Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+limo/full/checkpoint-515/tokenizer.json filter=lfs diff=lfs merge=lfs -text

limo/full/checkpoint-515/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

limo/full/checkpoint-515/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "/global_data/sft_intern/lh/huggingface_models/Qwen2.5-Math-7B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 4096,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

limo/full/checkpoint-515/generation_config.json

@@ -0,0 +1,9 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "transformers_version": "4.46.1"
+}

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9576622bd34465b97694265894b1ffda56e5af593de4fc5099f46bd929e6ade3
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0bb9655209fb11f164748ab89a941b07859803b765576aef8d5319f737e44a5
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fa338c33ba49ef763fde527342b61527f927065b9c41b7248c69c19bdd92238
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a21facf68e86bf29ccc6bb339ccd4ff71267c185248f3a2cd5229723aeba86eb
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfc445133211ef3d274babc664a39da3b0d1bf7b73342a260c2321c669919ef5
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa83611fa9807ef42900792d91525fc0909e1f37f49f7b45b833dc00700d26a6
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:029c5ed7c4807feb07e54c3753a3aa988ec4cdad8ba7d7e29c518e85c1dca15f
+size 11423430028

limo/full/checkpoint-515/global_step515/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8cc2914f5415a2ef31bdad188320c0c662865f2330a5caa2a0e31cda221e3a4
+size 11423430028

limo/full/checkpoint-515/global_step515/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ad46143ce4f22974afb3bee75c8d2dfd7d5797f92c55ebbd859b8b46a44f4c0
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a2a814ad293c753ea4acdd5f4218d64659212015e523cfe2058fd338f3afa99
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b295f036d44507abf42b8f7ecdfb5fba8ff0f51ab26070f8ca6237a86b9eef7f
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa969ef53ba8f160ae270a880b664fb084876f256fcc5bd1ae33b5b9ad0575fc
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:885b2a649d770130c3c983f7d4be2e018f8ddd7a8f636fc724c4be886beab224
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86138c77f1bc5496ca55be84380ad3cfadd9726e28c90d8de757700221b81433
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de06ee7bf066cbd7ba709eddd2b9daa58d1ba642ffb81f83bf949246086c4c46
+size 168277

limo/full/checkpoint-515/global_step515/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea1551314ab406fac9f39e77433e4c8be86a63a8807dc081f1b0954e540b57e2
+size 168277

limo/full/checkpoint-515/latest

@@ -0,0 +1 @@
+global_step515

limo/full/checkpoint-515/model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f469d5b29b55d4cf34a64764b9cf40644492479b5e48db6e4c83437072bb6c74
+size 4877660776

limo/full/checkpoint-515/model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8488e427d007b32a9c13318a96eee2a35f750ee0109364f920ac679c44724ad8
+size 4932751008

limo/full/checkpoint-515/model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d8cc286387dd53f5eba5a1d3e3222275cbb64759e3b2fc91643dcdaf8564f96
+size 4330865200

limo/full/checkpoint-515/model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8c051f4d079a1a347edaf21b2daa8506b3b5455768b1881cab68d37d6308bb8
+size 1089994880

limo/full/checkpoint-515/model.safetensors.index.json

@@ -0,0 +1,346 @@
+{
+  "metadata": {
+    "total_size": 15231233024
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.norm.weight": "model-00003-of-00004.safetensors"
+  }
+}

limo/full/checkpoint-515/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a0ef6f96a48e59aa52c4b471312c2a62378c19acc7ebbae839612b03a7d775a
+size 15984

limo/full/checkpoint-515/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab11d533c0fdad46ea8b8e295ba5fdb705e078eeb88cc28f37d82913508766e9
+size 15984

limo/full/checkpoint-515/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:615c168147e3465ce5bfab6da2ff4afc68566ce00ec0f0c6c9fc988038a58d0a
+size 15984

limo/full/checkpoint-515/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79f71e8f8674ecaef9f8cdcbf7ac457a8b8ff15b12694ba2a2fffcb4b43f0f08
+size 15984

limo/full/checkpoint-515/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88cf6d674dab5545c300a55135f08ca935730a3d35e2c419fb0b333f19482c19
+size 15984

limo/full/checkpoint-515/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2754f2cd8824702f027870d93748b3c0491b0ecd30f1e3d8e937116b2be6151f
+size 15984

limo/full/checkpoint-515/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1385124ac55604598f45ea6e2d141f29456647d3e7c10d12ca64ec93d312be8d
+size 15984

limo/full/checkpoint-515/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:416538efaec7391fa8fe782fb15146b83e5612d9e1961292c34c53e964806873
+size 15984

limo/full/checkpoint-515/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5d0919066ee6a14427d7be219c870467c623faeab11aea9e226fd92a7435d45
+size 1064

limo/full/checkpoint-515/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

limo/full/checkpoint-515/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

limo/full/checkpoint-515/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'Please reason step by step, and put your final answer within \\\\boxed{}.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nPlease reason step by step, and put your final answer within \\\\boxed{}.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 16384,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

limo/full/checkpoint-515/trainer_state.json

@@ -0,0 +1,3638 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 515,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.009708737864077669,
+      "grad_norm": 32.0991353240983,
+      "learning_rate": 4.999994831641374e-06,
+      "loss": 4.1052,
+      "step": 1
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 29.524023707455505,
+      "learning_rate": 4.9999793265868636e-06,
+      "loss": 3.2535,
+      "step": 2
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 25.915328826410054,
+      "learning_rate": 4.999953484900578e-06,
+      "loss": 3.0603,
+      "step": 3
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 26.925318996290773,
+      "learning_rate": 4.9999173066893655e-06,
+      "loss": 3.8736,
+      "step": 4
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 17.49179072124481,
+      "learning_rate": 4.9998707921028104e-06,
+      "loss": 3.2503,
+      "step": 5
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 11.174584909912863,
+      "learning_rate": 4.999813941333237e-06,
+      "loss": 2.0295,
+      "step": 6
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 9.90835381349434,
+      "learning_rate": 4.999746754615704e-06,
+      "loss": 1.9601,
+      "step": 7
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 2.252755813363907,
+      "learning_rate": 4.9996692322280085e-06,
+      "loss": 1.1375,
+      "step": 8
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 9.417040995705188,
+      "learning_rate": 4.999581374490681e-06,
+      "loss": 2.4889,
+      "step": 9
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 5.128297805941186,
+      "learning_rate": 4.999483181766986e-06,
+      "loss": 1.9038,
+      "step": 10
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 2.6424619563549627,
+      "learning_rate": 4.999374654462919e-06,
+      "loss": 1.3898,
+      "step": 11
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 5.089573735718794,
+      "learning_rate": 4.999255793027207e-06,
+      "loss": 2.1286,
+      "step": 12
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 2.251991573691883,
+      "learning_rate": 4.999126597951305e-06,
+      "loss": 1.0427,
+      "step": 13
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 2.2920823891389333,
+      "learning_rate": 4.998987069769394e-06,
+      "loss": 1.1071,
+      "step": 14
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 2.8377769456674247,
+      "learning_rate": 4.998837209058379e-06,
+      "loss": 1.6229,
+      "step": 15
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 7.727434690224115,
+      "learning_rate": 4.998677016437888e-06,
+      "loss": 1.3171,
+      "step": 16
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 3.74374589959768,
+      "learning_rate": 4.998506492570266e-06,
+      "loss": 1.7463,
+      "step": 17
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 3.837005912107839,
+      "learning_rate": 4.998325638160576e-06,
+      "loss": 1.3993,
+      "step": 18
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 5.797983985086783,
+      "learning_rate": 4.998134453956596e-06,
+      "loss": 1.1204,
+      "step": 19
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 2.7548345494912807,
+      "learning_rate": 4.997932940748811e-06,
+      "loss": 1.2081,
+      "step": 20
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 5.045371621372198,
+      "learning_rate": 4.997721099370416e-06,
+      "loss": 1.2124,
+      "step": 21
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 2.051110624952864,
+      "learning_rate": 4.997498930697308e-06,
+      "loss": 1.3137,
+      "step": 22
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 2.691454094294257,
+      "learning_rate": 4.997266435648086e-06,
+      "loss": 1.3098,
+      "step": 23
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 2.4865134695003017,
+      "learning_rate": 4.997023615184044e-06,
+      "loss": 1.6151,
+      "step": 24
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 2.8075788095017757,
+      "learning_rate": 4.996770470309167e-06,
+      "loss": 1.6531,
+      "step": 25
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 1.768040918463024,
+      "learning_rate": 4.996507002070131e-06,
+      "loss": 1.1902,
+      "step": 26
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 1.955363000769332,
+      "learning_rate": 4.996233211556295e-06,
+      "loss": 1.2454,
+      "step": 27
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 4.241739298469607,
+      "learning_rate": 4.9959490998996974e-06,
+      "loss": 0.8855,
+      "step": 28
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 7.222330025323884,
+      "learning_rate": 4.995654668275049e-06,
+      "loss": 1.2271,
+      "step": 29
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 2.3651169987619527,
+      "learning_rate": 4.995349917899735e-06,
+      "loss": 1.3329,
+      "step": 30
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 1.6948963634507868,
+      "learning_rate": 4.9950348500338005e-06,
+      "loss": 0.9562,
+      "step": 31
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.3136462562718183,
+      "learning_rate": 4.994709465979954e-06,
+      "loss": 0.9246,
+      "step": 32
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 3.6835006171200306,
+      "learning_rate": 4.994373767083556e-06,
+      "loss": 1.0724,
+      "step": 33
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 3.163430858725886,
+      "learning_rate": 4.994027754732616e-06,
+      "loss": 1.1139,
+      "step": 34
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 1.4746343819625023,
+      "learning_rate": 4.993671430357788e-06,
+      "loss": 0.9994,
+      "step": 35
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.9654224512814882,
+      "learning_rate": 4.99330479543236e-06,
+      "loss": 1.3667,
+      "step": 36
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 1.2495201306162087,
+      "learning_rate": 4.992927851472254e-06,
+      "loss": 1.1252,
+      "step": 37
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 1.5824371832462278,
+      "learning_rate": 4.992540600036014e-06,
+      "loss": 1.3831,
+      "step": 38
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 1.4702657605395504,
+      "learning_rate": 4.992143042724805e-06,
+      "loss": 1.1461,
+      "step": 39
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4506087672413808,
+      "learning_rate": 4.991735181182401e-06,
+      "loss": 1.4195,
+      "step": 40
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 1.075865370590295,
+      "learning_rate": 4.991317017095182e-06,
+      "loss": 0.9813,
+      "step": 41
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 1.2265855895293323,
+      "learning_rate": 4.990888552192126e-06,
+      "loss": 1.0049,
+      "step": 42
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 2.4151552794703437,
+      "learning_rate": 4.9904497882448004e-06,
+      "loss": 1.1099,
+      "step": 43
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6677746145821908,
+      "learning_rate": 4.990000727067357e-06,
+      "loss": 1.1481,
+      "step": 44
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 1.5396824114800838,
+      "learning_rate": 4.989541370516523e-06,
+      "loss": 0.9952,
+      "step": 45
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 1.3928969634273873,
+      "learning_rate": 4.989071720491595e-06,
+      "loss": 0.9147,
+      "step": 46
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.9492551299059091,
+      "learning_rate": 4.988591778934428e-06,
+      "loss": 0.6869,
+      "step": 47
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.2714456311245477,
+      "learning_rate": 4.9881015478294294e-06,
+      "loss": 0.8276,
+      "step": 48
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 2.6314825001613658,
+      "learning_rate": 4.987601029203553e-06,
+      "loss": 1.0247,
+      "step": 49
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 1.416971495316372,
+      "learning_rate": 4.987090225126285e-06,
+      "loss": 1.012,
+      "step": 50
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 1.361560829125586,
+      "learning_rate": 4.98656913770964e-06,
+      "loss": 0.8907,
+      "step": 51
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.3035984152044737,
+      "learning_rate": 4.986037769108154e-06,
+      "loss": 1.056,
+      "step": 52
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 1.0150205143363307,
+      "learning_rate": 4.9854961215188676e-06,
+      "loss": 1.0017,
+      "step": 53
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.9609277033036215,
+      "learning_rate": 4.984944197181324e-06,
+      "loss": 0.7601,
+      "step": 54
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 1.1898667428428673,
+      "learning_rate": 4.9843819983775575e-06,
+      "loss": 0.8858,
+      "step": 55
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.0139938825198491,
+      "learning_rate": 4.983809527432086e-06,
+      "loss": 0.8071,
+      "step": 56
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 1.247418732762796,
+      "learning_rate": 4.983226786711895e-06,
+      "loss": 0.9675,
+      "step": 57
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.8942637413037233,
+      "learning_rate": 4.982633778626437e-06,
+      "loss": 0.8187,
+      "step": 58
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 1.3943617986028647,
+      "learning_rate": 4.982030505627613e-06,
+      "loss": 1.0678,
+      "step": 59
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.8610131821728051,
+      "learning_rate": 4.98141697020977e-06,
+      "loss": 0.8306,
+      "step": 60
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 1.0768921632881472,
+      "learning_rate": 4.9807931749096836e-06,
+      "loss": 0.81,
+      "step": 61
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.9625381291373968,
+      "learning_rate": 4.980159122306551e-06,
+      "loss": 0.892,
+      "step": 62
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.9617828022691955,
+      "learning_rate": 4.979514815021984e-06,
+      "loss": 1.0243,
+      "step": 63
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.2606711145146943,
+      "learning_rate": 4.978860255719989e-06,
+      "loss": 0.7773,
+      "step": 64
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.8531956488376088,
+      "learning_rate": 4.978195447106965e-06,
+      "loss": 0.9458,
+      "step": 65
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.8576435230419439,
+      "learning_rate": 4.9775203919316864e-06,
+      "loss": 0.7812,
+      "step": 66
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 0.9030199864232913,
+      "learning_rate": 4.976835092985297e-06,
+      "loss": 0.9382,
+      "step": 67
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.8483972987027173,
+      "learning_rate": 4.976139553101291e-06,
+      "loss": 0.8671,
+      "step": 68
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.8960545150162272,
+      "learning_rate": 4.975433775155509e-06,
+      "loss": 0.8646,
+      "step": 69
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 1.0017538898162217,
+      "learning_rate": 4.974717762066123e-06,
+      "loss": 0.8805,
+      "step": 70
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 0.9212309895628247,
+      "learning_rate": 4.973991516793621e-06,
+      "loss": 1.0576,
+      "step": 71
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.9305642899936102,
+      "learning_rate": 4.973255042340801e-06,
+      "loss": 0.8486,
+      "step": 72
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 1.0773851273416222,
+      "learning_rate": 4.972508341752754e-06,
+      "loss": 1.0583,
+      "step": 73
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.892091799769842,
+      "learning_rate": 4.9717514181168534e-06,
+      "loss": 0.7527,
+      "step": 74
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.8481741026230679,
+      "learning_rate": 4.970984274562741e-06,
+      "loss": 0.7125,
+      "step": 75
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.9312066959231862,
+      "learning_rate": 4.970206914262315e-06,
+      "loss": 0.8687,
+      "step": 76
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.9215979104056321,
+      "learning_rate": 4.969419340429717e-06,
+      "loss": 0.7691,
+      "step": 77
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 1.227175476240732,
+      "learning_rate": 4.968621556321319e-06,
+      "loss": 0.92,
+      "step": 78
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 0.9330601553607981,
+      "learning_rate": 4.967813565235708e-06,
+      "loss": 0.8216,
+      "step": 79
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.8252798793274818,
+      "learning_rate": 4.966995370513675e-06,
+      "loss": 0.7061,
+      "step": 80
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.9922301648204198,
+      "learning_rate": 4.966166975538197e-06,
+      "loss": 1.0408,
+      "step": 81
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.7891923797199208,
+      "learning_rate": 4.965328383734429e-06,
+      "loss": 0.7595,
+      "step": 82
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 1.0034345889698408,
+      "learning_rate": 4.964479598569686e-06,
+      "loss": 1.0233,
+      "step": 83
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.9634692778615994,
+      "learning_rate": 4.963620623553428e-06,
+      "loss": 0.929,
+      "step": 84
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.8794808183269647,
+      "learning_rate": 4.962751462237248e-06,
+      "loss": 0.7247,
+      "step": 85
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.8991558039524955,
+      "learning_rate": 4.9618721182148564e-06,
+      "loss": 0.6248,
+      "step": 86
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.9291777621093169,
+      "learning_rate": 4.960982595122064e-06,
+      "loss": 0.7035,
+      "step": 87
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.85791569259972,
+      "learning_rate": 4.960082896636773e-06,
+      "loss": 0.9148,
+      "step": 88
+    },
+    {
+      "epoch": 0.8640776699029126,
+      "grad_norm": 0.9063445307146435,
+      "learning_rate": 4.959173026478952e-06,
+      "loss": 0.7805,
+      "step": 89
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.7587205124423355,
+      "learning_rate": 4.958252988410631e-06,
+      "loss": 0.6329,
+      "step": 90
+    },
+    {
+      "epoch": 0.883495145631068,
+      "grad_norm": 1.0713074090013364,
+      "learning_rate": 4.9573227862358794e-06,
+      "loss": 0.7955,
+      "step": 91
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.8477622625159322,
+      "learning_rate": 4.956382423800791e-06,
+      "loss": 0.8325,
+      "step": 92
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.9269059925838263,
+      "learning_rate": 4.955431904993471e-06,
+      "loss": 0.8194,
+      "step": 93
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 1.0475028792580197,
+      "learning_rate": 4.954471233744015e-06,
+      "loss": 0.5835,
+      "step": 94
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.8985260034505762,
+      "learning_rate": 4.9535004140245005e-06,
+      "loss": 0.8063,
+      "step": 95
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.0048572531965805,
+      "learning_rate": 4.952519449848962e-06,
+      "loss": 0.8127,
+      "step": 96
+    },
+    {
+      "epoch": 0.941747572815534,
+      "grad_norm": 0.9062750156316196,
+      "learning_rate": 4.951528345273379e-06,
+      "loss": 0.7181,
+      "step": 97
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.8778949835317967,
+      "learning_rate": 4.950527104395659e-06,
+      "loss": 0.8103,
+      "step": 98
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 0.9728187484090823,
+      "learning_rate": 4.9495157313556185e-06,
+      "loss": 0.6329,
+      "step": 99
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.7352370132544686,
+      "learning_rate": 4.94849423033497e-06,
+      "loss": 0.5231,
+      "step": 100
+    },
+    {
+      "epoch": 0.9805825242718447,
+      "grad_norm": 0.8479762742075609,
+      "learning_rate": 4.9474626055573e-06,
+      "loss": 0.9551,
+      "step": 101
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.7351142523348447,
+      "learning_rate": 4.946420861288051e-06,
+      "loss": 0.738,
+      "step": 102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9009004817884025,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.7789,
+      "step": 103
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 0.8057060436062744,
+      "learning_rate": 4.944307031545797e-06,
+      "loss": 0.6629,
+      "step": 104
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.8977736266341005,
+      "learning_rate": 4.943234954812812e-06,
+      "loss": 0.9053,
+      "step": 105
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 0.7717440949849311,
+      "learning_rate": 4.942152776068264e-06,
+      "loss": 0.6404,
+      "step": 106
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 0.890503065931168,
+      "learning_rate": 4.941060499786622e-06,
+      "loss": 0.9117,
+      "step": 107
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 0.8814259279520221,
+      "learning_rate": 4.939958130484106e-06,
+      "loss": 0.7055,
+      "step": 108
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.671081647050562,
+      "learning_rate": 4.938845672718668e-06,
+      "loss": 0.6671,
+      "step": 109
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.8278909841687843,
+      "learning_rate": 4.937723131089974e-06,
+      "loss": 0.7318,
+      "step": 110
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 0.9269878913315724,
+      "learning_rate": 4.93659051023938e-06,
+      "loss": 0.7646,
+      "step": 111
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.665235455755867,
+      "learning_rate": 4.93544781484992e-06,
+      "loss": 0.4998,
+      "step": 112
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.8487063874467249,
+      "learning_rate": 4.9342950496462815e-06,
+      "loss": 0.7435,
+      "step": 113
+    },
+    {
+      "epoch": 1.1067961165048543,
+      "grad_norm": 0.8430924624096062,
+      "learning_rate": 4.933132219394786e-06,
+      "loss": 0.6992,
+      "step": 114
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.8993084000328078,
+      "learning_rate": 4.931959328903376e-06,
+      "loss": 0.8961,
+      "step": 115
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 0.7373642852662377,
+      "learning_rate": 4.930776383021584e-06,
+      "loss": 0.6722,
+      "step": 116
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 0.8356211654383496,
+      "learning_rate": 4.92958338664052e-06,
+      "loss": 0.528,
+      "step": 117
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.8788366905889159,
+      "learning_rate": 4.928380344692853e-06,
+      "loss": 0.6369,
+      "step": 118
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.70400203786099,
+      "learning_rate": 4.927167262152784e-06,
+      "loss": 0.6961,
+      "step": 119
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.7933675823948615,
+      "learning_rate": 4.925944144036027e-06,
+      "loss": 0.7316,
+      "step": 120
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.7413576987999055,
+      "learning_rate": 4.924710995399796e-06,
+      "loss": 0.6764,
+      "step": 121
+    },
+    {
+      "epoch": 1.1844660194174756,
+      "grad_norm": 0.8385780569550869,
+      "learning_rate": 4.923467821342773e-06,
+      "loss": 0.7602,
+      "step": 122
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.8531356062842188,
+      "learning_rate": 4.922214627005092e-06,
+      "loss": 0.749,
+      "step": 123
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.8216291410900652,
+      "learning_rate": 4.920951417568323e-06,
+      "loss": 0.8079,
+      "step": 124
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.8320005263963047,
+      "learning_rate": 4.919678198255438e-06,
+      "loss": 0.6805,
+      "step": 125
+    },
+    {
+      "epoch": 1.2233009708737863,
+      "grad_norm": 0.8268979479586152,
+      "learning_rate": 4.918394974330801e-06,
+      "loss": 0.7583,
+      "step": 126
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.7645562011859167,
+      "learning_rate": 4.917101751100142e-06,
+      "loss": 0.7109,
+      "step": 127
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.8421630378850258,
+      "learning_rate": 4.915798533910534e-06,
+      "loss": 0.714,
+      "step": 128
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.7552330522382976,
+      "learning_rate": 4.9144853281503715e-06,
+      "loss": 0.5679,
+      "step": 129
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.852273134838057,
+      "learning_rate": 4.91316213924935e-06,
+      "loss": 0.7952,
+      "step": 130
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.8114602189230878,
+      "learning_rate": 4.911828972678441e-06,
+      "loss": 0.6176,
+      "step": 131
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.8164393262307038,
+      "learning_rate": 4.91048583394987e-06,
+      "loss": 0.6477,
+      "step": 132
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.8581250231339308,
+      "learning_rate": 4.909132728617095e-06,
+      "loss": 0.9476,
+      "step": 133
+    },
+    {
+      "epoch": 1.3009708737864076,
+      "grad_norm": 0.766359434418812,
+      "learning_rate": 4.907769662274785e-06,
+      "loss": 0.6528,
+      "step": 134
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.6976692264719603,
+      "learning_rate": 4.90639664055879e-06,
+      "loss": 0.5637,
+      "step": 135
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.7314627468283406,
+      "learning_rate": 4.905013669146127e-06,
+      "loss": 0.6096,
+      "step": 136
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.7773436570942428,
+      "learning_rate": 4.903620753754949e-06,
+      "loss": 0.7461,
+      "step": 137
+    },
+    {
+      "epoch": 1.3398058252427185,
+      "grad_norm": 0.8921198866699622,
+      "learning_rate": 4.902217900144524e-06,
+      "loss": 0.9358,
+      "step": 138
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.7375485723614339,
+      "learning_rate": 4.900805114115214e-06,
+      "loss": 0.8942,
+      "step": 139
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.7942794591041569,
+      "learning_rate": 4.899382401508446e-06,
+      "loss": 0.5492,
+      "step": 140
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.8078627057423822,
+      "learning_rate": 4.8979497682066916e-06,
+      "loss": 0.7808,
+      "step": 141
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.6837805173092772,
+      "learning_rate": 4.89650722013344e-06,
+      "loss": 0.5391,
+      "step": 142
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 0.9832481462127313,
+      "learning_rate": 4.895054763253177e-06,
+      "loss": 0.7406,
+      "step": 143
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.769076719995811,
+      "learning_rate": 4.8935924035713564e-06,
+      "loss": 0.5929,
+      "step": 144
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.8430007967016218,
+      "learning_rate": 4.892120147134378e-06,
+      "loss": 0.699,
+      "step": 145
+    },
+    {
+      "epoch": 1.4174757281553398,
+      "grad_norm": 0.9122194953590461,
+      "learning_rate": 4.8906380000295615e-06,
+      "loss": 0.6895,
+      "step": 146
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.7898786206966036,
+      "learning_rate": 4.889145968385121e-06,
+      "loss": 0.6528,
+      "step": 147
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.7666616277046007,
+      "learning_rate": 4.887644058370139e-06,
+      "loss": 0.531,
+      "step": 148
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.7653718997690864,
+      "learning_rate": 4.886132276194544e-06,
+      "loss": 0.5768,
+      "step": 149
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.90831243558374,
+      "learning_rate": 4.884610628109082e-06,
+      "loss": 0.5652,
+      "step": 150
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.7442744457963555,
+      "learning_rate": 4.883079120405292e-06,
+      "loss": 0.6688,
+      "step": 151
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.7344457110605118,
+      "learning_rate": 4.881537759415478e-06,
+      "loss": 0.5314,
+      "step": 152
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 1.1695813268075903,
+      "learning_rate": 4.879986551512684e-06,
+      "loss": 0.7273,
+      "step": 153
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.6738881315688398,
+      "learning_rate": 4.878425503110672e-06,
+      "loss": 0.5607,
+      "step": 154
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.8581543167197856,
+      "learning_rate": 4.876854620663887e-06,
+      "loss": 0.6943,
+      "step": 155
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 0.8060496260215426,
+      "learning_rate": 4.875273910667434e-06,
+      "loss": 0.6212,
+      "step": 156
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.8453811238639173,
+      "learning_rate": 4.873683379657057e-06,
+      "loss": 0.5456,
+      "step": 157
+    },
+    {
+      "epoch": 1.5339805825242718,
+      "grad_norm": 0.8844708129131296,
+      "learning_rate": 4.8720830342091015e-06,
+      "loss": 0.9448,
+      "step": 158
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.861907269409231,
+      "learning_rate": 4.870472880940496e-06,
+      "loss": 0.7925,
+      "step": 159
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 1.0829307275890252,
+      "learning_rate": 4.868852926508721e-06,
+      "loss": 0.8343,
+      "step": 160
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 0.875424595240487,
+      "learning_rate": 4.867223177611779e-06,
+      "loss": 0.7158,
+      "step": 161
+    },
+    {
+      "epoch": 1.5728155339805825,
+      "grad_norm": 0.6659133214470974,
+      "learning_rate": 4.865583640988173e-06,
+      "loss": 0.5684,
+      "step": 162
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.772588673794732,
+      "learning_rate": 4.863934323416871e-06,
+      "loss": 0.6486,
+      "step": 163
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.741499702229279,
+      "learning_rate": 4.862275231717288e-06,
+      "loss": 0.5758,
+      "step": 164
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.920336013887185,
+      "learning_rate": 4.860606372749247e-06,
+      "loss": 0.5468,
+      "step": 165
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.7093065195416415,
+      "learning_rate": 4.858927753412958e-06,
+      "loss": 0.6127,
+      "step": 166
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.8110174812695655,
+      "learning_rate": 4.857239380648985e-06,
+      "loss": 0.6017,
+      "step": 167
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.7375215564893128,
+      "learning_rate": 4.855541261438223e-06,
+      "loss": 0.6753,
+      "step": 168
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.7500973220432647,
+      "learning_rate": 4.8538334028018605e-06,
+      "loss": 0.8343,
+      "step": 169
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.6948713105727569,
+      "learning_rate": 4.8521158118013605e-06,
+      "loss": 0.5493,
+      "step": 170
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 0.773755912005905,
+      "learning_rate": 4.850388495538423e-06,
+      "loss": 0.6271,
+      "step": 171
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.8759327452228973,
+      "learning_rate": 4.84865146115496e-06,
+      "loss": 0.5641,
+      "step": 172
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 0.8673872477886911,
+      "learning_rate": 4.846904715833066e-06,
+      "loss": 0.5295,
+      "step": 173
+    },
+    {
+      "epoch": 1.6893203883495147,
+      "grad_norm": 0.7435589108212891,
+      "learning_rate": 4.8451482667949836e-06,
+      "loss": 0.7122,
+      "step": 174
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.7510040594244859,
+      "learning_rate": 4.843382121303082e-06,
+      "loss": 0.6579,
+      "step": 175
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.7254228155893363,
+      "learning_rate": 4.841606286659819e-06,
+      "loss": 0.5591,
+      "step": 176
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.7786787819553973,
+      "learning_rate": 4.839820770207714e-06,
+      "loss": 0.5417,
+      "step": 177
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.7424090664501856,
+      "learning_rate": 4.8380255793293195e-06,
+      "loss": 0.5679,
+      "step": 178
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 0.880425961769788,
+      "learning_rate": 4.8362207214471864e-06,
+      "loss": 0.5504,
+      "step": 179
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.7692052821337514,
+      "learning_rate": 4.83440620402384e-06,
+      "loss": 0.6217,
+      "step": 180
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.7571562044181578,
+      "learning_rate": 4.832582034561738e-06,
+      "loss": 0.641,
+      "step": 181
+    },
+    {
+      "epoch": 1.766990291262136,
+      "grad_norm": 0.8286514373124609,
+      "learning_rate": 4.830748220603251e-06,
+      "loss": 0.7462,
+      "step": 182
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 0.8663863630421635,
+      "learning_rate": 4.828904769730628e-06,
+      "loss": 0.7598,
+      "step": 183
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.8392430546644409,
+      "learning_rate": 4.827051689565958e-06,
+      "loss": 0.6278,
+      "step": 184
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.7431256674232911,
+      "learning_rate": 4.825188987771149e-06,
+      "loss": 0.7471,
+      "step": 185
+    },
+    {
+      "epoch": 1.8058252427184467,
+      "grad_norm": 0.775602420628486,
+      "learning_rate": 4.82331667204789e-06,
+      "loss": 0.4658,
+      "step": 186
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.8220707289832607,
+      "learning_rate": 4.821434750137619e-06,
+      "loss": 0.6212,
+      "step": 187
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.8721039703571108,
+      "learning_rate": 4.819543229821494e-06,
+      "loss": 0.7135,
+      "step": 188
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.7196046947557964,
+      "learning_rate": 4.8176421189203605e-06,
+      "loss": 0.5464,
+      "step": 189
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.7844213665577071,
+      "learning_rate": 4.815731425294716e-06,
+      "loss": 0.6536,
+      "step": 190
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.7921297235454372,
+      "learning_rate": 4.813811156844681e-06,
+      "loss": 0.7183,
+      "step": 191
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 0.863196612505686,
+      "learning_rate": 4.811881321509964e-06,
+      "loss": 0.6976,
+      "step": 192
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.7755881818234687,
+      "learning_rate": 4.809941927269829e-06,
+      "loss": 0.8491,
+      "step": 193
+    },
+    {
+      "epoch": 1.883495145631068,
+      "grad_norm": 0.8186186296199729,
+      "learning_rate": 4.807992982143064e-06,
+      "loss": 0.6343,
+      "step": 194
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.913496737962807,
+      "learning_rate": 4.806034494187949e-06,
+      "loss": 0.8436,
+      "step": 195
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.7590143443155779,
+      "learning_rate": 4.804066471502216e-06,
+      "loss": 0.603,
+      "step": 196
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.766119646590294,
+      "learning_rate": 4.802088922223024e-06,
+      "loss": 0.5167,
+      "step": 197
+    },
+    {
+      "epoch": 1.9223300970873787,
+      "grad_norm": 0.7274017571399753,
+      "learning_rate": 4.80010185452692e-06,
+      "loss": 0.6725,
+      "step": 198
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.73780689049117,
+      "learning_rate": 4.798105276629806e-06,
+      "loss": 0.521,
+      "step": 199
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.7791155049733209,
+      "learning_rate": 4.796099196786908e-06,
+      "loss": 0.583,
+      "step": 200
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.8004953963543238,
+      "learning_rate": 4.794083623292737e-06,
+      "loss": 0.7842,
+      "step": 201
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.7807030577892377,
+      "learning_rate": 4.792058564481058e-06,
+      "loss": 0.5534,
+      "step": 202
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.7092633632036252,
+      "learning_rate": 4.7900240287248554e-06,
+      "loss": 0.7083,
+      "step": 203
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.8875925228430702,
+      "learning_rate": 4.7879800244362975e-06,
+      "loss": 0.7782,
+      "step": 204
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.8909791320483021,
+      "learning_rate": 4.785926560066703e-06,
+      "loss": 0.7681,
+      "step": 205
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7275875515555469,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.649,
+      "step": 206
+    },
+    {
+      "epoch": 2.0097087378640777,
+      "grad_norm": 0.802116606621558,
+      "learning_rate": 4.781791285085209e-06,
+      "loss": 0.5568,
+      "step": 207
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.7406594271373391,
+      "learning_rate": 4.779709491571378e-06,
+      "loss": 0.5144,
+      "step": 208
+    },
+    {
+      "epoch": 2.029126213592233,
+      "grad_norm": 0.6857989185878482,
+      "learning_rate": 4.777618272172573e-06,
+      "loss": 0.6449,
+      "step": 209
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.734748210095666,
+      "learning_rate": 4.775517635535332e-06,
+      "loss": 0.7377,
+      "step": 210
+    },
+    {
+      "epoch": 2.0485436893203883,
+      "grad_norm": 0.7708218856857236,
+      "learning_rate": 4.77340759034513e-06,
+      "loss": 0.6855,
+      "step": 211
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 0.7279897754942517,
+      "learning_rate": 4.771288145326343e-06,
+      "loss": 0.6684,
+      "step": 212
+    },
+    {
+      "epoch": 2.0679611650485437,
+      "grad_norm": 0.674474791686522,
+      "learning_rate": 4.769159309242213e-06,
+      "loss": 0.438,
+      "step": 213
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 0.792505397680811,
+      "learning_rate": 4.767021090894809e-06,
+      "loss": 0.6831,
+      "step": 214
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.744545194492276,
+      "learning_rate": 4.764873499124997e-06,
+      "loss": 0.6976,
+      "step": 215
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.7354909274219694,
+      "learning_rate": 4.762716542812395e-06,
+      "loss": 0.5495,
+      "step": 216
+    },
+    {
+      "epoch": 2.1067961165048543,
+      "grad_norm": 0.7133200950602842,
+      "learning_rate": 4.7605502308753415e-06,
+      "loss": 0.5687,
+      "step": 217
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 0.7982098464573162,
+      "learning_rate": 4.758374572270859e-06,
+      "loss": 0.6886,
+      "step": 218
+    },
+    {
+      "epoch": 2.1262135922330097,
+      "grad_norm": 0.6913449166288441,
+      "learning_rate": 4.756189575994614e-06,
+      "loss": 0.3812,
+      "step": 219
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.7192629747712952,
+      "learning_rate": 4.753995251080884e-06,
+      "loss": 0.575,
+      "step": 220
+    },
+    {
+      "epoch": 2.145631067961165,
+      "grad_norm": 0.6248283732386594,
+      "learning_rate": 4.7517916066025126e-06,
+      "loss": 0.463,
+      "step": 221
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.8287445715621566,
+      "learning_rate": 4.7495786516708806e-06,
+      "loss": 0.7071,
+      "step": 222
+    },
+    {
+      "epoch": 2.1650485436893203,
+      "grad_norm": 0.8318311223138934,
+      "learning_rate": 4.747356395435865e-06,
+      "loss": 0.4322,
+      "step": 223
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.8533840933951986,
+      "learning_rate": 4.745124847085799e-06,
+      "loss": 0.568,
+      "step": 224
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.7901361363585824,
+      "learning_rate": 4.742884015847436e-06,
+      "loss": 0.5393,
+      "step": 225
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 0.8111627610336898,
+      "learning_rate": 4.740633910985911e-06,
+      "loss": 0.5969,
+      "step": 226
+    },
+    {
+      "epoch": 2.203883495145631,
+      "grad_norm": 0.7671540334640228,
+      "learning_rate": 4.738374541804704e-06,
+      "loss": 0.4576,
+      "step": 227
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.7242562019069873,
+      "learning_rate": 4.7361059176456e-06,
+      "loss": 0.5186,
+      "step": 228
+    },
+    {
+      "epoch": 2.2233009708737863,
+      "grad_norm": 0.7763865691421442,
+      "learning_rate": 4.733828047888647e-06,
+      "loss": 0.6682,
+      "step": 229
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.6906133600693856,
+      "learning_rate": 4.731540941952126e-06,
+      "loss": 0.4051,
+      "step": 230
+    },
+    {
+      "epoch": 2.2427184466019416,
+      "grad_norm": 0.7481490824065317,
+      "learning_rate": 4.7292446092925016e-06,
+      "loss": 0.4902,
+      "step": 231
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.7044106920537051,
+      "learning_rate": 4.726939059404392e-06,
+      "loss": 0.4499,
+      "step": 232
+    },
+    {
+      "epoch": 2.262135922330097,
+      "grad_norm": 0.6853728707124397,
+      "learning_rate": 4.724624301820524e-06,
+      "loss": 0.4757,
+      "step": 233
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 0.9364489830239873,
+      "learning_rate": 4.722300346111695e-06,
+      "loss": 0.6159,
+      "step": 234
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.8138300584150231,
+      "learning_rate": 4.719967201886734e-06,
+      "loss": 0.5365,
+      "step": 235
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.7475372553230331,
+      "learning_rate": 4.717624878792461e-06,
+      "loss": 0.7347,
+      "step": 236
+    },
+    {
+      "epoch": 2.3009708737864076,
+      "grad_norm": 0.9795714505586022,
+      "learning_rate": 4.715273386513651e-06,
+      "loss": 0.7175,
+      "step": 237
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 0.7332975150286841,
+      "learning_rate": 4.712912734772988e-06,
+      "loss": 0.5204,
+      "step": 238
+    },
+    {
+      "epoch": 2.320388349514563,
+      "grad_norm": 0.6488936041709285,
+      "learning_rate": 4.710542933331025e-06,
+      "loss": 0.5825,
+      "step": 239
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.7371958257036496,
+      "learning_rate": 4.708163991986152e-06,
+      "loss": 0.6679,
+      "step": 240
+    },
+    {
+      "epoch": 2.3398058252427183,
+      "grad_norm": 0.8214266450639852,
+      "learning_rate": 4.705775920574546e-06,
+      "loss": 0.6434,
+      "step": 241
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 0.7702926046174587,
+      "learning_rate": 4.703378728970134e-06,
+      "loss": 0.4755,
+      "step": 242
+    },
+    {
+      "epoch": 2.3592233009708736,
+      "grad_norm": 0.6996698003201337,
+      "learning_rate": 4.700972427084551e-06,
+      "loss": 0.592,
+      "step": 243
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.7560377791738715,
+      "learning_rate": 4.698557024867105e-06,
+      "loss": 0.4565,
+      "step": 244
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.7633818347510686,
+      "learning_rate": 4.696132532304727e-06,
+      "loss": 0.6659,
+      "step": 245
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 0.7234976470337168,
+      "learning_rate": 4.693698959421935e-06,
+      "loss": 0.5678,
+      "step": 246
+    },
+    {
+      "epoch": 2.3980582524271843,
+      "grad_norm": 0.750666487544434,
+      "learning_rate": 4.691256316280789e-06,
+      "loss": 0.8332,
+      "step": 247
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.7493376372555444,
+      "learning_rate": 4.688804612980855e-06,
+      "loss": 0.4837,
+      "step": 248
+    },
+    {
+      "epoch": 2.4174757281553396,
+      "grad_norm": 0.6751745421938132,
+      "learning_rate": 4.686343859659158e-06,
+      "loss": 0.6951,
+      "step": 249
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.734057362640619,
+      "learning_rate": 4.683874066490143e-06,
+      "loss": 0.5507,
+      "step": 250
+    },
+    {
+      "epoch": 2.436893203883495,
+      "grad_norm": 0.7227243519681933,
+      "learning_rate": 4.681395243685631e-06,
+      "loss": 0.56,
+      "step": 251
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.7751770457742055,
+      "learning_rate": 4.67890740149478e-06,
+      "loss": 0.518,
+      "step": 252
+    },
+    {
+      "epoch": 2.4563106796116507,
+      "grad_norm": 0.8010398356492873,
+      "learning_rate": 4.676410550204036e-06,
+      "loss": 0.6707,
+      "step": 253
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.6322959904355947,
+      "learning_rate": 4.673904700137098e-06,
+      "loss": 0.3992,
+      "step": 254
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.6951997535390679,
+      "learning_rate": 4.671389861654873e-06,
+      "loss": 0.4745,
+      "step": 255
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.7106909584689424,
+      "learning_rate": 4.668866045155428e-06,
+      "loss": 0.596,
+      "step": 256
+    },
+    {
+      "epoch": 2.4951456310679614,
+      "grad_norm": 0.8847918134161639,
+      "learning_rate": 4.666333261073956e-06,
+      "loss": 0.6131,
+      "step": 257
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 0.7875255376725511,
+      "learning_rate": 4.6637915198827265e-06,
+      "loss": 0.6808,
+      "step": 258
+    },
+    {
+      "epoch": 2.5145631067961167,
+      "grad_norm": 0.74457845575171,
+      "learning_rate": 4.661240832091042e-06,
+      "loss": 0.534,
+      "step": 259
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.850799682933559,
+      "learning_rate": 4.658681208245198e-06,
+      "loss": 0.6302,
+      "step": 260
+    },
+    {
+      "epoch": 2.533980582524272,
+      "grad_norm": 0.7621063847057238,
+      "learning_rate": 4.65611265892844e-06,
+      "loss": 0.5924,
+      "step": 261
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.8162383044494845,
+      "learning_rate": 4.653535194760912e-06,
+      "loss": 0.5497,
+      "step": 262
+    },
+    {
+      "epoch": 2.5533980582524274,
+      "grad_norm": 0.7920894927810362,
+      "learning_rate": 4.650948826399624e-06,
+      "loss": 0.392,
+      "step": 263
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.7383821620603919,
+      "learning_rate": 4.648353564538397e-06,
+      "loss": 0.5049,
+      "step": 264
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.7650056257507512,
+      "learning_rate": 4.645749419907829e-06,
+      "loss": 0.4973,
+      "step": 265
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 0.7120510594433922,
+      "learning_rate": 4.64313640327524e-06,
+      "loss": 0.406,
+      "step": 266
+    },
+    {
+      "epoch": 2.592233009708738,
+      "grad_norm": 0.7835448578972172,
+      "learning_rate": 4.640514525444637e-06,
+      "loss": 0.6122,
+      "step": 267
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 0.9134409217464837,
+      "learning_rate": 4.637883797256663e-06,
+      "loss": 0.845,
+      "step": 268
+    },
+    {
+      "epoch": 2.6116504854368934,
+      "grad_norm": 0.7387942613013898,
+      "learning_rate": 4.635244229588558e-06,
+      "loss": 0.5296,
+      "step": 269
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.6722442278200302,
+      "learning_rate": 4.632595833354105e-06,
+      "loss": 0.6111,
+      "step": 270
+    },
+    {
+      "epoch": 2.6310679611650487,
+      "grad_norm": 0.790100355706828,
+      "learning_rate": 4.629938619503593e-06,
+      "loss": 0.4931,
+      "step": 271
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.7457720631344418,
+      "learning_rate": 4.627272599023772e-06,
+      "loss": 0.6932,
+      "step": 272
+    },
+    {
+      "epoch": 2.650485436893204,
+      "grad_norm": 0.7742836603714502,
+      "learning_rate": 4.6245977829378e-06,
+      "loss": 0.8069,
+      "step": 273
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.7522977861605561,
+      "learning_rate": 4.6219141823052035e-06,
+      "loss": 0.4594,
+      "step": 274
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.7897785740037921,
+      "learning_rate": 4.619221808221833e-06,
+      "loss": 0.7502,
+      "step": 275
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.6542147168655412,
+      "learning_rate": 4.616520671819812e-06,
+      "loss": 0.455,
+      "step": 276
+    },
+    {
+      "epoch": 2.6893203883495147,
+      "grad_norm": 1.3324292561191493,
+      "learning_rate": 4.613810784267492e-06,
+      "loss": 0.615,
+      "step": 277
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.6366824455259196,
+      "learning_rate": 4.61109215676941e-06,
+      "loss": 0.4398,
+      "step": 278
+    },
+    {
+      "epoch": 2.70873786407767,
+      "grad_norm": 0.7464337608282651,
+      "learning_rate": 4.608364800566241e-06,
+      "loss": 0.587,
+      "step": 279
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.6895928222340304,
+      "learning_rate": 4.605628726934747e-06,
+      "loss": 0.6703,
+      "step": 280
+    },
+    {
+      "epoch": 2.7281553398058254,
+      "grad_norm": 0.6878653541621508,
+      "learning_rate": 4.602883947187738e-06,
+      "loss": 0.5948,
+      "step": 281
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.6926203162726917,
+      "learning_rate": 4.600130472674017e-06,
+      "loss": 0.4607,
+      "step": 282
+    },
+    {
+      "epoch": 2.7475728155339807,
+      "grad_norm": 0.7781936564011823,
+      "learning_rate": 4.5973683147783405e-06,
+      "loss": 0.7486,
+      "step": 283
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.7501081998810354,
+      "learning_rate": 4.594597484921365e-06,
+      "loss": 0.5782,
+      "step": 284
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.7647372977005877,
+      "learning_rate": 4.5918179945596055e-06,
+      "loss": 0.5159,
+      "step": 285
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.6977238852528742,
+      "learning_rate": 4.589029855185384e-06,
+      "loss": 0.5334,
+      "step": 286
+    },
+    {
+      "epoch": 2.7864077669902914,
+      "grad_norm": 0.9130101610006348,
+      "learning_rate": 4.586233078326785e-06,
+      "loss": 0.8354,
+      "step": 287
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.8318292434477851,
+      "learning_rate": 4.583427675547602e-06,
+      "loss": 0.6258,
+      "step": 288
+    },
+    {
+      "epoch": 2.8058252427184467,
+      "grad_norm": 0.7324981921403528,
+      "learning_rate": 4.580613658447301e-06,
+      "loss": 0.709,
+      "step": 289
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.7523751933345019,
+      "learning_rate": 4.577791038660959e-06,
+      "loss": 0.4851,
+      "step": 290
+    },
+    {
+      "epoch": 2.825242718446602,
+      "grad_norm": 0.7542503570491638,
+      "learning_rate": 4.574959827859226e-06,
+      "loss": 0.4493,
+      "step": 291
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.741187265691656,
+      "learning_rate": 4.572120037748273e-06,
+      "loss": 0.4682,
+      "step": 292
+    },
+    {
+      "epoch": 2.8446601941747574,
+      "grad_norm": 0.7141919069039983,
+      "learning_rate": 4.5692716800697415e-06,
+      "loss": 0.6235,
+      "step": 293
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.7414161387983556,
+      "learning_rate": 4.566414766600698e-06,
+      "loss": 0.5613,
+      "step": 294
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.700849509216115,
+      "learning_rate": 4.563549309153589e-06,
+      "loss": 0.443,
+      "step": 295
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.7854807058221909,
+      "learning_rate": 4.56067531957618e-06,
+      "loss": 0.4657,
+      "step": 296
+    },
+    {
+      "epoch": 2.883495145631068,
+      "grad_norm": 0.7744368603433763,
+      "learning_rate": 4.557792809751519e-06,
+      "loss": 0.6192,
+      "step": 297
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.811186979862698,
+      "learning_rate": 4.554901791597883e-06,
+      "loss": 0.5432,
+      "step": 298
+    },
+    {
+      "epoch": 2.9029126213592233,
+      "grad_norm": 0.7610230989372687,
+      "learning_rate": 4.552002277068725e-06,
+      "loss": 0.5689,
+      "step": 299
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.738268676455508,
+      "learning_rate": 4.549094278152631e-06,
+      "loss": 0.6102,
+      "step": 300
+    },
+    {
+      "epoch": 2.9223300970873787,
+      "grad_norm": 0.8265276954113333,
+      "learning_rate": 4.546177806873266e-06,
+      "loss": 0.4803,
+      "step": 301
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 0.7994330788168778,
+      "learning_rate": 4.543252875289326e-06,
+      "loss": 0.5232,
+      "step": 302
+    },
+    {
+      "epoch": 2.941747572815534,
+      "grad_norm": 0.7803228028067196,
+      "learning_rate": 4.540319495494486e-06,
+      "loss": 0.5785,
+      "step": 303
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.812661521852206,
+      "learning_rate": 4.537377679617353e-06,
+      "loss": 0.5857,
+      "step": 304
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.8017682940418599,
+      "learning_rate": 4.534427439821416e-06,
+      "loss": 0.4679,
+      "step": 305
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 0.7557388320666087,
+      "learning_rate": 4.531468788304992e-06,
+      "loss": 0.4511,
+      "step": 306
+    },
+    {
+      "epoch": 2.9805825242718447,
+      "grad_norm": 0.7990082356636408,
+      "learning_rate": 4.5285017373011784e-06,
+      "loss": 0.4999,
+      "step": 307
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.8410650831835784,
+      "learning_rate": 4.5255262990778024e-06,
+      "loss": 0.4279,
+      "step": 308
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.67842206115732,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.4778,
+      "step": 309
+    },
+    {
+      "epoch": 3.0097087378640777,
+      "grad_norm": 0.7208972107613384,
+      "learning_rate": 4.519550310217013e-06,
+      "loss": 0.5166,
+      "step": 310
+    },
+    {
+      "epoch": 3.0194174757281553,
+      "grad_norm": 0.73349262960244,
+      "learning_rate": 4.516549784288442e-06,
+      "loss": 0.4455,
+      "step": 311
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.6956083097152967,
+      "learning_rate": 4.513540920557892e-06,
+      "loss": 0.4371,
+      "step": 312
+    },
+    {
+      "epoch": 3.0388349514563107,
+      "grad_norm": 0.7915417702341847,
+      "learning_rate": 4.510523731466072e-06,
+      "loss": 0.6448,
+      "step": 313
+    },
+    {
+      "epoch": 3.0485436893203883,
+      "grad_norm": 0.5990172469379881,
+      "learning_rate": 4.507498229488116e-06,
+      "loss": 0.3384,
+      "step": 314
+    },
+    {
+      "epoch": 3.058252427184466,
+      "grad_norm": 0.8318855981134111,
+      "learning_rate": 4.504464427133527e-06,
+      "loss": 0.4508,
+      "step": 315
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 0.8758812451211815,
+      "learning_rate": 4.501422336946126e-06,
+      "loss": 0.4328,
+      "step": 316
+    },
+    {
+      "epoch": 3.0776699029126213,
+      "grad_norm": 0.7844299628970943,
+      "learning_rate": 4.498371971504005e-06,
+      "loss": 0.4222,
+      "step": 317
+    },
+    {
+      "epoch": 3.087378640776699,
+      "grad_norm": 0.7195150581975005,
+      "learning_rate": 4.49531334341947e-06,
+      "loss": 0.4417,
+      "step": 318
+    },
+    {
+      "epoch": 3.0970873786407767,
+      "grad_norm": 0.7671602770728359,
+      "learning_rate": 4.49224646533899e-06,
+      "loss": 0.42,
+      "step": 319
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 0.6929615407642132,
+      "learning_rate": 4.489171349943144e-06,
+      "loss": 0.4332,
+      "step": 320
+    },
+    {
+      "epoch": 3.116504854368932,
+      "grad_norm": 0.7086258123510658,
+      "learning_rate": 4.486088009946575e-06,
+      "loss": 0.3632,
+      "step": 321
+    },
+    {
+      "epoch": 3.1262135922330097,
+      "grad_norm": 0.7708151250786913,
+      "learning_rate": 4.482996458097926e-06,
+      "loss": 0.4975,
+      "step": 322
+    },
+    {
+      "epoch": 3.1359223300970873,
+      "grad_norm": 0.846155524161367,
+      "learning_rate": 4.479896707179796e-06,
+      "loss": 0.6871,
+      "step": 323
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 0.8231242232551517,
+      "learning_rate": 4.476788770008685e-06,
+      "loss": 0.6574,
+      "step": 324
+    },
+    {
+      "epoch": 3.1553398058252426,
+      "grad_norm": 0.8549632502448101,
+      "learning_rate": 4.473672659434941e-06,
+      "loss": 0.5856,
+      "step": 325
+    },
+    {
+      "epoch": 3.1650485436893203,
+      "grad_norm": 0.7228128426745009,
+      "learning_rate": 4.470548388342704e-06,
+      "loss": 0.4776,
+      "step": 326
+    },
+    {
+      "epoch": 3.174757281553398,
+      "grad_norm": 0.8108949680823984,
+      "learning_rate": 4.467415969649858e-06,
+      "loss": 0.4874,
+      "step": 327
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 0.6851857911673899,
+      "learning_rate": 4.464275416307973e-06,
+      "loss": 0.4994,
+      "step": 328
+    },
+    {
+      "epoch": 3.1941747572815533,
+      "grad_norm": 0.761950009953729,
+      "learning_rate": 4.461126741302253e-06,
+      "loss": 0.4929,
+      "step": 329
+    },
+    {
+      "epoch": 3.203883495145631,
+      "grad_norm": 0.8432338612654883,
+      "learning_rate": 4.457969957651485e-06,
+      "loss": 0.4137,
+      "step": 330
+    },
+    {
+      "epoch": 3.2135922330097086,
+      "grad_norm": 0.7741138798177974,
+      "learning_rate": 4.454805078407979e-06,
+      "loss": 0.6696,
+      "step": 331
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.71765697779996,
+      "learning_rate": 4.451632116657521e-06,
+      "loss": 0.4506,
+      "step": 332
+    },
+    {
+      "epoch": 3.233009708737864,
+      "grad_norm": 0.6253829454649015,
+      "learning_rate": 4.448451085519314e-06,
+      "loss": 0.3586,
+      "step": 333
+    },
+    {
+      "epoch": 3.2427184466019416,
+      "grad_norm": 0.9686569714580476,
+      "learning_rate": 4.445261998145927e-06,
+      "loss": 0.4832,
+      "step": 334
+    },
+    {
+      "epoch": 3.2524271844660193,
+      "grad_norm": 0.682456331535341,
+      "learning_rate": 4.442064867723236e-06,
+      "loss": 0.4737,
+      "step": 335
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 0.801934212380896,
+      "learning_rate": 4.438859707470376e-06,
+      "loss": 0.4988,
+      "step": 336
+    },
+    {
+      "epoch": 3.2718446601941746,
+      "grad_norm": 0.6210847995583545,
+      "learning_rate": 4.435646530639679e-06,
+      "loss": 0.3549,
+      "step": 337
+    },
+    {
+      "epoch": 3.2815533980582523,
+      "grad_norm": 0.7693142736109597,
+      "learning_rate": 4.432425350516627e-06,
+      "loss": 0.4612,
+      "step": 338
+    },
+    {
+      "epoch": 3.29126213592233,
+      "grad_norm": 0.7469011823723339,
+      "learning_rate": 4.42919618041979e-06,
+      "loss": 0.3343,
+      "step": 339
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 0.6080757362122666,
+      "learning_rate": 4.425959033700776e-06,
+      "loss": 0.2422,
+      "step": 340
+    },
+    {
+      "epoch": 3.3106796116504853,
+      "grad_norm": 0.7696858723921102,
+      "learning_rate": 4.422713923744174e-06,
+      "loss": 0.5103,
+      "step": 341
+    },
+    {
+      "epoch": 3.320388349514563,
+      "grad_norm": 0.8778656215024975,
+      "learning_rate": 4.419460863967496e-06,
+      "loss": 0.5335,
+      "step": 342
+    },
+    {
+      "epoch": 3.3300970873786406,
+      "grad_norm": 0.6994532036876416,
+      "learning_rate": 4.416199867821126e-06,
+      "loss": 0.5022,
+      "step": 343
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 0.7969656894134531,
+      "learning_rate": 4.412930948788263e-06,
+      "loss": 0.4246,
+      "step": 344
+    },
+    {
+      "epoch": 3.349514563106796,
+      "grad_norm": 0.8154555970057127,
+      "learning_rate": 4.409654120384863e-06,
+      "loss": 0.6664,
+      "step": 345
+    },
+    {
+      "epoch": 3.3592233009708736,
+      "grad_norm": 0.7113779335682774,
+      "learning_rate": 4.406369396159585e-06,
+      "loss": 0.6024,
+      "step": 346
+    },
+    {
+      "epoch": 3.3689320388349513,
+      "grad_norm": 0.7946176349594999,
+      "learning_rate": 4.403076789693735e-06,
+      "loss": 0.6273,
+      "step": 347
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 0.6725243345900541,
+      "learning_rate": 4.399776314601212e-06,
+      "loss": 0.422,
+      "step": 348
+    },
+    {
+      "epoch": 3.3883495145631066,
+      "grad_norm": 0.8474457763132652,
+      "learning_rate": 4.396467984528445e-06,
+      "loss": 0.5515,
+      "step": 349
+    },
+    {
+      "epoch": 3.3980582524271843,
+      "grad_norm": 0.7635758657443303,
+      "learning_rate": 4.393151813154345e-06,
+      "loss": 0.417,
+      "step": 350
+    },
+    {
+      "epoch": 3.407766990291262,
+      "grad_norm": 0.7301687492460268,
+      "learning_rate": 4.3898278141902396e-06,
+      "loss": 0.4335,
+      "step": 351
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 0.8217888155242994,
+      "learning_rate": 4.386496001379826e-06,
+      "loss": 0.5301,
+      "step": 352
+    },
+    {
+      "epoch": 3.4271844660194173,
+      "grad_norm": 0.700440895441813,
+      "learning_rate": 4.383156388499106e-06,
+      "loss": 0.5289,
+      "step": 353
+    },
+    {
+      "epoch": 3.436893203883495,
+      "grad_norm": 0.676777993038823,
+      "learning_rate": 4.3798089893563335e-06,
+      "loss": 0.4079,
+      "step": 354
+    },
+    {
+      "epoch": 3.4466019417475726,
+      "grad_norm": 0.7359945465771945,
+      "learning_rate": 4.3764538177919555e-06,
+      "loss": 0.3024,
+      "step": 355
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 0.7431315486816701,
+      "learning_rate": 4.3730908876785574e-06,
+      "loss": 0.4715,
+      "step": 356
+    },
+    {
+      "epoch": 3.466019417475728,
+      "grad_norm": 0.858728130994808,
+      "learning_rate": 4.3697202129208e-06,
+      "loss": 0.5126,
+      "step": 357
+    },
+    {
+      "epoch": 3.475728155339806,
+      "grad_norm": 0.7006328621067651,
+      "learning_rate": 4.36634180745537e-06,
+      "loss": 0.3219,
+      "step": 358
+    },
+    {
+      "epoch": 3.4854368932038833,
+      "grad_norm": 0.8347983820993932,
+      "learning_rate": 4.3629556852509145e-06,
+      "loss": 0.7038,
+      "step": 359
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.6736438089942041,
+      "learning_rate": 4.35956186030799e-06,
+      "loss": 0.388,
+      "step": 360
+    },
+    {
+      "epoch": 3.5048543689320386,
+      "grad_norm": 0.7313432635941632,
+      "learning_rate": 4.356160346659001e-06,
+      "loss": 0.4803,
+      "step": 361
+    },
+    {
+      "epoch": 3.5145631067961167,
+      "grad_norm": 0.914378556446873,
+      "learning_rate": 4.3527511583681384e-06,
+      "loss": 0.9472,
+      "step": 362
+    },
+    {
+      "epoch": 3.524271844660194,
+      "grad_norm": 0.7928555788899265,
+      "learning_rate": 4.34933430953133e-06,
+      "loss": 0.5195,
+      "step": 363
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 0.8552957566999786,
+      "learning_rate": 4.345909814276177e-06,
+      "loss": 0.5432,
+      "step": 364
+    },
+    {
+      "epoch": 3.5436893203883493,
+      "grad_norm": 0.6593786360114855,
+      "learning_rate": 4.3424776867618935e-06,
+      "loss": 0.5773,
+      "step": 365
+    },
+    {
+      "epoch": 3.5533980582524274,
+      "grad_norm": 0.7453073470450546,
+      "learning_rate": 4.339037941179253e-06,
+      "loss": 0.5414,
+      "step": 366
+    },
+    {
+      "epoch": 3.5631067961165046,
+      "grad_norm": 0.7285347424646024,
+      "learning_rate": 4.335590591750526e-06,
+      "loss": 0.4336,
+      "step": 367
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 0.7229430966497608,
+      "learning_rate": 4.332135652729423e-06,
+      "loss": 0.4226,
+      "step": 368
+    },
+    {
+      "epoch": 3.58252427184466,
+      "grad_norm": 0.7264295125746805,
+      "learning_rate": 4.328673138401036e-06,
+      "loss": 0.5502,
+      "step": 369
+    },
+    {
+      "epoch": 3.592233009708738,
+      "grad_norm": 0.7134753381323626,
+      "learning_rate": 4.325203063081776e-06,
+      "loss": 0.2876,
+      "step": 370
+    },
+    {
+      "epoch": 3.6019417475728153,
+      "grad_norm": 0.7636004423408037,
+      "learning_rate": 4.32172544111932e-06,
+      "loss": 0.3651,
+      "step": 371
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 0.731718208729643,
+      "learning_rate": 4.318240286892544e-06,
+      "loss": 0.4391,
+      "step": 372
+    },
+    {
+      "epoch": 3.6213592233009706,
+      "grad_norm": 0.6392786785401195,
+      "learning_rate": 4.314747614811471e-06,
+      "loss": 0.4575,
+      "step": 373
+    },
+    {
+      "epoch": 3.6310679611650487,
+      "grad_norm": 0.7082359191028809,
+      "learning_rate": 4.3112474393172055e-06,
+      "loss": 0.3473,
+      "step": 374
+    },
+    {
+      "epoch": 3.6407766990291264,
+      "grad_norm": 0.704495139070046,
+      "learning_rate": 4.307739774881878e-06,
+      "loss": 0.4346,
+      "step": 375
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 0.7500124236230435,
+      "learning_rate": 4.304224636008582e-06,
+      "loss": 0.2937,
+      "step": 376
+    },
+    {
+      "epoch": 3.6601941747572817,
+      "grad_norm": 0.7473030765489458,
+      "learning_rate": 4.300702037231318e-06,
+      "loss": 0.5837,
+      "step": 377
+    },
+    {
+      "epoch": 3.6699029126213594,
+      "grad_norm": 0.6924606297437474,
+      "learning_rate": 4.297171993114927e-06,
+      "loss": 0.2863,
+      "step": 378
+    },
+    {
+      "epoch": 3.679611650485437,
+      "grad_norm": 0.7683874913737846,
+      "learning_rate": 4.2936345182550365e-06,
+      "loss": 0.5933,
+      "step": 379
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.738704129251105,
+      "learning_rate": 4.290089627277998e-06,
+      "loss": 0.3695,
+      "step": 380
+    },
+    {
+      "epoch": 3.6990291262135924,
+      "grad_norm": 0.730122979591364,
+      "learning_rate": 4.286537334840825e-06,
+      "loss": 0.5314,
+      "step": 381
+    },
+    {
+      "epoch": 3.70873786407767,
+      "grad_norm": 0.8561211215862192,
+      "learning_rate": 4.2829776556311355e-06,
+      "loss": 0.4077,
+      "step": 382
+    },
+    {
+      "epoch": 3.7184466019417477,
+      "grad_norm": 0.8670647072409673,
+      "learning_rate": 4.279410604367088e-06,
+      "loss": 0.6157,
+      "step": 383
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 0.8439911656034701,
+      "learning_rate": 4.275836195797323e-06,
+      "loss": 0.5611,
+      "step": 384
+    },
+    {
+      "epoch": 3.737864077669903,
+      "grad_norm": 0.8111642418779228,
+      "learning_rate": 4.2722544447008995e-06,
+      "loss": 0.6242,
+      "step": 385
+    },
+    {
+      "epoch": 3.7475728155339807,
+      "grad_norm": 0.8234444876560004,
+      "learning_rate": 4.268665365887238e-06,
+      "loss": 0.626,
+      "step": 386
+    },
+    {
+      "epoch": 3.7572815533980584,
+      "grad_norm": 0.7499486492298939,
+      "learning_rate": 4.265068974196056e-06,
+      "loss": 0.3372,
+      "step": 387
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 0.8486583520088408,
+      "learning_rate": 4.261465284497307e-06,
+      "loss": 0.3682,
+      "step": 388
+    },
+    {
+      "epoch": 3.7766990291262137,
+      "grad_norm": 0.8004979648644168,
+      "learning_rate": 4.257854311691118e-06,
+      "loss": 0.5395,
+      "step": 389
+    },
+    {
+      "epoch": 3.7864077669902914,
+      "grad_norm": 0.7151044131463972,
+      "learning_rate": 4.254236070707734e-06,
+      "loss": 0.3921,
+      "step": 390
+    },
+    {
+      "epoch": 3.796116504854369,
+      "grad_norm": 0.6664683762506176,
+      "learning_rate": 4.250610576507445e-06,
+      "loss": 0.357,
+      "step": 391
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 0.8605859568498817,
+      "learning_rate": 4.246977844080537e-06,
+      "loss": 0.5304,
+      "step": 392
+    },
+    {
+      "epoch": 3.8155339805825244,
+      "grad_norm": 0.8012964647317763,
+      "learning_rate": 4.24333788844722e-06,
+      "loss": 0.4708,
+      "step": 393
+    },
+    {
+      "epoch": 3.825242718446602,
+      "grad_norm": 0.8135374110450408,
+      "learning_rate": 4.239690724657571e-06,
+      "loss": 0.4547,
+      "step": 394
+    },
+    {
+      "epoch": 3.8349514563106797,
+      "grad_norm": 1.1239630061318624,
+      "learning_rate": 4.236036367791471e-06,
+      "loss": 0.5222,
+      "step": 395
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 0.8761219541832255,
+      "learning_rate": 4.23237483295854e-06,
+      "loss": 0.584,
+      "step": 396
+    },
+    {
+      "epoch": 3.854368932038835,
+      "grad_norm": 0.7205334687878558,
+      "learning_rate": 4.228706135298081e-06,
+      "loss": 0.383,
+      "step": 397
+    },
+    {
+      "epoch": 3.8640776699029127,
+      "grad_norm": 0.778638883459709,
+      "learning_rate": 4.225030289979006e-06,
+      "loss": 0.3611,
+      "step": 398
+    },
+    {
+      "epoch": 3.8737864077669903,
+      "grad_norm": 0.775480657637448,
+      "learning_rate": 4.221347312199788e-06,
+      "loss": 0.3474,
+      "step": 399
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.9124181162931588,
+      "learning_rate": 4.2176572171883865e-06,
+      "loss": 0.5554,
+      "step": 400
+    },
+    {
+      "epoch": 3.8932038834951457,
+      "grad_norm": 0.7674226802327085,
+      "learning_rate": 4.213960020202187e-06,
+      "loss": 0.5775,
+      "step": 401
+    },
+    {
+      "epoch": 3.9029126213592233,
+      "grad_norm": 0.8374874478064829,
+      "learning_rate": 4.2102557365279435e-06,
+      "loss": 0.6738,
+      "step": 402
+    },
+    {
+      "epoch": 3.912621359223301,
+      "grad_norm": 0.789430401614793,
+      "learning_rate": 4.206544381481708e-06,
+      "loss": 0.5645,
+      "step": 403
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 0.8301680910378548,
+      "learning_rate": 4.202825970408772e-06,
+      "loss": 0.3362,
+      "step": 404
+    },
+    {
+      "epoch": 3.9320388349514563,
+      "grad_norm": 0.7434161207431128,
+      "learning_rate": 4.199100518683601e-06,
+      "loss": 0.3865,
+      "step": 405
+    },
+    {
+      "epoch": 3.941747572815534,
+      "grad_norm": 0.8023276174001709,
+      "learning_rate": 4.195368041709772e-06,
+      "loss": 0.557,
+      "step": 406
+    },
+    {
+      "epoch": 3.9514563106796117,
+      "grad_norm": 0.6897944866344006,
+      "learning_rate": 4.191628554919907e-06,
+      "loss": 0.4306,
+      "step": 407
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 0.7587292046885276,
+      "learning_rate": 4.187882073775615e-06,
+      "loss": 0.4791,
+      "step": 408
+    },
+    {
+      "epoch": 3.970873786407767,
+      "grad_norm": 0.7402776540544128,
+      "learning_rate": 4.184128613767422e-06,
+      "loss": 0.4587,
+      "step": 409
+    },
+    {
+      "epoch": 3.9805825242718447,
+      "grad_norm": 0.918829002078532,
+      "learning_rate": 4.18036819041471e-06,
+      "loss": 0.4834,
+      "step": 410
+    },
+    {
+      "epoch": 3.9902912621359223,
+      "grad_norm": 0.6951873113715458,
+      "learning_rate": 4.17660081926565e-06,
+      "loss": 0.3421,
+      "step": 411
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6863501093124053,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.3388,
+      "step": 412
+    },
+    {
+      "epoch": 4.009708737864078,
+      "grad_norm": 0.7114971026737045,
+      "learning_rate": 4.169045295914757e-06,
+      "loss": 0.5117,
+      "step": 413
+    },
+    {
+      "epoch": 4.019417475728155,
+      "grad_norm": 0.669639982894413,
+      "learning_rate": 4.165257174952647e-06,
+      "loss": 0.2925,
+      "step": 414
+    },
+    {
+      "epoch": 4.029126213592233,
+      "grad_norm": 0.7192571477195556,
+      "learning_rate": 4.161462168673508e-06,
+      "loss": 0.4189,
+      "step": 415
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 0.7186614597357849,
+      "learning_rate": 4.157660292768502e-06,
+      "loss": 0.5147,
+      "step": 416
+    },
+    {
+      "epoch": 4.048543689320389,
+      "grad_norm": 0.7615450499237213,
+      "learning_rate": 4.1538515629571985e-06,
+      "loss": 0.3839,
+      "step": 417
+    },
+    {
+      "epoch": 4.058252427184466,
+      "grad_norm": 0.7825364795958354,
+      "learning_rate": 4.1500359949875e-06,
+      "loss": 0.5142,
+      "step": 418
+    },
+    {
+      "epoch": 4.067961165048544,
+      "grad_norm": 0.7567301120413946,
+      "learning_rate": 4.1462136046355864e-06,
+      "loss": 0.3761,
+      "step": 419
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 0.7496382866591231,
+      "learning_rate": 4.142384407705846e-06,
+      "loss": 0.6578,
+      "step": 420
+    },
+    {
+      "epoch": 4.087378640776699,
+      "grad_norm": 0.6832493076636806,
+      "learning_rate": 4.138548420030808e-06,
+      "loss": 0.2888,
+      "step": 421
+    },
+    {
+      "epoch": 4.097087378640777,
+      "grad_norm": 0.8082055763247596,
+      "learning_rate": 4.13470565747108e-06,
+      "loss": 0.4464,
+      "step": 422
+    },
+    {
+      "epoch": 4.106796116504855,
+      "grad_norm": 0.6922107618500224,
+      "learning_rate": 4.130856135915282e-06,
+      "loss": 0.414,
+      "step": 423
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 0.8568390451993608,
+      "learning_rate": 4.126999871279982e-06,
+      "loss": 0.3524,
+      "step": 424
+    },
+    {
+      "epoch": 4.12621359223301,
+      "grad_norm": 0.7555701643934601,
+      "learning_rate": 4.123136879509626e-06,
+      "loss": 0.4294,
+      "step": 425
+    },
+    {
+      "epoch": 4.135922330097087,
+      "grad_norm": 0.7172542927002353,
+      "learning_rate": 4.119267176576475e-06,
+      "loss": 0.464,
+      "step": 426
+    },
+    {
+      "epoch": 4.145631067961165,
+      "grad_norm": 0.6682558575648311,
+      "learning_rate": 4.11539077848054e-06,
+      "loss": 0.4873,
+      "step": 427
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 0.8598473815632454,
+      "learning_rate": 4.111507701249513e-06,
+      "loss": 0.4819,
+      "step": 428
+    },
+    {
+      "epoch": 4.165048543689321,
+      "grad_norm": 0.9167731886550053,
+      "learning_rate": 4.107617960938702e-06,
+      "loss": 0.1954,
+      "step": 429
+    },
+    {
+      "epoch": 4.174757281553398,
+      "grad_norm": 0.8392679998978078,
+      "learning_rate": 4.103721573630965e-06,
+      "loss": 0.3716,
+      "step": 430
+    },
+    {
+      "epoch": 4.184466019417476,
+      "grad_norm": 0.7039689710893157,
+      "learning_rate": 4.099818555436645e-06,
+      "loss": 0.399,
+      "step": 431
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 0.8962803885867973,
+      "learning_rate": 4.095908922493499e-06,
+      "loss": 0.6665,
+      "step": 432
+    },
+    {
+      "epoch": 4.203883495145631,
+      "grad_norm": 1.919580266091257,
+      "learning_rate": 4.091992690966636e-06,
+      "loss": 0.2248,
+      "step": 433
+    },
+    {
+      "epoch": 4.213592233009709,
+      "grad_norm": 0.7602103224056724,
+      "learning_rate": 4.088069877048447e-06,
+      "loss": 0.3179,
+      "step": 434
+    },
+    {
+      "epoch": 4.223300970873787,
+      "grad_norm": 0.6467738319616958,
+      "learning_rate": 4.084140496958539e-06,
+      "loss": 0.2963,
+      "step": 435
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 0.7220900194096829,
+      "learning_rate": 4.080204566943668e-06,
+      "loss": 0.2475,
+      "step": 436
+    },
+    {
+      "epoch": 4.242718446601942,
+      "grad_norm": 1.189054603182352,
+      "learning_rate": 4.076262103277673e-06,
+      "loss": 0.2943,
+      "step": 437
+    },
+    {
+      "epoch": 4.252427184466019,
+      "grad_norm": 0.7169713980642173,
+      "learning_rate": 4.072313122261406e-06,
+      "loss": 0.6236,
+      "step": 438
+    },
+    {
+      "epoch": 4.262135922330097,
+      "grad_norm": 0.631689118975417,
+      "learning_rate": 4.068357640222668e-06,
+      "loss": 0.2852,
+      "step": 439
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 0.756930350000182,
+      "learning_rate": 4.06439567351614e-06,
+      "loss": 0.3246,
+      "step": 440
+    },
+    {
+      "epoch": 4.281553398058253,
+      "grad_norm": 0.6706899017753607,
+      "learning_rate": 4.0604272385233105e-06,
+      "loss": 0.3887,
+      "step": 441
+    },
+    {
+      "epoch": 4.29126213592233,
+      "grad_norm": 0.6775487389857546,
+      "learning_rate": 4.056452351652418e-06,
+      "loss": 0.2223,
+      "step": 442
+    },
+    {
+      "epoch": 4.300970873786408,
+      "grad_norm": 0.8187523228643651,
+      "learning_rate": 4.052471029338375e-06,
+      "loss": 0.4288,
+      "step": 443
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 0.6200886929947741,
+      "learning_rate": 4.048483288042703e-06,
+      "loss": 0.24,
+      "step": 444
+    },
+    {
+      "epoch": 4.320388349514563,
+      "grad_norm": 0.7372457408545906,
+      "learning_rate": 4.0444891442534615e-06,
+      "loss": 0.3116,
+      "step": 445
+    },
+    {
+      "epoch": 4.330097087378641,
+      "grad_norm": 0.7260528356835414,
+      "learning_rate": 4.040488614485187e-06,
+      "loss": 0.4287,
+      "step": 446
+    },
+    {
+      "epoch": 4.339805825242719,
+      "grad_norm": 0.8887283733391886,
+      "learning_rate": 4.036481715278818e-06,
+      "loss": 0.4366,
+      "step": 447
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 0.7417768044969674,
+      "learning_rate": 4.032468463201626e-06,
+      "loss": 0.4489,
+      "step": 448
+    },
+    {
+      "epoch": 4.359223300970874,
+      "grad_norm": 0.767167243913028,
+      "learning_rate": 4.028448874847152e-06,
+      "loss": 0.1889,
+      "step": 449
+    },
+    {
+      "epoch": 4.368932038834951,
+      "grad_norm": 0.7548738530081098,
+      "learning_rate": 4.024422966835137e-06,
+      "loss": 0.3836,
+      "step": 450
+    },
+    {
+      "epoch": 4.378640776699029,
+      "grad_norm": 0.7267980574762086,
+      "learning_rate": 4.0203907558114475e-06,
+      "loss": 0.3113,
+      "step": 451
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 0.7384964389802123,
+      "learning_rate": 4.016352258448016e-06,
+      "loss": 0.2558,
+      "step": 452
+    },
+    {
+      "epoch": 4.398058252427185,
+      "grad_norm": 0.7354848571749719,
+      "learning_rate": 4.0123074914427635e-06,
+      "loss": 0.2631,
+      "step": 453
+    },
+    {
+      "epoch": 4.407766990291262,
+      "grad_norm": 0.6121502281910712,
+      "learning_rate": 4.008256471519536e-06,
+      "loss": 0.2496,
+      "step": 454
+    },
+    {
+      "epoch": 4.41747572815534,
+      "grad_norm": 0.8179123108880375,
+      "learning_rate": 4.004199215428032e-06,
+      "loss": 0.3859,
+      "step": 455
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 0.9831209055765248,
+      "learning_rate": 4.000135739943735e-06,
+      "loss": 0.4455,
+      "step": 456
+    },
+    {
+      "epoch": 4.436893203883495,
+      "grad_norm": 0.8368630657095354,
+      "learning_rate": 3.996066061867844e-06,
+      "loss": 0.4183,
+      "step": 457
+    },
+    {
+      "epoch": 4.446601941747573,
+      "grad_norm": 0.7510366020026782,
+      "learning_rate": 3.991990198027203e-06,
+      "loss": 0.3243,
+      "step": 458
+    },
+    {
+      "epoch": 4.456310679611651,
+      "grad_norm": 0.7739088378814717,
+      "learning_rate": 3.987908165274233e-06,
+      "loss": 0.3937,
+      "step": 459
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 0.7451064795923457,
+      "learning_rate": 3.9838199804868635e-06,
+      "loss": 0.4762,
+      "step": 460
+    },
+    {
+      "epoch": 4.475728155339806,
+      "grad_norm": 0.8005653610588523,
+      "learning_rate": 3.979725660568456e-06,
+      "loss": 0.4203,
+      "step": 461
+    },
+    {
+      "epoch": 4.485436893203883,
+      "grad_norm": 0.7571415862327683,
+      "learning_rate": 3.975625222447742e-06,
+      "loss": 0.2966,
+      "step": 462
+    },
+    {
+      "epoch": 4.495145631067961,
+      "grad_norm": 0.9407682824067278,
+      "learning_rate": 3.97151868307875e-06,
+      "loss": 0.4631,
+      "step": 463
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 0.8280322987446443,
+      "learning_rate": 3.9674060594407345e-06,
+      "loss": 0.3439,
+      "step": 464
+    },
+    {
+      "epoch": 4.514563106796117,
+      "grad_norm": 0.7056881797171722,
+      "learning_rate": 3.963287368538105e-06,
+      "loss": 0.2049,
+      "step": 465
+    },
+    {
+      "epoch": 4.524271844660194,
+      "grad_norm": 0.9332990417689767,
+      "learning_rate": 3.959162627400361e-06,
+      "loss": 0.4755,
+      "step": 466
+    },
+    {
+      "epoch": 4.533980582524272,
+      "grad_norm": 1.1403012923588804,
+      "learning_rate": 3.9550318530820145e-06,
+      "loss": 0.4185,
+      "step": 467
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 0.689075963184153,
+      "learning_rate": 3.9508950626625244e-06,
+      "loss": 0.2702,
+      "step": 468
+    },
+    {
+      "epoch": 4.553398058252427,
+      "grad_norm": 0.8020665017742961,
+      "learning_rate": 3.946752273246224e-06,
+      "loss": 0.4359,
+      "step": 469
+    },
+    {
+      "epoch": 4.563106796116505,
+      "grad_norm": 0.8212257599019372,
+      "learning_rate": 3.942603501962249e-06,
+      "loss": 0.352,
+      "step": 470
+    },
+    {
+      "epoch": 4.572815533980583,
+      "grad_norm": 0.7742300518258042,
+      "learning_rate": 3.9384487659644716e-06,
+      "loss": 0.2489,
+      "step": 471
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 0.7529500735779854,
+      "learning_rate": 3.934288082431423e-06,
+      "loss": 0.4064,
+      "step": 472
+    },
+    {
+      "epoch": 4.592233009708738,
+      "grad_norm": 0.7285209333659659,
+      "learning_rate": 3.930121468566227e-06,
+      "loss": 0.4903,
+      "step": 473
+    },
+    {
+      "epoch": 4.601941747572815,
+      "grad_norm": 0.702088268251986,
+      "learning_rate": 3.925948941596528e-06,
+      "loss": 0.3524,
+      "step": 474
+    },
+    {
+      "epoch": 4.611650485436893,
+      "grad_norm": 0.6123252794922119,
+      "learning_rate": 3.92177051877442e-06,
+      "loss": 0.291,
+      "step": 475
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 0.743156646498327,
+      "learning_rate": 3.917586217376369e-06,
+      "loss": 0.435,
+      "step": 476
+    },
+    {
+      "epoch": 4.631067961165049,
+      "grad_norm": 0.704106227231604,
+      "learning_rate": 3.913396054703155e-06,
+      "loss": 0.4408,
+      "step": 477
+    },
+    {
+      "epoch": 4.640776699029126,
+      "grad_norm": 0.7127662571747935,
+      "learning_rate": 3.909200048079786e-06,
+      "loss": 0.4177,
+      "step": 478
+    },
+    {
+      "epoch": 4.650485436893204,
+      "grad_norm": 0.748418825692663,
+      "learning_rate": 3.9049982148554384e-06,
+      "loss": 0.5271,
+      "step": 479
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 1.0250685486452722,
+      "learning_rate": 3.900790572403376e-06,
+      "loss": 0.3461,
+      "step": 480
+    },
+    {
+      "epoch": 4.669902912621359,
+      "grad_norm": 0.7829872249544201,
+      "learning_rate": 3.896577138120881e-06,
+      "loss": 0.4164,
+      "step": 481
+    },
+    {
+      "epoch": 4.679611650485437,
+      "grad_norm": 0.9232004605724133,
+      "learning_rate": 3.892357929429187e-06,
+      "loss": 0.3579,
+      "step": 482
+    },
+    {
+      "epoch": 4.689320388349515,
+      "grad_norm": 0.784529648802751,
+      "learning_rate": 3.8881329637734e-06,
+      "loss": 0.2236,
+      "step": 483
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 0.7262690017605888,
+      "learning_rate": 3.883902258622431e-06,
+      "loss": 0.3756,
+      "step": 484
+    },
+    {
+      "epoch": 4.70873786407767,
+      "grad_norm": 0.8102856355194789,
+      "learning_rate": 3.8796658314689205e-06,
+      "loss": 0.4178,
+      "step": 485
+    },
+    {
+      "epoch": 4.718446601941747,
+      "grad_norm": 0.7486005625149885,
+      "learning_rate": 3.875423699829168e-06,
+      "loss": 0.2835,
+      "step": 486
+    },
+    {
+      "epoch": 4.728155339805825,
+      "grad_norm": 0.7379335654656654,
+      "learning_rate": 3.871175881243061e-06,
+      "loss": 0.3489,
+      "step": 487
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 0.8104226180405947,
+      "learning_rate": 3.866922393273999e-06,
+      "loss": 0.4974,
+      "step": 488
+    },
+    {
+      "epoch": 4.747572815533981,
+      "grad_norm": 0.7845427197820135,
+      "learning_rate": 3.862663253508822e-06,
+      "loss": 0.3453,
+      "step": 489
+    },
+    {
+      "epoch": 4.757281553398058,
+      "grad_norm": 0.7436234031119154,
+      "learning_rate": 3.858398479557739e-06,
+      "loss": 0.3273,
+      "step": 490
+    },
+    {
+      "epoch": 4.766990291262136,
+      "grad_norm": 2.217227869082823,
+      "learning_rate": 3.8541280890542565e-06,
+      "loss": 0.3229,
+      "step": 491
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 1.0428680689060519,
+      "learning_rate": 3.849852099655102e-06,
+      "loss": 0.4476,
+      "step": 492
+    },
+    {
+      "epoch": 4.786407766990291,
+      "grad_norm": 0.7093952303779135,
+      "learning_rate": 3.845570529040151e-06,
+      "loss": 0.3531,
+      "step": 493
+    },
+    {
+      "epoch": 4.796116504854369,
+      "grad_norm": 0.7919634533614799,
+      "learning_rate": 3.841283394912361e-06,
+      "loss": 0.3435,
+      "step": 494
+    },
+    {
+      "epoch": 4.805825242718447,
+      "grad_norm": 0.7136080403274071,
+      "learning_rate": 3.836990714997686e-06,
+      "loss": 0.4444,
+      "step": 495
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 0.8376006583714606,
+      "learning_rate": 3.832692507045015e-06,
+      "loss": 0.4478,
+      "step": 496
+    },
+    {
+      "epoch": 4.825242718446602,
+      "grad_norm": 0.6616123052279387,
+      "learning_rate": 3.828388788826091e-06,
+      "loss": 0.5166,
+      "step": 497
+    },
+    {
+      "epoch": 4.834951456310679,
+      "grad_norm": 0.7483368464194841,
+      "learning_rate": 3.824079578135442e-06,
+      "loss": 0.4151,
+      "step": 498
+    },
+    {
+      "epoch": 4.844660194174757,
+      "grad_norm": 0.646917604155803,
+      "learning_rate": 3.819764892790307e-06,
+      "loss": 0.4058,
+      "step": 499
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 0.8267691480427513,
+      "learning_rate": 3.815444750630555e-06,
+      "loss": 0.3406,
+      "step": 500
+    },
+    {
+      "epoch": 4.864077669902913,
+      "grad_norm": 0.8928434754303194,
+      "learning_rate": 3.811119169518624e-06,
+      "loss": 0.541,
+      "step": 501
+    },
+    {
+      "epoch": 4.87378640776699,
+      "grad_norm": 0.9334290516881728,
+      "learning_rate": 3.8067881673394363e-06,
+      "loss": 0.6994,
+      "step": 502
+    },
+    {
+      "epoch": 4.883495145631068,
+      "grad_norm": 0.7272706791055442,
+      "learning_rate": 3.802451762000331e-06,
+      "loss": 0.3112,
+      "step": 503
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 0.6753535996733946,
+      "learning_rate": 3.7981099714309856e-06,
+      "loss": 0.2749,
+      "step": 504
+    },
+    {
+      "epoch": 4.902912621359223,
+      "grad_norm": 0.8446207256750283,
+      "learning_rate": 3.7937628135833453e-06,
+      "loss": 0.354,
+      "step": 505
+    },
+    {
+      "epoch": 4.9126213592233015,
+      "grad_norm": 0.8427902134243664,
+      "learning_rate": 3.7894103064315463e-06,
+      "loss": 0.3765,
+      "step": 506
+    },
+    {
+      "epoch": 4.922330097087379,
+      "grad_norm": 0.7801637629685254,
+      "learning_rate": 3.7850524679718424e-06,
+      "loss": 0.3014,
+      "step": 507
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 0.7448345484733,
+      "learning_rate": 3.7806893162225328e-06,
+      "loss": 0.2862,
+      "step": 508
+    },
+    {
+      "epoch": 4.941747572815534,
+      "grad_norm": 0.7506722271057215,
+      "learning_rate": 3.7763208692238818e-06,
+      "loss": 0.4005,
+      "step": 509
+    },
+    {
+      "epoch": 4.951456310679612,
+      "grad_norm": 0.7507341810173735,
+      "learning_rate": 3.7719471450380518e-06,
+      "loss": 0.3154,
+      "step": 510
+    },
+    {
+      "epoch": 4.961165048543689,
+      "grad_norm": 0.7509927539126146,
+      "learning_rate": 3.7675681617490212e-06,
+      "loss": 0.5283,
+      "step": 511
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 0.7780851180094496,
+      "learning_rate": 3.7631839374625167e-06,
+      "loss": 0.5371,
+      "step": 512
+    },
+    {
+      "epoch": 4.980582524271845,
+      "grad_norm": 0.7269269443801724,
+      "learning_rate": 3.758794490305932e-06,
+      "loss": 0.3471,
+      "step": 513
+    },
+    {
+      "epoch": 4.990291262135923,
+      "grad_norm": 0.7023346099879195,
+      "learning_rate": 3.7543998384282565e-06,
+      "loss": 0.3844,
+      "step": 514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.6668852497704912,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4468,
+      "step": 515
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 57257695182848.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

limo/full/checkpoint-515/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cfd645031663b96977d94be89e3ef836f67c72ab2e36456032dd786e721780e
+size 7416

limo/full/checkpoint-515/zero_to_fp32.py

@@ -0,0 +1,674 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import json
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering Sharded Weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        state_dict_split = split_torch_state_dict_into_shards(state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard, output_path)
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)