"num_tokens": 10504396.0,
|
| 1798 |
+
"step": 1480
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"entropy": 0.7413169213570654,
|
| 1802 |
+
"epoch": 0.596,
|
| 1803 |
+
"grad_norm": 32.0,
|
| 1804 |
+
"learning_rate": 1.8913333333333334e-05,
|
| 1805 |
+
"loss": 11.959,
|
| 1806 |
+
"mean_token_accuracy": 0.7982309013605118,
|
| 1807 |
+
"num_tokens": 10576150.0,
|
| 1808 |
+
"step": 1490
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"entropy": 0.7089840611442924,
|
| 1812 |
+
"epoch": 0.6,
|
| 1813 |
+
"grad_norm": 35.0,
|
| 1814 |
+
"learning_rate": 1.8891111111111115e-05,
|
| 1815 |
+
"loss": 11.6029,
|
| 1816 |
+
"mean_token_accuracy": 0.8050658291205763,
|
| 1817 |
+
"num_tokens": 10644641.0,
|
| 1818 |
+
"step": 1500
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.6,
|
| 1822 |
+
"eval_biology_entropy": 1.301250998735428,
|
| 1823 |
+
"eval_biology_loss": 1.3366436958312988,
|
| 1824 |
+
"eval_biology_mean_token_accuracy": 0.6739367877840996,
|
| 1825 |
+
"eval_biology_num_tokens": 10644641.0,
|
| 1826 |
+
"eval_biology_runtime": 40.559,
|
| 1827 |
+
"eval_biology_samples_per_second": 12.328,
|
| 1828 |
+
"eval_biology_steps_per_second": 12.328,
|
| 1829 |
+
"step": 1500
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.6,
|
| 1833 |
+
"eval_math_entropy": 0.7238052343726158,
|
| 1834 |
+
"eval_math_loss": 0.7381152510643005,
|
| 1835 |
+
"eval_math_mean_token_accuracy": 0.8017449014186859,
|
| 1836 |
+
"eval_math_num_tokens": 10644641.0,
|
| 1837 |
+
"eval_math_runtime": 48.6353,
|
| 1838 |
+
"eval_math_samples_per_second": 10.281,
|
| 1839 |
+
"eval_math_steps_per_second": 10.281,
|
| 1840 |
+
"step": 1500
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"entropy": 0.7384224113076925,
|
| 1844 |
+
"epoch": 0.604,
|
| 1845 |
+
"grad_norm": 31.375,
|
| 1846 |
+
"learning_rate": 1.886888888888889e-05,
|
| 1847 |
+
"loss": 11.6601,
|
| 1848 |
+
"mean_token_accuracy": 0.8054828774183989,
|
| 1849 |
+
"num_tokens": 10718879.0,
|
| 1850 |
+
"step": 1510
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"entropy": 0.7157115396112204,
|
| 1854 |
+
"epoch": 0.608,
|
| 1855 |
+
"grad_norm": 38.0,
|
| 1856 |
+
"learning_rate": 1.884666666666667e-05,
|
| 1857 |
+
"loss": 11.7253,
|
| 1858 |
+
"mean_token_accuracy": 0.805310570448637,
|
| 1859 |
+
"num_tokens": 10793126.0,
|
| 1860 |
+
"step": 1520
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"entropy": 0.7220246642827988,
|
| 1864 |
+
"epoch": 0.612,
|
| 1865 |
+
"grad_norm": 31.875,
|
| 1866 |
+
"learning_rate": 1.8824444444444445e-05,
|
| 1867 |
+
"loss": 11.4871,
|
| 1868 |
+
"mean_token_accuracy": 0.8036613553762436,
|
| 1869 |
+
"num_tokens": 10868501.0,
|
| 1870 |
+
"step": 1530
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"entropy": 0.7019952710717916,
|
| 1874 |
+
"epoch": 0.616,
|
| 1875 |
+
"grad_norm": 37.75,
|
| 1876 |
+
"learning_rate": 1.8802222222222226e-05,
|
| 1877 |
+
"loss": 11.3501,
|
| 1878 |
+
"mean_token_accuracy": 0.8096903506666422,
|
| 1879 |
+
"num_tokens": 10941399.0,
|
| 1880 |
+
"step": 1540
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"entropy": 0.7124224284663796,
|
| 1884 |
+
"epoch": 0.62,
|
| 1885 |
+
"grad_norm": 28.0,
|
| 1886 |
+
"learning_rate": 1.878e-05,
|
| 1887 |
+
"loss": 11.2568,
|
| 1888 |
+
"mean_token_accuracy": 0.8094121634960174,
|
| 1889 |
+
"num_tokens": 11013227.0,
|
| 1890 |
+
"step": 1550
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"entropy": 0.6985575210303068,
|
| 1894 |
+
"epoch": 0.624,
|
| 1895 |
+
"grad_norm": 41.0,
|
| 1896 |
+
"learning_rate": 1.875777777777778e-05,
|
| 1897 |
+
"loss": 11.3067,
|
| 1898 |
+
"mean_token_accuracy": 0.8099827248603105,
|
| 1899 |
+
"num_tokens": 11081768.0,
|
| 1900 |
+
"step": 1560
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"entropy": 0.7057704029604792,
|
| 1904 |
+
"epoch": 0.628,
|
| 1905 |
+
"grad_norm": 35.5,
|
| 1906 |
+
"learning_rate": 1.873555555555556e-05,
|
| 1907 |
+
"loss": 11.2109,
|
| 1908 |
+
"mean_token_accuracy": 0.8099541310220957,
|
| 1909 |
+
"num_tokens": 11155362.0,
|
| 1910 |
+
"step": 1570
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"entropy": 0.6665219226852059,
|
| 1914 |
+
"epoch": 0.632,
|
| 1915 |
+
"grad_norm": 33.5,
|
| 1916 |
+
"learning_rate": 1.8713333333333336e-05,
|
| 1917 |
+
"loss": 10.7041,
|
| 1918 |
+
"mean_token_accuracy": 0.8133308228105307,
|
| 1919 |
+
"num_tokens": 11226419.0,
|
| 1920 |
+
"step": 1580
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"entropy": 0.6918639421463013,
|
| 1924 |
+
"epoch": 0.636,
|
| 1925 |
+
"grad_norm": 28.0,
|
| 1926 |
+
"learning_rate": 1.8691111111111114e-05,
|
| 1927 |
+
"loss": 11.1789,
|
| 1928 |
+
"mean_token_accuracy": 0.810427363216877,
|
| 1929 |
+
"num_tokens": 11298800.0,
|
| 1930 |
+
"step": 1590
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"entropy": 0.7074929376132786,
|
| 1934 |
+
"epoch": 0.64,
|
| 1935 |
+
"grad_norm": 32.0,
|
| 1936 |
+
"learning_rate": 1.866888888888889e-05,
|
| 1937 |
+
"loss": 11.4364,
|
| 1938 |
+
"mean_token_accuracy": 0.8062872476875782,
|
| 1939 |
+
"num_tokens": 11370320.0,
|
| 1940 |
+
"step": 1600
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 0.64,
|
| 1944 |
+
"eval_biology_entropy": 1.2752835777401925,
|
| 1945 |
+
"eval_biology_loss": 1.3371055126190186,
|
| 1946 |
+
"eval_biology_mean_token_accuracy": 0.6743674901127815,
|
| 1947 |
+
"eval_biology_num_tokens": 11370320.0,
|
| 1948 |
+
"eval_biology_runtime": 41.1859,
|
| 1949 |
+
"eval_biology_samples_per_second": 12.14,
|
| 1950 |
+
"eval_biology_steps_per_second": 12.14,
|
| 1951 |
+
"step": 1600
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.64,
|
| 1955 |
+
"eval_math_entropy": 0.7216429207324981,
|
| 1956 |
+
"eval_math_loss": 0.7321795225143433,
|
| 1957 |
+
"eval_math_mean_token_accuracy": 0.8027968648672104,
|
| 1958 |
+
"eval_math_num_tokens": 11370320.0,
|
| 1959 |
+
"eval_math_runtime": 50.051,
|
| 1960 |
+
"eval_math_samples_per_second": 9.99,
|
| 1961 |
+
"eval_math_steps_per_second": 9.99,
|
| 1962 |
+
"step": 1600
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"entropy": 0.7110869940370321,
|
| 1966 |
+
"epoch": 0.644,
|
| 1967 |
+
"grad_norm": 33.5,
|
| 1968 |
+
"learning_rate": 1.864666666666667e-05,
|
| 1969 |
+
"loss": 11.4826,
|
| 1970 |
+
"mean_token_accuracy": 0.8066194530576467,
|
| 1971 |
+
"num_tokens": 11445577.0,
|
| 1972 |
+
"step": 1610
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"entropy": 0.6680682392790913,
|
| 1976 |
+
"epoch": 0.648,
|
| 1977 |
+
"grad_norm": 41.75,
|
| 1978 |
+
"learning_rate": 1.8624444444444446e-05,
|
| 1979 |
+
"loss": 10.6001,
|
| 1980 |
+
"mean_token_accuracy": 0.8180687021464109,
|
| 1981 |
+
"num_tokens": 11515332.0,
|
| 1982 |
+
"step": 1620
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"entropy": 0.6972724428400398,
|
| 1986 |
+
"epoch": 0.652,
|
| 1987 |
+
"grad_norm": 28.375,
|
| 1988 |
+
"learning_rate": 1.8602222222222224e-05,
|
| 1989 |
+
"loss": 11.2625,
|
| 1990 |
+
"mean_token_accuracy": 0.8118615668267012,
|
| 1991 |
+
"num_tokens": 11589344.0,
|
| 1992 |
+
"step": 1630
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"entropy": 0.7117041515186429,
|
| 1996 |
+
"epoch": 0.656,
|
| 1997 |
+
"grad_norm": 30.125,
|
| 1998 |
+
"learning_rate": 1.858e-05,
|
| 1999 |
+
"loss": 11.2332,
|
| 2000 |
+
"mean_token_accuracy": 0.8057843204587698,
|
| 2001 |
+
"num_tokens": 11657835.0,
|
| 2002 |
+
"step": 1640
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"entropy": 0.7337123662233352,
|
| 2006 |
+
"epoch": 0.66,
|
| 2007 |
+
"grad_norm": 35.0,
|
| 2008 |
+
"learning_rate": 1.855777777777778e-05,
|
| 2009 |
+
"loss": 11.9959,
|
| 2010 |
+
"mean_token_accuracy": 0.7986002914607525,
|
| 2011 |
+
"num_tokens": 11728761.0,
|
| 2012 |
+
"step": 1650
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"entropy": 0.6996571996249259,
|
| 2016 |
+
"epoch": 0.664,
|
| 2017 |
+
"grad_norm": 36.0,
|
| 2018 |
+
"learning_rate": 1.8535555555555557e-05,
|
| 2019 |
+
"loss": 11.225,
|
| 2020 |
+
"mean_token_accuracy": 0.807379187270999,
|
| 2021 |
+
"num_tokens": 11799325.0,
|
| 2022 |
+
"step": 1660
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"entropy": 0.7201200375333429,
|
| 2026 |
+
"epoch": 0.668,
|
| 2027 |
+
"grad_norm": 26.625,
|
| 2028 |
+
"learning_rate": 1.8513333333333335e-05,
|
| 2029 |
+
"loss": 11.4655,
|
| 2030 |
+
"mean_token_accuracy": 0.8065472077578306,
|
| 2031 |
+
"num_tokens": 11875761.0,
|
| 2032 |
+
"step": 1670
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"entropy": 0.6706034051254391,
|
| 2036 |
+
"epoch": 0.672,
|
| 2037 |
+
"grad_norm": 29.5,
|
| 2038 |
+
"learning_rate": 1.8491111111111112e-05,
|
| 2039 |
+
"loss": 11.0248,
|
| 2040 |
+
"mean_token_accuracy": 0.8108973000198603,
|
| 2041 |
+
"num_tokens": 11949262.0,
|
| 2042 |
+
"step": 1680
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"entropy": 0.7067649567499756,
|
| 2046 |
+
"epoch": 0.676,
|
| 2047 |
+
"grad_norm": 39.0,
|
| 2048 |
+
"learning_rate": 1.846888888888889e-05,
|
| 2049 |
+
"loss": 11.2029,
|
| 2050 |
+
"mean_token_accuracy": 0.8082552354782819,
|
| 2051 |
+
"num_tokens": 12017021.0,
|
| 2052 |
+
"step": 1690
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"entropy": 0.7115105454809963,
|
| 2056 |
+
"epoch": 0.68,
|
| 2057 |
+
"grad_norm": 28.375,
|
| 2058 |
+
"learning_rate": 1.8446666666666667e-05,
|
| 2059 |
+
"loss": 11.5299,
|
| 2060 |
+
"mean_token_accuracy": 0.8070931307971477,
|
| 2061 |
+
"num_tokens": 12090029.0,
|
| 2062 |
+
"step": 1700
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.68,
|
| 2066 |
+
"eval_biology_entropy": 1.3190836807489394,
|
| 2067 |
+
"eval_biology_loss": 1.3399405479431152,
|
| 2068 |
+
"eval_biology_mean_token_accuracy": 0.6731415511965751,
|
| 2069 |
+
"eval_biology_num_tokens": 12090029.0,
|
| 2070 |
+
"eval_biology_runtime": 41.5684,
|
| 2071 |
+
"eval_biology_samples_per_second": 12.028,
|
| 2072 |
+
"eval_biology_steps_per_second": 12.028,
|
| 2073 |
+
"step": 1700
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.68,
|
| 2077 |
+
"eval_math_entropy": 0.7365337611436844,
|
| 2078 |
+
"eval_math_loss": 0.7248461842536926,
|
| 2079 |
+
"eval_math_mean_token_accuracy": 0.8049273964166641,
|
| 2080 |
+
"eval_math_num_tokens": 12090029.0,
|
| 2081 |
+
"eval_math_runtime": 48.658,
|
| 2082 |
+
"eval_math_samples_per_second": 10.276,
|
| 2083 |
+
"eval_math_steps_per_second": 10.276,
|
| 2084 |
+
"step": 1700
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"entropy": 0.7276539742946625,
|
| 2088 |
+
"epoch": 0.684,
|
| 2089 |
+
"grad_norm": 45.75,
|
| 2090 |
+
"learning_rate": 1.842444444444445e-05,
|
| 2091 |
+
"loss": 11.6271,
|
| 2092 |
+
"mean_token_accuracy": 0.8031581114977598,
|
| 2093 |
+
"num_tokens": 12157148.0,
|
| 2094 |
+
"step": 1710
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"entropy": 0.692380516603589,
|
| 2098 |
+
"epoch": 0.688,
|
| 2099 |
+
"grad_norm": 32.75,
|
| 2100 |
+
"learning_rate": 1.8402222222222223e-05,
|
| 2101 |
+
"loss": 11.0596,
|
| 2102 |
+
"mean_token_accuracy": 0.8091447809711099,
|
| 2103 |
+
"num_tokens": 12227640.0,
|
| 2104 |
+
"step": 1720
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"entropy": 0.7150112164206803,
|
| 2108 |
+
"epoch": 0.692,
|
| 2109 |
+
"grad_norm": 31.5,
|
| 2110 |
+
"learning_rate": 1.8380000000000004e-05,
|
| 2111 |
+
"loss": 11.4079,
|
| 2112 |
+
"mean_token_accuracy": 0.8083063390105962,
|
| 2113 |
+
"num_tokens": 12300485.0,
|
| 2114 |
+
"step": 1730
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"entropy": 0.707249708659947,
|
| 2118 |
+
"epoch": 0.696,
|
| 2119 |
+
"grad_norm": 44.5,
|
| 2120 |
+
"learning_rate": 1.8357777777777778e-05,
|
| 2121 |
+
"loss": 11.4663,
|
| 2122 |
+
"mean_token_accuracy": 0.8062439143657685,
|
| 2123 |
+
"num_tokens": 12371560.0,
|
| 2124 |
+
"step": 1740
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"entropy": 0.6466306034475565,
|
| 2128 |
+
"epoch": 0.7,
|
| 2129 |
+
"grad_norm": 31.5,
|
| 2130 |
+
"learning_rate": 1.833555555555556e-05,
|
| 2131 |
+
"loss": 10.4753,
|
| 2132 |
+
"mean_token_accuracy": 0.8210283190011978,
|
| 2133 |
+
"num_tokens": 12444485.0,
|
| 2134 |
+
"step": 1750
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"entropy": 0.7072455288842321,
|
| 2138 |
+
"epoch": 0.704,
|
| 2139 |
+
"grad_norm": 28.625,
|
| 2140 |
+
"learning_rate": 1.8313333333333333e-05,
|
| 2141 |
+
"loss": 11.2585,
|
| 2142 |
+
"mean_token_accuracy": 0.8080747354775667,
|
| 2143 |
+
"num_tokens": 12516641.0,
|
| 2144 |
+
"step": 1760
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"entropy": 0.6972917837090791,
|
| 2148 |
+
"epoch": 0.708,
|
| 2149 |
+
"grad_norm": 31.875,
|
| 2150 |
+
"learning_rate": 1.8291111111111114e-05,
|
| 2151 |
+
"loss": 11.1016,
|
| 2152 |
+
"mean_token_accuracy": 0.8106624346226454,
|
| 2153 |
+
"num_tokens": 12586969.0,
|
| 2154 |
+
"step": 1770
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"entropy": 0.733701079338789,
|
| 2158 |
+
"epoch": 0.712,
|
| 2159 |
+
"grad_norm": 33.0,
|
| 2160 |
+
"learning_rate": 1.8268888888888888e-05,
|
| 2161 |
+
"loss": 11.9791,
|
| 2162 |
+
"mean_token_accuracy": 0.8003675427287817,
|
| 2163 |
+
"num_tokens": 12656623.0,
|
| 2164 |
+
"step": 1780
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"entropy": 0.7007612373679877,
|
| 2168 |
+
"epoch": 0.716,
|
| 2169 |
+
"grad_norm": 32.5,
|
| 2170 |
+
"learning_rate": 1.824666666666667e-05,
|
| 2171 |
+
"loss": 11.3258,
|
| 2172 |
+
"mean_token_accuracy": 0.8079633958637714,
|
| 2173 |
+
"num_tokens": 12724402.0,
|
| 2174 |
+
"step": 1790
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"entropy": 0.7202634891495109,
|
| 2178 |
+
"epoch": 0.72,
|
| 2179 |
+
"grad_norm": 45.25,
|
| 2180 |
+
"learning_rate": 1.8224444444444447e-05,
|
| 2181 |
+
"loss": 11.4906,
|
| 2182 |
+
"mean_token_accuracy": 0.8054792225360871,
|
| 2183 |
+
"num_tokens": 12793343.0,
|
| 2184 |
+
"step": 1800
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.72,
|
| 2188 |
+
"eval_biology_entropy": 1.2490641915798186,
|
| 2189 |
+
"eval_biology_loss": 1.3441540002822876,
|
| 2190 |
+
"eval_biology_mean_token_accuracy": 0.6730483353734017,
|
| 2191 |
+
"eval_biology_num_tokens": 12793343.0,
|
| 2192 |
+
"eval_biology_runtime": 41.3736,
|
| 2193 |
+
"eval_biology_samples_per_second": 12.085,
|
| 2194 |
+
"eval_biology_steps_per_second": 12.085,
|
| 2195 |
+
"step": 1800
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.72,
|
| 2199 |
+
"eval_math_entropy": 0.6915989356935024,
|
| 2200 |
+
"eval_math_loss": 0.7225330471992493,
|
| 2201 |
+
"eval_math_mean_token_accuracy": 0.8046257610321045,
|
| 2202 |
+
"eval_math_num_tokens": 12793343.0,
|
| 2203 |
+
"eval_math_runtime": 48.3235,
|
| 2204 |
+
"eval_math_samples_per_second": 10.347,
|
| 2205 |
+
"eval_math_steps_per_second": 10.347,
|
| 2206 |
+
"step": 1800
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"entropy": 0.689653181284666,
|
| 2210 |
+
"epoch": 0.724,
|
| 2211 |
+
"grad_norm": 30.0,
|
| 2212 |
+
"learning_rate": 1.8202222222222225e-05,
|
| 2213 |
+
"loss": 11.095,
|
| 2214 |
+
"mean_token_accuracy": 0.8116659399121999,
|
| 2215 |
+
"num_tokens": 12862261.0,
|
| 2216 |
+
"step": 1810
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"entropy": 0.6818234785459936,
|
| 2220 |
+
"epoch": 0.728,
|
| 2221 |
+
"grad_norm": 33.25,
|
| 2222 |
+
"learning_rate": 1.8180000000000002e-05,
|
| 2223 |
+
"loss": 11.0911,
|
| 2224 |
+
"mean_token_accuracy": 0.812104906514287,
|
| 2225 |
+
"num_tokens": 12932479.0,
|
| 2226 |
+
"step": 1820
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"entropy": 0.7217532116919756,
|
| 2230 |
+
"epoch": 0.732,
|
| 2231 |
+
"grad_norm": 39.0,
|
| 2232 |
+
"learning_rate": 1.815777777777778e-05,
|
| 2233 |
+
"loss": 11.5997,
|
| 2234 |
+
"mean_token_accuracy": 0.80367177054286,
|
| 2235 |
+
"num_tokens": 13005640.0,
|
| 2236 |
+
"step": 1830
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"entropy": 0.678720659762621,
|
| 2240 |
+
"epoch": 0.736,
|
| 2241 |
+
"grad_norm": 30.25,
|
| 2242 |
+
"learning_rate": 1.8135555555555557e-05,
|
| 2243 |
+
"loss": 10.8064,
|
| 2244 |
+
"mean_token_accuracy": 0.817896718531847,
|
| 2245 |
+
"num_tokens": 13077981.0,
|
| 2246 |
+
"step": 1840
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"entropy": 0.6725755118764937,
|
| 2250 |
+
"epoch": 0.74,
|
| 2251 |
+
"grad_norm": 32.25,
|
| 2252 |
+
"learning_rate": 1.8113333333333335e-05,
|
| 2253 |
+
"loss": 10.8668,
|
| 2254 |
+
"mean_token_accuracy": 0.8124937634915114,
|
| 2255 |
+
"num_tokens": 13147662.0,
|
| 2256 |
+
"step": 1850
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"entropy": 0.6931444002315402,
|
| 2260 |
+
"epoch": 0.744,
|
| 2261 |
+
"grad_norm": 29.75,
|
| 2262 |
+
"learning_rate": 1.8091111111111113e-05,
|
| 2263 |
+
"loss": 10.9939,
|
| 2264 |
+
"mean_token_accuracy": 0.8126753181219101,
|
| 2265 |
+
"num_tokens": 13217888.0,
|
| 2266 |
+
"step": 1860
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"entropy": 0.6799735228531063,
|
| 2270 |
+
"epoch": 0.748,
|
| 2271 |
+
"grad_norm": 30.625,
|
| 2272 |
+
"learning_rate": 1.806888888888889e-05,
|
| 2273 |
+
"loss": 11.1209,
|
| 2274 |
+
"mean_token_accuracy": 0.8113108247518539,
|
| 2275 |
+
"num_tokens": 13290215.0,
|
| 2276 |
+
"step": 1870
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"entropy": 0.6741954877972602,
|
| 2280 |
+
"epoch": 0.752,
|
| 2281 |
+
"grad_norm": 37.25,
|
| 2282 |
+
"learning_rate": 1.8046666666666668e-05,
|
| 2283 |
+
"loss": 10.68,
|
| 2284 |
+
"mean_token_accuracy": 0.8178396210074425,
|
| 2285 |
+
"num_tokens": 13358957.0,
|
| 2286 |
+
"step": 1880
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"entropy": 0.6795736599713564,
|
| 2290 |
+
"epoch": 0.756,
|
| 2291 |
+
"grad_norm": 27.5,
|
| 2292 |
+
"learning_rate": 1.8024444444444445e-05,
|
| 2293 |
+
"loss": 10.7669,
|
| 2294 |
+
"mean_token_accuracy": 0.8167448993772268,
|
| 2295 |
+
"num_tokens": 13435490.0,
|
| 2296 |
+
"step": 1890
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"entropy": 0.6350338911637664,
|
| 2300 |
+
"epoch": 0.76,
|
| 2301 |
+
"grad_norm": 33.75,
|
| 2302 |
+
"learning_rate": 1.8002222222222223e-05,
|
| 2303 |
+
"loss": 10.293,
|
| 2304 |
+
"mean_token_accuracy": 0.8226836711168289,
|
| 2305 |
+
"num_tokens": 13510850.0,
|
| 2306 |
+
"step": 1900
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.76,
|
| 2310 |
+
"eval_biology_entropy": 1.2716083939671516,
|
| 2311 |
+
"eval_biology_loss": 1.343294620513916,
|
| 2312 |
+
"eval_biology_mean_token_accuracy": 0.6733359169363976,
|
| 2313 |
+
"eval_biology_num_tokens": 13510850.0,
|
| 2314 |
+
"eval_biology_runtime": 40.8464,
|
| 2315 |
+
"eval_biology_samples_per_second": 12.241,
|
| 2316 |
+
"eval_biology_steps_per_second": 12.241,
|
| 2317 |
+
"step": 1900
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 0.76,
|
| 2321 |
+
"eval_math_entropy": 0.7108331805765629,
|
| 2322 |
+
"eval_math_loss": 0.7166345715522766,
|
| 2323 |
+
"eval_math_mean_token_accuracy": 0.8057290604114532,
|
| 2324 |
+
"eval_math_num_tokens": 13510850.0,
|
| 2325 |
+
"eval_math_runtime": 47.8127,
|
| 2326 |
+
"eval_math_samples_per_second": 10.457,
|
| 2327 |
+
"eval_math_steps_per_second": 10.457,
|
| 2328 |
+
"step": 1900
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"entropy": 0.6852852246724069,
|
| 2332 |
+
"epoch": 0.764,
|
| 2333 |
+
"grad_norm": 26.0,
|
| 2334 |
+
"learning_rate": 1.798e-05,
|
| 2335 |
+
"loss": 10.9532,
|
| 2336 |
+
"mean_token_accuracy": 0.8128071770071983,
|
| 2337 |
+
"num_tokens": 13581848.0,
|
| 2338 |
+
"step": 1910
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"entropy": 0.6663387446664274,
|
| 2342 |
+
"epoch": 0.768,
|
| 2343 |
+
"grad_norm": 29.125,
|
| 2344 |
+
"learning_rate": 1.7957777777777778e-05,
|
| 2345 |
+
"loss": 10.8527,
|
| 2346 |
+
"mean_token_accuracy": 0.8174721624702215,
|
| 2347 |
+
"num_tokens": 13653412.0,
|
| 2348 |
+
"step": 1920
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"entropy": 0.7165529853664339,
|
| 2352 |
+
"epoch": 0.772,
|
| 2353 |
+
"grad_norm": 31.125,
|
| 2354 |
+
"learning_rate": 1.7935555555555556e-05,
|
| 2355 |
+
"loss": 11.2507,
|
| 2356 |
+
"mean_token_accuracy": 0.8098738107830286,
|
| 2357 |
+
"num_tokens": 13721740.0,
|
| 2358 |
+
"step": 1930
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"entropy": 0.6845185202546418,
|
| 2362 |
+
"epoch": 0.776,
|
| 2363 |
+
"grad_norm": 27.25,
|
| 2364 |
+
"learning_rate": 1.7913333333333337e-05,
|
| 2365 |
+
"loss": 11.1313,
|
| 2366 |
+
"mean_token_accuracy": 0.810427762567997,
|
| 2367 |
+
"num_tokens": 13794252.0,
|
| 2368 |
+
"step": 1940
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"entropy": 0.693409715499729,
|
| 2372 |
+
"epoch": 0.78,
|
| 2373 |
+
"grad_norm": 37.25,
|
| 2374 |
+
"learning_rate": 1.789111111111111e-05,
|
| 2375 |
+
"loss": 11.1187,
|
| 2376 |
+
"mean_token_accuracy": 0.8099165614694357,
|
| 2377 |
+
"num_tokens": 13865839.0,
|
| 2378 |
+
"step": 1950
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"entropy": 0.707060054782778,
|
| 2382 |
+
"epoch": 0.784,
|
| 2383 |
+
"grad_norm": 81.5,
|
| 2384 |
+
"learning_rate": 1.7868888888888892e-05,
|
| 2385 |
+
"loss": 11.428,
|
| 2386 |
+
"mean_token_accuracy": 0.8070406079292297,
|
| 2387 |
+
"num_tokens": 13940856.0,
|
| 2388 |
+
"step": 1960
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"entropy": 0.7144587634131312,
|
| 2392 |
+
"epoch": 0.788,
|
| 2393 |
+
"grad_norm": 34.0,
|
| 2394 |
+
"learning_rate": 1.7846666666666666e-05,
|
| 2395 |
+
"loss": 11.4499,
|
| 2396 |
+
"mean_token_accuracy": 0.8067739550024271,
|
| 2397 |
+
"num_tokens": 14014568.0,
|
| 2398 |
+
"step": 1970
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"entropy": 0.6812865046784282,
|
| 2402 |
+
"epoch": 0.792,
|
| 2403 |
+
"grad_norm": 29.75,
|
| 2404 |
+
"learning_rate": 1.7824444444444447e-05,
|
| 2405 |
+
"loss": 10.9827,
|
| 2406 |
+
"mean_token_accuracy": 0.8135220106691122,
|
| 2407 |
+
"num_tokens": 14086117.0,
|
| 2408 |
+
"step": 1980
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"entropy": 0.7018590277060867,
|
| 2412 |
+
"epoch": 0.796,
|
| 2413 |
+
"grad_norm": 28.875,
|
| 2414 |
+
"learning_rate": 1.780222222222222e-05,
|
| 2415 |
+
"loss": 11.4353,
|
| 2416 |
+
"mean_token_accuracy": 0.8069973647594452,
|
| 2417 |
+
"num_tokens": 14160050.0,
|
| 2418 |
+
"step": 1990
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"entropy": 0.7179917026311159,
|
| 2422 |
+
"epoch": 0.8,
|
| 2423 |
+
"grad_norm": 33.0,
|
| 2424 |
+
"learning_rate": 1.7780000000000003e-05,
|
| 2425 |
+
"loss": 11.3243,
|
| 2426 |
+
"mean_token_accuracy": 0.8104559704661369,
|
| 2427 |
+
"num_tokens": 14230754.0,
|
| 2428 |
+
"step": 2000
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 0.8,
|
| 2432 |
+
"eval_biology_entropy": 1.2935789892077445,
|
| 2433 |
+
"eval_biology_loss": 1.3476186990737915,
|
| 2434 |
+
"eval_biology_mean_token_accuracy": 0.672012362420559,
|
| 2435 |
+
"eval_biology_num_tokens": 14230754.0,
|
| 2436 |
+
"eval_biology_runtime": 44.053,
|
| 2437 |
+
"eval_biology_samples_per_second": 11.35,
|
| 2438 |
+
"eval_biology_steps_per_second": 11.35,
|
| 2439 |
+
"step": 2000
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.8,
|
| 2443 |
+
"eval_math_entropy": 0.6968656893670558,
|
| 2444 |
+
"eval_math_loss": 0.7122905850410461,
|
| 2445 |
+
"eval_math_mean_token_accuracy": 0.8068100419044495,
|
| 2446 |
+
"eval_math_num_tokens": 14230754.0,
|
| 2447 |
+
"eval_math_runtime": 48.0745,
|
| 2448 |
+
"eval_math_samples_per_second": 10.401,
|
| 2449 |
+
"eval_math_steps_per_second": 10.401,
|
| 2450 |
+
"step": 2000
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"entropy": 0.6868740394711494,
|
| 2454 |
+
"epoch": 0.804,
|
| 2455 |
+
"grad_norm": 35.75,
|
| 2456 |
+
"learning_rate": 1.7757777777777777e-05,
|
| 2457 |
+
"loss": 11.1864,
|
| 2458 |
+
"mean_token_accuracy": 0.8112314622849226,
|
| 2459 |
+
"num_tokens": 14301887.0,
|
| 2460 |
+
"step": 2010
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"entropy": 0.6712423311546445,
|
| 2464 |
+
"epoch": 0.808,
|
| 2465 |
+
"grad_norm": 28.5,
|
| 2466 |
+
"learning_rate": 1.7735555555555558e-05,
|
| 2467 |
+
"loss": 10.6354,
|
| 2468 |
+
"mean_token_accuracy": 0.8163831599056721,
|
| 2469 |
+
"num_tokens": 14374548.0,
|
| 2470 |
+
"step": 2020
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"entropy": 0.7165202729403972,
|
| 2474 |
+
"epoch": 0.812,
|
| 2475 |
+
"grad_norm": 31.625,
|
| 2476 |
+
"learning_rate": 1.7713333333333335e-05,
|
| 2477 |
+
"loss": 11.6733,
|
| 2478 |
+
"mean_token_accuracy": 0.8052945740520954,
|
| 2479 |
+
"num_tokens": 14451445.0,
|
| 2480 |
+
"step": 2030
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"entropy": 0.663957865908742,
|
| 2484 |
+
"epoch": 0.816,
|
| 2485 |
+
"grad_norm": 28.625,
|
| 2486 |
+
"learning_rate": 1.7691111111111113e-05,
|
| 2487 |
+
"loss": 10.5899,
|
| 2488 |
+
"mean_token_accuracy": 0.8158421032130718,
|
| 2489 |
+
"num_tokens": 14519893.0,
|
| 2490 |
+
"step": 2040
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"entropy": 0.7052116150036454,
|
| 2494 |
+
"epoch": 0.82,
|
| 2495 |
+
"grad_norm": 36.5,
|
| 2496 |
+
"learning_rate": 1.766888888888889e-05,
|
| 2497 |
+
"loss": 11.6679,
|
| 2498 |
+
"mean_token_accuracy": 0.8059296425431967,
|
| 2499 |
+
"num_tokens": 14594443.0,
|
| 2500 |
+
"step": 2050
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"entropy": 0.675534434337169,
|
| 2504 |
+
"epoch": 0.824,
|
| 2505 |
+
"grad_norm": 39.5,
|
| 2506 |
+
"learning_rate": 1.7646666666666668e-05,
|
| 2507 |
+
"loss": 10.7127,
|
| 2508 |
+
"mean_token_accuracy": 0.8169379521161317,
|
| 2509 |
+
"num_tokens": 14664170.0,
|
| 2510 |
+
"step": 2060
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"entropy": 0.6607738406397402,
|
| 2514 |
+
"epoch": 0.828,
|
| 2515 |
+
"grad_norm": 30.375,
|
| 2516 |
+
"learning_rate": 1.7624444444444446e-05,
|
| 2517 |
+
"loss": 10.8303,
|
| 2518 |
+
"mean_token_accuracy": 0.8149679381400347,
|
| 2519 |
+
"num_tokens": 14735227.0,
|
| 2520 |
+
"step": 2070
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"entropy": 0.686363017372787,
|
| 2524 |
+
"epoch": 0.832,
|
| 2525 |
+
"grad_norm": 28.875,
|
| 2526 |
+
"learning_rate": 1.7602222222222223e-05,
|
| 2527 |
+
"loss": 11.0101,
|
| 2528 |
+
"mean_token_accuracy": 0.8137267474085093,
|
| 2529 |
+
"num_tokens": 14805088.0,
|
| 2530 |
+
"step": 2080
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"entropy": 0.6817418249323964,
|
| 2534 |
+
"epoch": 0.836,
|
| 2535 |
+
"grad_norm": 31.125,
|
| 2536 |
+
"learning_rate": 1.758e-05,
|
| 2537 |
+
"loss": 11.1016,
|
| 2538 |
+
"mean_token_accuracy": 0.811983397975564,
|
| 2539 |
+
"num_tokens": 14875740.0,
|
| 2540 |
+
"step": 2090
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"entropy": 0.698756986297667,
|
| 2544 |
+
"epoch": 0.84,
|
| 2545 |
+
"grad_norm": 46.25,
|
| 2546 |
+
"learning_rate": 1.755777777777778e-05,
|
| 2547 |
+
"loss": 11.3359,
|
| 2548 |
+
"mean_token_accuracy": 0.8067192178219557,
|
| 2549 |
+
"num_tokens": 14944711.0,
|
| 2550 |
+
"step": 2100
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.84,
|
| 2554 |
+
"eval_biology_entropy": 1.268996791422367,
|
| 2555 |
+
"eval_biology_loss": 1.3460168838500977,
|
| 2556 |
+
"eval_biology_mean_token_accuracy": 0.673141638636589,
|
| 2557 |
+
"eval_biology_num_tokens": 14944711.0,
|
| 2558 |
+
"eval_biology_runtime": 41.4449,
|
| 2559 |
+
"eval_biology_samples_per_second": 12.064,
|
| 2560 |
+
"eval_biology_steps_per_second": 12.064,
|
| 2561 |
+
"step": 2100
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 0.84,
|
| 2565 |
+
"eval_math_entropy": 0.6983661242723465,
|
| 2566 |
+
"eval_math_loss": 0.7085059881210327,
|
| 2567 |
+
"eval_math_mean_token_accuracy": 0.8069341617822647,
|
| 2568 |
+
"eval_math_num_tokens": 14944711.0,
|
| 2569 |
+
"eval_math_runtime": 48.3919,
|
| 2570 |
+
"eval_math_samples_per_second": 10.332,
|
| 2571 |
+
"eval_math_steps_per_second": 10.332,
|
| 2572 |
+
"step": 2100
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"entropy": 0.6973433203995227,
|
| 2576 |
+
"epoch": 0.844,
|
| 2577 |
+
"grad_norm": 26.5,
|
| 2578 |
+
"learning_rate": 1.7535555555555556e-05,
|
| 2579 |
+
"loss": 11.1804,
|
| 2580 |
+
"mean_token_accuracy": 0.8072816133499146,
|
| 2581 |
+
"num_tokens": 15014711.0,
|
| 2582 |
+
"step": 2110
|
| 2583 |
+
},
|
| 2584 |
+
{
|
| 2585 |
+
"entropy": 0.6738097196444869,
|
| 2586 |
+
"epoch": 0.848,
|
| 2587 |
+
"grad_norm": 31.125,
|
| 2588 |
+
"learning_rate": 1.7513333333333334e-05,
|
| 2589 |
+
"loss": 10.8837,
|
| 2590 |
+
"mean_token_accuracy": 0.8131649781018495,
|
| 2591 |
+
"num_tokens": 15086306.0,
|
| 2592 |
+
"step": 2120
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"entropy": 0.7231233287602663,
|
| 2596 |
+
"epoch": 0.852,
|
| 2597 |
+
"grad_norm": 34.75,
|
| 2598 |
+
"learning_rate": 1.749111111111111e-05,
|
| 2599 |
+
"loss": 11.505,
|
| 2600 |
+
"mean_token_accuracy": 0.8036668822169304,
|
| 2601 |
+
"num_tokens": 15155311.0,
|
| 2602 |
+
"step": 2130
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"entropy": 0.7016849497333169,
|
| 2606 |
+
"epoch": 0.856,
|
| 2607 |
+
"grad_norm": 35.5,
|
| 2608 |
+
"learning_rate": 1.746888888888889e-05,
|
| 2609 |
+
"loss": 11.5304,
|
| 2610 |
+
"mean_token_accuracy": 0.8056407071650028,
|
| 2611 |
+
"num_tokens": 15228872.0,
|
| 2612 |
+
"step": 2140
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"entropy": 0.6963578762486577,
|
| 2616 |
+
"epoch": 0.86,
|
| 2617 |
+
"grad_norm": 34.0,
|
| 2618 |
+
"learning_rate": 1.7446666666666667e-05,
|
| 2619 |
+
"loss": 11.1786,
|
| 2620 |
+
"mean_token_accuracy": 0.8113155175000429,
|
| 2621 |
+
"num_tokens": 15301140.0,
|
| 2622 |
+
"step": 2150
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"entropy": 0.6930005135014653,
|
| 2626 |
+
"epoch": 0.864,
|
| 2627 |
+
"grad_norm": 29.25,
|
| 2628 |
+
"learning_rate": 1.7424444444444444e-05,
|
| 2629 |
+
"loss": 11.0449,
|
| 2630 |
+
"mean_token_accuracy": 0.810348815843463,
|
| 2631 |
+
"num_tokens": 15370985.0,
|
| 2632 |
+
"step": 2160
|
| 2633 |
+
},
|
| 2634 |
+
{
|
| 2635 |
+
"entropy": 0.6782323656603694,
|
| 2636 |
+
"epoch": 0.868,
|
| 2637 |
+
"grad_norm": 33.75,
|
| 2638 |
+
"learning_rate": 1.7402222222222222e-05,
|
| 2639 |
+
"loss": 11.0558,
|
| 2640 |
+
"mean_token_accuracy": 0.8125830505043268,
|
| 2641 |
+
"num_tokens": 15440581.0,
|
| 2642 |
+
"step": 2170
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"entropy": 0.681379459425807,
|
| 2646 |
+
"epoch": 0.872,
|
| 2647 |
+
"grad_norm": 32.5,
|
| 2648 |
+
"learning_rate": 1.7380000000000003e-05,
|
| 2649 |
+
"loss": 11.028,
|
| 2650 |
+
"mean_token_accuracy": 0.8101276684552431,
|
| 2651 |
+
"num_tokens": 15513360.0,
|
| 2652 |
+
"step": 2180
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"entropy": 0.7348810566589237,
|
| 2656 |
+
"epoch": 0.876,
|
| 2657 |
+
"grad_norm": 40.75,
|
| 2658 |
+
"learning_rate": 1.735777777777778e-05,
|
| 2659 |
+
"loss": 11.6674,
|
| 2660 |
+
"mean_token_accuracy": 0.8031103067100048,
|
| 2661 |
+
"num_tokens": 15585619.0,
|
| 2662 |
+
"step": 2190
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"entropy": 0.7099354591220617,
|
| 2666 |
+
"epoch": 0.88,
|
| 2667 |
+
"grad_norm": 36.25,
|
| 2668 |
+
"learning_rate": 1.7335555555555558e-05,
|
| 2669 |
+
"loss": 11.585,
|
| 2670 |
+
"mean_token_accuracy": 0.8052990060299635,
|
| 2671 |
+
"num_tokens": 15653836.0,
|
| 2672 |
+
"step": 2200
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 0.88,
|
| 2676 |
+
"eval_biology_entropy": 1.3010711609721184,
|
| 2677 |
+
"eval_biology_loss": 1.3462131023406982,
|
| 2678 |
+
"eval_biology_mean_token_accuracy": 0.6750594420433045,
|
| 2679 |
+
"eval_biology_num_tokens": 15653836.0,
|
| 2680 |
+
"eval_biology_runtime": 42.0081,
|
| 2681 |
+
"eval_biology_samples_per_second": 11.902,
|
| 2682 |
+
"eval_biology_steps_per_second": 11.902,
|
| 2683 |
+
"step": 2200
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.88,
|
| 2687 |
+
"eval_math_entropy": 0.7088503445386887,
|
| 2688 |
+
"eval_math_loss": 0.7035099864006042,
|
| 2689 |
+
"eval_math_mean_token_accuracy": 0.8091315562725068,
|
| 2690 |
+
"eval_math_num_tokens": 15653836.0,
|
| 2691 |
+
"eval_math_runtime": 52.4596,
|
| 2692 |
+
"eval_math_samples_per_second": 9.531,
|
| 2693 |
+
"eval_math_steps_per_second": 9.531,
|
| 2694 |
+
"step": 2200
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"entropy": 0.6961297648027539,
|
| 2698 |
+
"epoch": 0.884,
|
| 2699 |
+
"grad_norm": 38.5,
|
| 2700 |
+
"learning_rate": 1.7313333333333336e-05,
|
| 2701 |
+
"loss": 11.055,
|
| 2702 |
+
"mean_token_accuracy": 0.811257503926754,
|
| 2703 |
+
"num_tokens": 15721058.0,
|
| 2704 |
+
"step": 2210
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"entropy": 0.6642772014252841,
|
| 2708 |
+
"epoch": 0.888,
|
| 2709 |
+
"grad_norm": 28.5,
|
| 2710 |
+
"learning_rate": 1.7291111111111113e-05,
|
| 2711 |
+
"loss": 10.7968,
|
| 2712 |
+
"mean_token_accuracy": 0.8171007815748453,
|
| 2713 |
+
"num_tokens": 15791102.0,
|
| 2714 |
+
"step": 2220
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"entropy": 0.7127458658069372,
|
| 2718 |
+
"epoch": 0.892,
|
| 2719 |
+
"grad_norm": 28.375,
|
| 2720 |
+
"learning_rate": 1.726888888888889e-05,
|
| 2721 |
+
"loss": 11.5136,
|
| 2722 |
+
"mean_token_accuracy": 0.803728886321187,
|
| 2723 |
+
"num_tokens": 15861638.0,
|
| 2724 |
+
"step": 2230
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"entropy": 0.6892058169469237,
|
| 2728 |
+
"epoch": 0.896,
|
| 2729 |
+
"grad_norm": 39.75,
|
| 2730 |
+
"learning_rate": 1.724666666666667e-05,
|
| 2731 |
+
"loss": 11.2148,
|
| 2732 |
+
"mean_token_accuracy": 0.8115568786859513,
|
| 2733 |
+
"num_tokens": 15932179.0,
|
| 2734 |
+
"step": 2240
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"entropy": 0.6717539247125387,
|
| 2738 |
+
"epoch": 0.9,
|
| 2739 |
+
"grad_norm": 32.75,
|
| 2740 |
+
"learning_rate": 1.7224444444444446e-05,
|
| 2741 |
+
"loss": 10.7604,
|
| 2742 |
+
"mean_token_accuracy": 0.8124691691249609,
|
| 2743 |
+
"num_tokens": 16005581.0,
|
| 2744 |
+
"step": 2250
|
| 2745 |
+
},
|
| 2746 |
+
{
|
| 2747 |
+
"entropy": 0.6672627380117774,
|
| 2748 |
+
"epoch": 0.904,
|
| 2749 |
+
"grad_norm": 28.75,
|
| 2750 |
+
"learning_rate": 1.7202222222222224e-05,
|
| 2751 |
+
"loss": 10.6646,
|
| 2752 |
+
"mean_token_accuracy": 0.8165866218507289,
|
| 2753 |
+
"num_tokens": 16074409.0,
|
| 2754 |
+
"step": 2260
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"entropy": 0.7063423937186599,
|
| 2758 |
+
"epoch": 0.908,
|
| 2759 |
+
"grad_norm": 31.5,
|
| 2760 |
+
"learning_rate": 1.718e-05,
|
| 2761 |
+
"loss": 11.431,
|
| 2762 |
+
"mean_token_accuracy": 0.8058984253555537,
|
| 2763 |
+
"num_tokens": 16147328.0,
|
| 2764 |
+
"step": 2270
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"entropy": 0.6730604227632284,
|
| 2768 |
+
"epoch": 0.912,
|
| 2769 |
+
"grad_norm": 32.75,
|
| 2770 |
+
"learning_rate": 1.715777777777778e-05,
|
| 2771 |
+
"loss": 10.9602,
|
| 2772 |
+
"mean_token_accuracy": 0.8139959491789341,
|
| 2773 |
+
"num_tokens": 16219842.0,
|
| 2774 |
+
"step": 2280
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"entropy": 0.6916197570040822,
|
| 2778 |
+
"epoch": 0.916,
|
| 2779 |
+
"grad_norm": 34.0,
|
| 2780 |
+
"learning_rate": 1.7135555555555557e-05,
|
| 2781 |
+
"loss": 11.2951,
|
| 2782 |
+
"mean_token_accuracy": 0.81037244386971,
|
| 2783 |
+
"num_tokens": 16287996.0,
|
| 2784 |
+
"step": 2290
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"entropy": 0.6452676621265709,
|
| 2788 |
+
"epoch": 0.92,
|
| 2789 |
+
"grad_norm": 32.75,
|
| 2790 |
+
"learning_rate": 1.7113333333333334e-05,
|
| 2791 |
+
"loss": 10.48,
|
| 2792 |
+
"mean_token_accuracy": 0.8181278888136149,
|
| 2793 |
+
"num_tokens": 16359249.0,
|
| 2794 |
+
"step": 2300
|
| 2795 |
+
},
|
| 2796 |
+
{
|
| 2797 |
+
"epoch": 0.92,
|
| 2798 |
+
"eval_biology_entropy": 1.3079236669540406,
|
| 2799 |
+
"eval_biology_loss": 1.3488103151321411,
|
| 2800 |
+
"eval_biology_mean_token_accuracy": 0.6741236344575882,
|
| 2801 |
+
"eval_biology_num_tokens": 16359249.0,
|
| 2802 |
+
"eval_biology_runtime": 41.0049,
|
| 2803 |
+
"eval_biology_samples_per_second": 12.194,
|
| 2804 |
+
"eval_biology_steps_per_second": 12.194,
|
| 2805 |
+
"step": 2300
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 0.92,
|
| 2809 |
+
"eval_math_entropy": 0.7039950062930584,
|
| 2810 |
+
"eval_math_loss": 0.7003913521766663,
|
| 2811 |
+
"eval_math_mean_token_accuracy": 0.8092928317785263,
|
| 2812 |
+
"eval_math_num_tokens": 16359249.0,
|
| 2813 |
+
"eval_math_runtime": 48.3023,
|
| 2814 |
+
"eval_math_samples_per_second": 10.351,
|
| 2815 |
+
"eval_math_steps_per_second": 10.351,
|
| 2816 |
+
"step": 2300
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"entropy": 0.6840571843087673,
|
| 2820 |
+
"epoch": 0.924,
|
| 2821 |
+
"grad_norm": 30.375,
|
| 2822 |
+
"learning_rate": 1.7091111111111112e-05,
|
| 2823 |
+
"loss": 11.0281,
|
| 2824 |
+
"mean_token_accuracy": 0.8124604672193527,
|
| 2825 |
+
"num_tokens": 16430183.0,
|
| 2826 |
+
"step": 2310
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"entropy": 0.6504545219242572,
|
| 2830 |
+
"epoch": 0.928,
|
| 2831 |
+
"grad_norm": 40.75,
|
| 2832 |
+
"learning_rate": 1.706888888888889e-05,
|
| 2833 |
+
"loss": 10.4019,
|
| 2834 |
+
"mean_token_accuracy": 0.8224281292408705,
|
| 2835 |
+
"num_tokens": 16499842.0,
|
| 2836 |
+
"step": 2320
|
| 2837 |
+
},
|
| 2838 |
+
{
|
| 2839 |
+
"entropy": 0.6886620251461864,
|
| 2840 |
+
"epoch": 0.932,
|
| 2841 |
+
"grad_norm": 35.5,
|
| 2842 |
+
"learning_rate": 1.704666666666667e-05,
|
| 2843 |
+
"loss": 11.2413,
|
| 2844 |
+
"mean_token_accuracy": 0.8109501354396343,
|
| 2845 |
+
"num_tokens": 16569436.0,
|
| 2846 |
+
"step": 2330
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"entropy": 0.6863971030339598,
|
| 2850 |
+
"epoch": 0.936,
|
| 2851 |
+
"grad_norm": 37.5,
|
| 2852 |
+
"learning_rate": 1.7024444444444445e-05,
|
| 2853 |
+
"loss": 10.976,
|
| 2854 |
+
"mean_token_accuracy": 0.812617215141654,
|
| 2855 |
+
"num_tokens": 16640499.0,
|
| 2856 |
+
"step": 2340
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"entropy": 0.6811121774837374,
|
| 2860 |
+
"epoch": 0.94,
|
| 2861 |
+
"grad_norm": 29.0,
|
| 2862 |
+
"learning_rate": 1.7002222222222226e-05,
|
| 2863 |
+
"loss": 10.9927,
|
| 2864 |
+
"mean_token_accuracy": 0.812858996540308,
|
| 2865 |
+
"num_tokens": 16708143.0,
|
| 2866 |
+
"step": 2350
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"entropy": 0.6965519505552947,
|
| 2870 |
+
"epoch": 0.944,
|
| 2871 |
+
"grad_norm": 29.625,
|
| 2872 |
+
"learning_rate": 1.698e-05,
|
| 2873 |
+
"loss": 11.2976,
|
| 2874 |
+
"mean_token_accuracy": 0.8083064600825309,
|
| 2875 |
+
"num_tokens": 16781882.0,
|
| 2876 |
+
"step": 2360
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"entropy": 0.6884551353752613,
|
| 2880 |
+
"epoch": 0.948,
|
| 2881 |
+
"grad_norm": 34.0,
|
| 2882 |
+
"learning_rate": 1.695777777777778e-05,
|
| 2883 |
+
"loss": 11.1246,
|
| 2884 |
+
"mean_token_accuracy": 0.8140901360660792,
|
| 2885 |
+
"num_tokens": 16855942.0,
|
| 2886 |
+
"step": 2370
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"entropy": 0.682703816704452,
|
| 2890 |
+
"epoch": 0.952,
|
| 2891 |
+
"grad_norm": 37.5,
|
| 2892 |
+
"learning_rate": 1.6935555555555555e-05,
|
| 2893 |
+
"loss": 10.9073,
|
| 2894 |
+
"mean_token_accuracy": 0.8152716889977455,
|
| 2895 |
+
"num_tokens": 16922913.0,
|
| 2896 |
+
"step": 2380
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"entropy": 0.7374881992116571,
|
| 2900 |
+
"epoch": 0.956,
|
| 2901 |
+
"grad_norm": 42.25,
|
| 2902 |
+
"learning_rate": 1.6913333333333336e-05,
|
| 2903 |
+
"loss": 11.9923,
|
| 2904 |
+
"mean_token_accuracy": 0.7996022880077363,
|
| 2905 |
+
"num_tokens": 16991940.0,
|
| 2906 |
+
"step": 2390
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"entropy": 0.679273992497474,
|
| 2910 |
+
"epoch": 0.96,
|
| 2911 |
+
"grad_norm": 26.75,
|
| 2912 |
+
"learning_rate": 1.689111111111111e-05,
|
| 2913 |
+
"loss": 10.8099,
|
| 2914 |
+
"mean_token_accuracy": 0.8152121867984533,
|
| 2915 |
+
"num_tokens": 17067407.0,
|
| 2916 |
+
"step": 2400
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.96,
|
| 2920 |
+
"eval_biology_entropy": 1.2861202019453049,
|
| 2921 |
+
"eval_biology_loss": 1.348455786705017,
|
| 2922 |
+
"eval_biology_mean_token_accuracy": 0.6744914083480835,
|
| 2923 |
+
"eval_biology_num_tokens": 17067407.0,
|
| 2924 |
+
"eval_biology_runtime": 40.276,
|
| 2925 |
+
"eval_biology_samples_per_second": 12.414,
|
| 2926 |
+
"eval_biology_steps_per_second": 12.414,
|
| 2927 |
+
"step": 2400
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 0.96,
|
| 2931 |
+
"eval_math_entropy": 0.6839256008565426,
|
| 2932 |
+
"eval_math_loss": 0.6964585185050964,
|
| 2933 |
+
"eval_math_mean_token_accuracy": 0.8100001140832901,
|
| 2934 |
+
"eval_math_num_tokens": 17067407.0,
|
| 2935 |
+
"eval_math_runtime": 48.0956,
|
| 2936 |
+
"eval_math_samples_per_second": 10.396,
|
| 2937 |
+
"eval_math_steps_per_second": 10.396,
|
| 2938 |
+
"step": 2400
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"entropy": 0.6621893520466984,
|
| 2942 |
+
"epoch": 0.964,
|
| 2943 |
+
"grad_norm": 31.625,
|
| 2944 |
+
"learning_rate": 1.686888888888889e-05,
|
| 2945 |
+
"loss": 10.733,
|
| 2946 |
+
"mean_token_accuracy": 0.8197611160576344,
|
| 2947 |
+
"num_tokens": 17135614.0,
|
| 2948 |
+
"step": 2410
|
| 2949 |
+
},
|
| 2950 |
+
{
|
| 2951 |
+
"entropy": 0.6837352143600584,
|
| 2952 |
+
"epoch": 0.968,
|
| 2953 |
+
"grad_norm": 32.0,
|
| 2954 |
+
"learning_rate": 1.684666666666667e-05,
|
| 2955 |
+
"loss": 10.9487,
|
| 2956 |
+
"mean_token_accuracy": 0.811904390156269,
|
| 2957 |
+
"num_tokens": 17211153.0,
|
| 2958 |
+
"step": 2420
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"entropy": 0.6816597860306501,
|
| 2962 |
+
"epoch": 0.972,
|
| 2963 |
+
"grad_norm": 33.5,
|
| 2964 |
+
"learning_rate": 1.6824444444444447e-05,
|
| 2965 |
+
"loss": 11.1352,
|
| 2966 |
+
"mean_token_accuracy": 0.8097947511821986,
|
| 2967 |
+
"num_tokens": 17280414.0,
|
| 2968 |
+
"step": 2430
|
| 2969 |
+
},
|
| 2970 |
+
{
|
| 2971 |
+
"entropy": 0.6748472723178566,
|
| 2972 |
+
"epoch": 0.976,
|
| 2973 |
+
"grad_norm": 41.5,
|
| 2974 |
+
"learning_rate": 1.6802222222222224e-05,
|
| 2975 |
+
"loss": 10.8497,
|
| 2976 |
+
"mean_token_accuracy": 0.8141761176288128,
|
| 2977 |
+
"num_tokens": 17350994.0,
|
| 2978 |
+
"step": 2440
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"entropy": 0.6920185225084424,
|
| 2982 |
+
"epoch": 0.98,
|
| 2983 |
+
"grad_norm": 37.75,
|
| 2984 |
+
"learning_rate": 1.6780000000000002e-05,
|
| 2985 |
+
"loss": 11.0203,
|
| 2986 |
+
"mean_token_accuracy": 0.8110847111791373,
|
| 2987 |
+
"num_tokens": 17422905.0,
|
| 2988 |
+
"step": 2450
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"entropy": 0.6656286349520087,
|
| 2992 |
+
"epoch": 0.984,
|
| 2993 |
+
"grad_norm": 32.0,
|
| 2994 |
+
"learning_rate": 1.675777777777778e-05,
|
| 2995 |
+
"loss": 10.8741,
|
| 2996 |
+
"mean_token_accuracy": 0.8118232063949108,
|
| 2997 |
+
"num_tokens": 17492960.0,
|
| 2998 |
+
"step": 2460
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"entropy": 0.7282194011844695,
|
| 3002 |
+
"epoch": 0.988,
|
| 3003 |
+
"grad_norm": 29.0,
|
| 3004 |
+
"learning_rate": 1.6735555555555557e-05,
|
| 3005 |
+
"loss": 11.4374,
|
| 3006 |
+
"mean_token_accuracy": 0.8076034974306822,
|
| 3007 |
+
"num_tokens": 17564626.0,
|
| 3008 |
+
"step": 2470
|
| 3009 |
+
},
|
| 3010 |
+
{
|
| 3011 |
+
"entropy": 0.656241605617106,
|
| 3012 |
+
"epoch": 0.992,
|
| 3013 |
+
"grad_norm": 37.25,
|
| 3014 |
+
"learning_rate": 1.6713333333333335e-05,
|
| 3015 |
+
"loss": 10.7226,
|
| 3016 |
+
"mean_token_accuracy": 0.814904909953475,
|
| 3017 |
+
"num_tokens": 17637514.0,
|
| 3018 |
+
"step": 2480
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"entropy": 0.7101638291031123,
|
| 3022 |
+
"epoch": 0.996,
|
| 3023 |
+
"grad_norm": 26.75,
|
| 3024 |
+
"learning_rate": 1.6691111111111112e-05,
|
| 3025 |
+
"loss": 11.4156,
|
| 3026 |
+
"mean_token_accuracy": 0.8066222406923771,
|
| 3027 |
+
"num_tokens": 17708953.0,
|
| 3028 |
+
"step": 2490
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"entropy": 0.6990812286734581,
|
| 3032 |
+
"epoch": 1.0,
|
| 3033 |
+
"grad_norm": 28.5,
|
| 3034 |
+
"learning_rate": 1.666888888888889e-05,
|
| 3035 |
+
"loss": 11.3146,
|
| 3036 |
+
"mean_token_accuracy": 0.8074303966015577,
|
| 3037 |
+
"num_tokens": 17780432.0,
|
| 3038 |
+
"step": 2500
|
| 3039 |
+
},
|
| 3040 |
+
{
|
| 3041 |
+
"epoch": 1.0,
|
| 3042 |
+
"eval_biology_entropy": 1.278903669476509,
|
| 3043 |
+
"eval_biology_loss": 1.3472909927368164,
|
| 3044 |
+
"eval_biology_mean_token_accuracy": 0.6741317790150643,
|
| 3045 |
+
"eval_biology_num_tokens": 17780432.0,
|
| 3046 |
+
"eval_biology_runtime": 41.8672,
|
| 3047 |
+
"eval_biology_samples_per_second": 11.943,
|
| 3048 |
+
"eval_biology_steps_per_second": 11.943,
|
| 3049 |
+
"step": 2500
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 1.0,
|
| 3053 |
+
"eval_math_entropy": 0.6809976233839988,
|
| 3054 |
+
"eval_math_loss": 0.6932579278945923,
|
| 3055 |
+
"eval_math_mean_token_accuracy": 0.8099925496578216,
|
| 3056 |
+
"eval_math_num_tokens": 17780432.0,
|
| 3057 |
+
"eval_math_runtime": 51.1798,
|
| 3058 |
+
"eval_math_samples_per_second": 9.769,
|
| 3059 |
+
"eval_math_steps_per_second": 9.769,
|
| 3060 |
+
"step": 2500
|
| 3061 |
+
}
|
| 3062 |
+
],
|
| 3063 |
+
"logging_steps": 10,
|
| 3064 |
+
"max_steps": 10000,
|
| 3065 |
+
"num_input_tokens_seen": 0,
|
| 3066 |
+
"num_train_epochs": 4,
|
| 3067 |
+
"save_steps": 500,
|
| 3068 |
+
"stateful_callbacks": {
|
| 3069 |
+
"TrainerControl": {
|
| 3070 |
+
"args": {
|
| 3071 |
+
"should_epoch_stop": false,
|
| 3072 |
+
"should_evaluate": false,
|
| 3073 |
+
"should_log": false,
|
| 3074 |
+
"should_save": true,
|
| 3075 |
+
"should_training_stop": false
|
| 3076 |
+
},
|
| 3077 |
+
"attributes": {}
|
| 3078 |
+
}
|
| 3079 |
+
},
|
| 3080 |
+
"total_flos": 1.1922556015093102e+18,
|
| 3081 |
+
"train_batch_size": 1,
|
| 3082 |
+
"trial_name": null,
|
| 3083 |
+
"trial_params": null
|
| 3084 |
+
}
|
checkpoint-2500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6df9046062202b486a4fcb0ad2a415dd6b0a1a521e0667c7f40799812ff8d06e
|
| 3 |
+
size 6481
|