Upload folder using huggingface_hub

.gitattributes

@@ -38,3 +38,4 @@ s1k/full/checkpoint-930/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 limo/full/checkpoint-1030/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 limo/full/checkpoint-1545/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 s1k/full/checkpoint-310/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+limo/full/checkpoint-515/tokenizer.json filter=lfs diff=lfs merge=lfs -text

limo/full/checkpoint-515/config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "/mnt/public/data/lh/models/DeepSeek-R1-Distill-Qwen-7B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 131072,
+  "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,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

limo/full/checkpoint-515/generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 151646,
+  "do_sample": true,
+  "eos_token_id": 151643,
+  "temperature": 0.6,
+  "top_p": 0.95,
+  "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 sha256:1afcc0e4c15d5de35bd71d4288e5ed759ead0c41550cfe323da52413ae6ac824
+size 11423429568

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:a37033f432949fdf7b9bec4f35ddbb3e67ce5ee3ab5b301084e2fc95f46cab97
+size 11423429568

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 sha256:313791de2f39491e310cba3e261a00ed48524de8e4444be9fcfd80d29e419b7a
+size 11423429568

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:c92d14685b37d08fcf7ae7d589f26da970efd86d48f69daf7924960d1ff80596
+size 11423429568

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:bc68cda34df5e9c6312fc72b7b1db88405312a30d476f61834f644c2f2c6d858
+size 11423429568

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 sha256:1146e8c9707853ee5ff2607a9b0f9fd927a29b97b9d18df2cd771b79ef1c6988
+size 11423429568

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:d019c29f6d06f93f7f749958fe76f1103cf59a464f0763ea1f55cb8d0531459d
+size 11423429568

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 sha256:1259e832e4a9ae45831c58815a3e00c97216756927aec7b2c5f476cf85b1f087
+size 11423429568

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:bf286f41254d5b215a2532c2426dd8faadd8233dfcd74064cd070985e724a9ad
+size 168405

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:e932638b3b1f388f852f042f0d99fa6e53523c710b6e97703cb6f78e3ca43992
+size 168405

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 sha256:098d43c3604043c8f88e47b19c10a6da09179bb50356a92f426ebf3b282f404f
+size 168405

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 sha256:96cdda6725cf780ef3264ae44d4f9e52afffba48f21c85d5e424da60789d6b2f
+size 168405

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 sha256:67aab74410b3f61e677b2edcd3c04d047f2a4c19f198e4dfc59cf0925c30113d
+size 168405

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 sha256:012cdcc9074166ec431530b3b08621e2ad7ebcbb326a88e8ec38f45013061c1c
+size 168405

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:bc8fafd9e82d3ea0ad125ebbf1cc5b72dab33df3a0dd8bace9c22321eb4fa9a8
+size 168405

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:bf24bcb2ce59c1d44ead1260362eff03bc072fe9a3dd30043ba2cd4d8157b9f8
+size 168405

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 sha256:116c08decc31cc1ebbcc992e2375f1728d749f446355db9b2742e8de071beb04
+size 4877660776

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70628c87aa74d0b5660171596809680e30bafc4907853d9b0c279077ea20d857
+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:ba0d0234ed7781e0f1db5634192355e9ddc4d128569692788e95723f87f4ea4d
+size 4330865200

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a111fd0d196f762bf543003c951dde0f7f8dd20e3b1b0fbacceb1c2c0e52d95
+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 +1,32 @@
+{
+  "additional_special_tokens": [
+    {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
+  "bos_token": {
+    "content": "<｜begin▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "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 sha256:02643f00207dfc5ed248992486bde04314c21dca556bf65ce520690962b8db63
+size 11422965

limo/full/checkpoint-515/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<｜end▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<｜User｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151645": {
+      "content": "<｜Assistant｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151646": {
+      "content": "<｜begin▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|EOT|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151648": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151649": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "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
+    },
+    "151665": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_end|>"
+  ],
+  "bos_token": "<｜begin▁of▁sentence｜>",
+  "chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set ns = namespace(is_first=false, is_tool=false, is_output_first=true, system_prompt='') %}{%- for message in messages %}{%- if message['role'] == 'system' %}{% set ns.system_prompt = message['content'] %}{%- endif %}{%- endfor %}{{bos_token}}{{ns.system_prompt}}{%- for message in messages %}{%- if message['role'] == 'user' %}{%- set ns.is_tool = false -%}{{'<｜User｜>' + message['content']}}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is none %}{%- set ns.is_tool = false -%}{%- for tool in message['tool_calls']%}{%- if not ns.is_first %}{{'<｜Assistant｜><｜tool▁calls▁begin｜><｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\\n' + '```json' + '\\n' + tool['function']['arguments'] + '\\n' + '```' + '<｜tool▁call▁end｜>'}}{%- set ns.is_first = true -%}{%- else %}{{'\\n' + '<｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\\n' + '```json' + '\\n' + tool['function']['arguments'] + '\\n' + '```' + '<｜tool▁call▁end｜>'}}{{'<｜tool▁calls▁end｜><｜end▁of▁sentence｜>'}}{%- endif %}{%- endfor %}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is not none %}{%- if ns.is_tool %}{{'<｜tool▁outputs▁end｜>' + message['content'] + '<｜end▁of▁sentence｜>'}}{%- set ns.is_tool = false -%}{%- else %}{% set content = message['content'] %}{% if '</think>' in content %}{% set content = content.split('</think>')[-1] %}{% endif %}{{'<｜Assistant｜>' + content + '<｜end▁of▁sentence｜>'}}{%- endif %}{%- endif %}{%- if message['role'] == 'tool' %}{%- set ns.is_tool = true -%}{%- if ns.is_output_first %}{{'<｜tool▁outputs▁begin｜><｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- set ns.is_output_first = false %}{%- else %}{{'\\n<｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- endif %}{%- endif %}{%- endfor -%}{% if ns.is_tool %}{{'<｜tool▁outputs▁end｜>'}}{% endif %}{% if add_generation_prompt and not ns.is_tool %}{{'<｜Assistant｜>'}}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<｜end▁of▁sentence｜>",
+  "legacy": true,
+  "model_max_length": 16384,
+  "pad_token": "<｜end▁of▁sentence｜>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "split_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": null,
+  "use_default_system_prompt": false
+}

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": 2.750988483428955,
+      "learning_rate": 4.999994831641374e-06,
+      "loss": 0.5337,
+      "step": 1
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 2.219198703765869,
+      "learning_rate": 4.9999793265868636e-06,
+      "loss": 0.3878,
+      "step": 2
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 2.154571533203125,
+      "learning_rate": 4.999953484900578e-06,
+      "loss": 0.5907,
+      "step": 3
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 2.1270041465759277,
+      "learning_rate": 4.9999173066893655e-06,
+      "loss": 0.4606,
+      "step": 4
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 1.8446154594421387,
+      "learning_rate": 4.9998707921028104e-06,
+      "loss": 0.5342,
+      "step": 5
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 1.504712462425232,
+      "learning_rate": 4.999813941333237e-06,
+      "loss": 0.3683,
+      "step": 6
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 1.5787030458450317,
+      "learning_rate": 4.999746754615704e-06,
+      "loss": 0.5031,
+      "step": 7
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 0.9261472225189209,
+      "learning_rate": 4.9996692322280085e-06,
+      "loss": 0.4372,
+      "step": 8
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 0.7603314518928528,
+      "learning_rate": 4.999581374490681e-06,
+      "loss": 0.45,
+      "step": 9
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 1.050328254699707,
+      "learning_rate": 4.999483181766986e-06,
+      "loss": 0.6373,
+      "step": 10
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 1.1623330116271973,
+      "learning_rate": 4.999374654462919e-06,
+      "loss": 0.5608,
+      "step": 11
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 0.9819350838661194,
+      "learning_rate": 4.999255793027207e-06,
+      "loss": 0.5828,
+      "step": 12
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 0.7598986625671387,
+      "learning_rate": 4.999126597951305e-06,
+      "loss": 0.3634,
+      "step": 13
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 0.7111096382141113,
+      "learning_rate": 4.998987069769394e-06,
+      "loss": 0.3395,
+      "step": 14
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 0.9073492288589478,
+      "learning_rate": 4.998837209058379e-06,
+      "loss": 0.5404,
+      "step": 15
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 1.2447338104248047,
+      "learning_rate": 4.998677016437888e-06,
+      "loss": 0.4153,
+      "step": 16
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 0.5936450958251953,
+      "learning_rate": 4.998506492570266e-06,
+      "loss": 0.3405,
+      "step": 17
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 1.0477153062820435,
+      "learning_rate": 4.998325638160576e-06,
+      "loss": 0.6277,
+      "step": 18
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 0.9470447301864624,
+      "learning_rate": 4.998134453956596e-06,
+      "loss": 0.4055,
+      "step": 19
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.2609833478927612,
+      "learning_rate": 4.997932940748811e-06,
+      "loss": 0.5094,
+      "step": 20
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 1.033146619796753,
+      "learning_rate": 4.997721099370416e-06,
+      "loss": 0.4799,
+      "step": 21
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 0.9227539300918579,
+      "learning_rate": 4.997498930697308e-06,
+      "loss": 0.4716,
+      "step": 22
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 0.8762221336364746,
+      "learning_rate": 4.997266435648086e-06,
+      "loss": 0.4087,
+      "step": 23
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.0111958980560303,
+      "learning_rate": 4.997023615184044e-06,
+      "loss": 0.6779,
+      "step": 24
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 0.7958911657333374,
+      "learning_rate": 4.996770470309167e-06,
+      "loss": 0.6274,
+      "step": 25
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 0.7777181267738342,
+      "learning_rate": 4.996507002070131e-06,
+      "loss": 0.4815,
+      "step": 26
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 0.7622829675674438,
+      "learning_rate": 4.996233211556295e-06,
+      "loss": 0.5668,
+      "step": 27
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 0.7621707320213318,
+      "learning_rate": 4.9959490998996974e-06,
+      "loss": 0.3447,
+      "step": 28
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 0.6936230659484863,
+      "learning_rate": 4.995654668275049e-06,
+      "loss": 0.4305,
+      "step": 29
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.6379275918006897,
+      "learning_rate": 4.995349917899735e-06,
+      "loss": 0.3779,
+      "step": 30
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 0.714606761932373,
+      "learning_rate": 4.9950348500338005e-06,
+      "loss": 0.3723,
+      "step": 31
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 0.8575965166091919,
+      "learning_rate": 4.994709465979954e-06,
+      "loss": 0.5068,
+      "step": 32
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 0.8307704925537109,
+      "learning_rate": 4.994373767083556e-06,
+      "loss": 0.4475,
+      "step": 33
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 0.978870689868927,
+      "learning_rate": 4.994027754732616e-06,
+      "loss": 0.5223,
+      "step": 34
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.6880773305892944,
+      "learning_rate": 4.993671430357788e-06,
+      "loss": 0.379,
+      "step": 35
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 0.7792438268661499,
+      "learning_rate": 4.99330479543236e-06,
+      "loss": 0.4358,
+      "step": 36
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 0.7110076546669006,
+      "learning_rate": 4.992927851472254e-06,
+      "loss": 0.5785,
+      "step": 37
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 0.6907165050506592,
+      "learning_rate": 4.992540600036014e-06,
+      "loss": 0.6123,
+      "step": 38
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 0.6032484769821167,
+      "learning_rate": 4.992143042724805e-06,
+      "loss": 0.4347,
+      "step": 39
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.9119201302528381,
+      "learning_rate": 4.991735181182401e-06,
+      "loss": 0.7102,
+      "step": 40
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 0.7313927412033081,
+      "learning_rate": 4.991317017095182e-06,
+      "loss": 0.4537,
+      "step": 41
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 0.6860741376876831,
+      "learning_rate": 4.990888552192126e-06,
+      "loss": 0.4216,
+      "step": 42
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 0.6899588704109192,
+      "learning_rate": 4.9904497882448004e-06,
+      "loss": 0.5287,
+      "step": 43
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 0.6557210087776184,
+      "learning_rate": 4.990000727067357e-06,
+      "loss": 0.4969,
+      "step": 44
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.6218708157539368,
+      "learning_rate": 4.989541370516523e-06,
+      "loss": 0.4038,
+      "step": 45
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 0.5744000673294067,
+      "learning_rate": 4.989071720491595e-06,
+      "loss": 0.3772,
+      "step": 46
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.6886637210845947,
+      "learning_rate": 4.988591778934428e-06,
+      "loss": 0.3736,
+      "step": 47
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.330315113067627,
+      "learning_rate": 4.9881015478294294e-06,
+      "loss": 0.4809,
+      "step": 48
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 0.6803296208381653,
+      "learning_rate": 4.987601029203553e-06,
+      "loss": 0.3859,
+      "step": 49
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.6322437524795532,
+      "learning_rate": 4.987090225126285e-06,
+      "loss": 0.5004,
+      "step": 50
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 0.6261456608772278,
+      "learning_rate": 4.98656913770964e-06,
+      "loss": 0.4145,
+      "step": 51
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 0.6072115302085876,
+      "learning_rate": 4.986037769108154e-06,
+      "loss": 0.4377,
+      "step": 52
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 0.8084361553192139,
+      "learning_rate": 4.9854961215188676e-06,
+      "loss": 0.6331,
+      "step": 53
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.6779961585998535,
+      "learning_rate": 4.984944197181324e-06,
+      "loss": 0.3618,
+      "step": 54
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.669882595539093,
+      "learning_rate": 4.9843819983775575e-06,
+      "loss": 0.4005,
+      "step": 55
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 0.7368363738059998,
+      "learning_rate": 4.983809527432086e-06,
+      "loss": 0.3951,
+      "step": 56
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 0.7703456878662109,
+      "learning_rate": 4.983226786711895e-06,
+      "loss": 0.4821,
+      "step": 57
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.6797823905944824,
+      "learning_rate": 4.982633778626437e-06,
+      "loss": 0.4389,
+      "step": 58
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 0.5815659761428833,
+      "learning_rate": 4.982030505627613e-06,
+      "loss": 0.5409,
+      "step": 59
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.766441285610199,
+      "learning_rate": 4.98141697020977e-06,
+      "loss": 0.4977,
+      "step": 60
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 0.6255026459693909,
+      "learning_rate": 4.9807931749096836e-06,
+      "loss": 0.3994,
+      "step": 61
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.7539398670196533,
+      "learning_rate": 4.980159122306551e-06,
+      "loss": 0.5768,
+      "step": 62
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.7843878269195557,
+      "learning_rate": 4.979514815021984e-06,
+      "loss": 0.6804,
+      "step": 63
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 0.7728163599967957,
+      "learning_rate": 4.978860255719989e-06,
+      "loss": 0.4552,
+      "step": 64
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.7498195171356201,
+      "learning_rate": 4.978195447106965e-06,
+      "loss": 0.5952,
+      "step": 65
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.7298671007156372,
+      "learning_rate": 4.9775203919316864e-06,
+      "loss": 0.4846,
+      "step": 66
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 1.0157232284545898,
+      "learning_rate": 4.976835092985297e-06,
+      "loss": 0.615,
+      "step": 67
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.6613374352455139,
+      "learning_rate": 4.976139553101291e-06,
+      "loss": 0.518,
+      "step": 68
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.754895806312561,
+      "learning_rate": 4.975433775155509e-06,
+      "loss": 0.4902,
+      "step": 69
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.6307485699653625,
+      "learning_rate": 4.974717762066123e-06,
+      "loss": 0.4122,
+      "step": 70
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 0.866858959197998,
+      "learning_rate": 4.973991516793621e-06,
+      "loss": 0.6873,
+      "step": 71
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.7766331434249878,
+      "learning_rate": 4.973255042340801e-06,
+      "loss": 0.4722,
+      "step": 72
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 0.8586907386779785,
+      "learning_rate": 4.972508341752754e-06,
+      "loss": 0.6483,
+      "step": 73
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.7376808524131775,
+      "learning_rate": 4.9717514181168534e-06,
+      "loss": 0.4555,
+      "step": 74
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.5241130590438843,
+      "learning_rate": 4.970984274562741e-06,
+      "loss": 0.3603,
+      "step": 75
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.7115001082420349,
+      "learning_rate": 4.970206914262315e-06,
+      "loss": 0.5182,
+      "step": 76
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.688653826713562,
+      "learning_rate": 4.969419340429717e-06,
+      "loss": 0.4654,
+      "step": 77
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 0.5328500866889954,
+      "learning_rate": 4.968621556321319e-06,
+      "loss": 0.4655,
+      "step": 78
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 1.0476230382919312,
+      "learning_rate": 4.967813565235708e-06,
+      "loss": 0.5689,
+      "step": 79
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.6169328689575195,
+      "learning_rate": 4.966995370513675e-06,
+      "loss": 0.4068,
+      "step": 80
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.8216828107833862,
+      "learning_rate": 4.966166975538197e-06,
+      "loss": 0.68,
+      "step": 81
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.6563589572906494,
+      "learning_rate": 4.965328383734429e-06,
+      "loss": 0.4559,
+      "step": 82
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 0.8646124005317688,
+      "learning_rate": 4.964479598569686e-06,
+      "loss": 0.6455,
+      "step": 83
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.9952420592308044,
+      "learning_rate": 4.963620623553428e-06,
+      "loss": 0.5929,
+      "step": 84
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.6527711749076843,
+      "learning_rate": 4.962751462237248e-06,
+      "loss": 0.4333,
+      "step": 85
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.7421233057975769,
+      "learning_rate": 4.9618721182148564e-06,
+      "loss": 0.3732,
+      "step": 86
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.7303851246833801,
+      "learning_rate": 4.960982595122064e-06,
+      "loss": 0.4007,
+      "step": 87
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.6574289798736572,
+      "learning_rate": 4.960082896636773e-06,
+      "loss": 0.5971,
+      "step": 88
+    },
+    {
+      "epoch": 0.8640776699029126,
+      "grad_norm": 0.7417539954185486,
+      "learning_rate": 4.959173026478952e-06,
+      "loss": 0.4952,
+      "step": 89
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.6207308769226074,
+      "learning_rate": 4.958252988410631e-06,
+      "loss": 0.3605,
+      "step": 90
+    },
+    {
+      "epoch": 0.883495145631068,
+      "grad_norm": 1.2858132123947144,
+      "learning_rate": 4.9573227862358794e-06,
+      "loss": 0.5067,
+      "step": 91
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.7016294002532959,
+      "learning_rate": 4.956382423800791e-06,
+      "loss": 0.5049,
+      "step": 92
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.6535316109657288,
+      "learning_rate": 4.955431904993471e-06,
+      "loss": 0.5478,
+      "step": 93
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 0.581110417842865,
+      "learning_rate": 4.954471233744015e-06,
+      "loss": 0.3442,
+      "step": 94
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.7974981069564819,
+      "learning_rate": 4.9535004140245005e-06,
+      "loss": 0.5455,
+      "step": 95
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 0.6826483011245728,
+      "learning_rate": 4.952519449848962e-06,
+      "loss": 0.471,
+      "step": 96
+    },
+    {
+      "epoch": 0.941747572815534,
+      "grad_norm": 0.6755044460296631,
+      "learning_rate": 4.951528345273379e-06,
+      "loss": 0.4695,
+      "step": 97
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.7678257822990417,
+      "learning_rate": 4.950527104395659e-06,
+      "loss": 0.583,
+      "step": 98
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 0.7510711550712585,
+      "learning_rate": 4.9495157313556185e-06,
+      "loss": 0.3671,
+      "step": 99
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.5573858618736267,
+      "learning_rate": 4.94849423033497e-06,
+      "loss": 0.2934,
+      "step": 100
+    },
+    {
+      "epoch": 0.9805825242718447,
+      "grad_norm": 0.7562112212181091,
+      "learning_rate": 4.9474626055573e-06,
+      "loss": 0.6547,
+      "step": 101
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.5529704689979553,
+      "learning_rate": 4.946420861288051e-06,
+      "loss": 0.47,
+      "step": 102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9349260330200195,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.4291,
+      "step": 103
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 0.5343674421310425,
+      "learning_rate": 4.944307031545797e-06,
+      "loss": 0.3439,
+      "step": 104
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.9405943155288696,
+      "learning_rate": 4.943234954812812e-06,
+      "loss": 0.5853,
+      "step": 105
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 0.6383285522460938,
+      "learning_rate": 4.942152776068264e-06,
+      "loss": 0.3842,
+      "step": 106
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 0.859075665473938,
+      "learning_rate": 4.941060499786622e-06,
+      "loss": 0.5976,
+      "step": 107
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 0.7414010167121887,
+      "learning_rate": 4.939958130484106e-06,
+      "loss": 0.4295,
+      "step": 108
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.5165299773216248,
+      "learning_rate": 4.938845672718668e-06,
+      "loss": 0.4116,
+      "step": 109
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.6870267987251282,
+      "learning_rate": 4.937723131089974e-06,
+      "loss": 0.4893,
+      "step": 110
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 0.7484614253044128,
+      "learning_rate": 4.93659051023938e-06,
+      "loss": 0.4745,
+      "step": 111
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.4287586808204651,
+      "learning_rate": 4.93544781484992e-06,
+      "loss": 0.2543,
+      "step": 112
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.7693856954574585,
+      "learning_rate": 4.9342950496462815e-06,
+      "loss": 0.4679,
+      "step": 113
+    },
+    {
+      "epoch": 1.1067961165048543,
+      "grad_norm": 0.7279606461524963,
+      "learning_rate": 4.933132219394786e-06,
+      "loss": 0.468,
+      "step": 114
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.787359893321991,
+      "learning_rate": 4.931959328903376e-06,
+      "loss": 0.5892,
+      "step": 115
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 0.6111868023872375,
+      "learning_rate": 4.930776383021584e-06,
+      "loss": 0.4062,
+      "step": 116
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 0.6948837637901306,
+      "learning_rate": 4.92958338664052e-06,
+      "loss": 0.3093,
+      "step": 117
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.45196858048439026,
+      "learning_rate": 4.928380344692853e-06,
+      "loss": 0.3166,
+      "step": 118
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.6264169812202454,
+      "learning_rate": 4.927167262152784e-06,
+      "loss": 0.4513,
+      "step": 119
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.5587222576141357,
+      "learning_rate": 4.925944144036027e-06,
+      "loss": 0.4497,
+      "step": 120
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.6254664063453674,
+      "learning_rate": 4.924710995399796e-06,
+      "loss": 0.4199,
+      "step": 121
+    },
+    {
+      "epoch": 1.1844660194174756,
+      "grad_norm": 0.5662532448768616,
+      "learning_rate": 4.923467821342773e-06,
+      "loss": 0.4158,
+      "step": 122
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.6218069791793823,
+      "learning_rate": 4.922214627005092e-06,
+      "loss": 0.4801,
+      "step": 123
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.7494342923164368,
+      "learning_rate": 4.920951417568323e-06,
+      "loss": 0.5588,
+      "step": 124
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.5926786661148071,
+      "learning_rate": 4.919678198255438e-06,
+      "loss": 0.3545,
+      "step": 125
+    },
+    {
+      "epoch": 1.2233009708737863,
+      "grad_norm": 0.5640999674797058,
+      "learning_rate": 4.918394974330801e-06,
+      "loss": 0.4607,
+      "step": 126
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.52383953332901,
+      "learning_rate": 4.917101751100142e-06,
+      "loss": 0.4278,
+      "step": 127
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.7073279023170471,
+      "learning_rate": 4.915798533910534e-06,
+      "loss": 0.4067,
+      "step": 128
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.6415435075759888,
+      "learning_rate": 4.9144853281503715e-06,
+      "loss": 0.3647,
+      "step": 129
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.869964063167572,
+      "learning_rate": 4.91316213924935e-06,
+      "loss": 0.5479,
+      "step": 130
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.5213742256164551,
+      "learning_rate": 4.911828972678441e-06,
+      "loss": 0.3092,
+      "step": 131
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.5948150157928467,
+      "learning_rate": 4.91048583394987e-06,
+      "loss": 0.3913,
+      "step": 132
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.695085346698761,
+      "learning_rate": 4.909132728617095e-06,
+      "loss": 0.666,
+      "step": 133
+    },
+    {
+      "epoch": 1.3009708737864076,
+      "grad_norm": 0.609943687915802,
+      "learning_rate": 4.907769662274785e-06,
+      "loss": 0.3837,
+      "step": 134
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.5892361998558044,
+      "learning_rate": 4.90639664055879e-06,
+      "loss": 0.3312,
+      "step": 135
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.5192832350730896,
+      "learning_rate": 4.905013669146127e-06,
+      "loss": 0.376,
+      "step": 136
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.6771908402442932,
+      "learning_rate": 4.903620753754949e-06,
+      "loss": 0.5261,
+      "step": 137
+    },
+    {
+      "epoch": 1.3398058252427185,
+      "grad_norm": 0.8547617197036743,
+      "learning_rate": 4.902217900144524e-06,
+      "loss": 0.6715,
+      "step": 138
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.6083062291145325,
+      "learning_rate": 4.900805114115214e-06,
+      "loss": 0.5999,
+      "step": 139
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.6749557256698608,
+      "learning_rate": 4.899382401508446e-06,
+      "loss": 0.3316,
+      "step": 140
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.6656121015548706,
+      "learning_rate": 4.8979497682066916e-06,
+      "loss": 0.4876,
+      "step": 141
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.5090677738189697,
+      "learning_rate": 4.89650722013344e-06,
+      "loss": 0.3391,
+      "step": 142
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 0.5438173413276672,
+      "learning_rate": 4.895054763253177e-06,
+      "loss": 0.436,
+      "step": 143
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.6639130115509033,
+      "learning_rate": 4.8935924035713564e-06,
+      "loss": 0.3602,
+      "step": 144
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.5315232276916504,
+      "learning_rate": 4.892120147134378e-06,
+      "loss": 0.3503,
+      "step": 145
+    },
+    {
+      "epoch": 1.4174757281553398,
+      "grad_norm": 0.7320194840431213,
+      "learning_rate": 4.8906380000295615e-06,
+      "loss": 0.4673,
+      "step": 146
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.6032155156135559,
+      "learning_rate": 4.889145968385121e-06,
+      "loss": 0.3787,
+      "step": 147
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.6545783877372742,
+      "learning_rate": 4.887644058370139e-06,
+      "loss": 0.3103,
+      "step": 148
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.5894538760185242,
+      "learning_rate": 4.886132276194544e-06,
+      "loss": 0.3413,
+      "step": 149
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.6232749223709106,
+      "learning_rate": 4.884610628109082e-06,
+      "loss": 0.3436,
+      "step": 150
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.4999227523803711,
+      "learning_rate": 4.883079120405292e-06,
+      "loss": 0.3365,
+      "step": 151
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.6769834160804749,
+      "learning_rate": 4.881537759415478e-06,
+      "loss": 0.3026,
+      "step": 152
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 0.7683031558990479,
+      "learning_rate": 4.879986551512684e-06,
+      "loss": 0.4916,
+      "step": 153
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.6288866996765137,
+      "learning_rate": 4.878425503110672e-06,
+      "loss": 0.3623,
+      "step": 154
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.7626834511756897,
+      "learning_rate": 4.876854620663887e-06,
+      "loss": 0.4129,
+      "step": 155
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 0.6667632460594177,
+      "learning_rate": 4.875273910667434e-06,
+      "loss": 0.4238,
+      "step": 156
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.582093358039856,
+      "learning_rate": 4.873683379657057e-06,
+      "loss": 0.2778,
+      "step": 157
+    },
+    {
+      "epoch": 1.5339805825242718,
+      "grad_norm": 0.7995895743370056,
+      "learning_rate": 4.8720830342091015e-06,
+      "loss": 0.7007,
+      "step": 158
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.7098531126976013,
+      "learning_rate": 4.870472880940496e-06,
+      "loss": 0.5393,
+      "step": 159
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 1.1624517440795898,
+      "learning_rate": 4.868852926508721e-06,
+      "loss": 0.6077,
+      "step": 160
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 0.80013507604599,
+      "learning_rate": 4.867223177611779e-06,
+      "loss": 0.4601,
+      "step": 161
+    },
+    {
+      "epoch": 1.5728155339805825,
+      "grad_norm": 0.5787361860275269,
+      "learning_rate": 4.865583640988173e-06,
+      "loss": 0.3338,
+      "step": 162
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.5957843661308289,
+      "learning_rate": 4.863934323416871e-06,
+      "loss": 0.4314,
+      "step": 163
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.5439764857292175,
+      "learning_rate": 4.862275231717288e-06,
+      "loss": 0.3712,
+      "step": 164
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.7053182125091553,
+      "learning_rate": 4.860606372749247e-06,
+      "loss": 0.3109,
+      "step": 165
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.5710446238517761,
+      "learning_rate": 4.858927753412958e-06,
+      "loss": 0.3766,
+      "step": 166
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.6648595333099365,
+      "learning_rate": 4.857239380648985e-06,
+      "loss": 0.3825,
+      "step": 167
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.6290114521980286,
+      "learning_rate": 4.855541261438223e-06,
+      "loss": 0.4474,
+      "step": 168
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.8073899745941162,
+      "learning_rate": 4.8538334028018605e-06,
+      "loss": 0.5414,
+      "step": 169
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.4455365240573883,
+      "learning_rate": 4.8521158118013605e-06,
+      "loss": 0.2714,
+      "step": 170
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 0.7584375143051147,
+      "learning_rate": 4.850388495538423e-06,
+      "loss": 0.3977,
+      "step": 171
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.7362307906150818,
+      "learning_rate": 4.84865146115496e-06,
+      "loss": 0.361,
+      "step": 172
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 0.6931505799293518,
+      "learning_rate": 4.846904715833066e-06,
+      "loss": 0.3181,
+      "step": 173
+    },
+    {
+      "epoch": 1.6893203883495147,
+      "grad_norm": 0.6749580502510071,
+      "learning_rate": 4.8451482667949836e-06,
+      "loss": 0.5013,
+      "step": 174
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.6496002078056335,
+      "learning_rate": 4.843382121303082e-06,
+      "loss": 0.4404,
+      "step": 175
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.5231767892837524,
+      "learning_rate": 4.841606286659819e-06,
+      "loss": 0.2943,
+      "step": 176
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.6066251993179321,
+      "learning_rate": 4.839820770207714e-06,
+      "loss": 0.3056,
+      "step": 177
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.6501230001449585,
+      "learning_rate": 4.8380255793293195e-06,
+      "loss": 0.3317,
+      "step": 178
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 0.6353102326393127,
+      "learning_rate": 4.8362207214471864e-06,
+      "loss": 0.3367,
+      "step": 179
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.7967941164970398,
+      "learning_rate": 4.83440620402384e-06,
+      "loss": 0.4452,
+      "step": 180
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.6307664513587952,
+      "learning_rate": 4.832582034561738e-06,
+      "loss": 0.4285,
+      "step": 181
+    },
+    {
+      "epoch": 1.766990291262136,
+      "grad_norm": 0.6044855117797852,
+      "learning_rate": 4.830748220603251e-06,
+      "loss": 0.4512,
+      "step": 182
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 0.6435297727584839,
+      "learning_rate": 4.828904769730628e-06,
+      "loss": 0.4549,
+      "step": 183
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.702921986579895,
+      "learning_rate": 4.827051689565958e-06,
+      "loss": 0.4001,
+      "step": 184
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.668709397315979,
+      "learning_rate": 4.825188987771149e-06,
+      "loss": 0.5234,
+      "step": 185
+    },
+    {
+      "epoch": 1.8058252427184467,
+      "grad_norm": 0.6227490305900574,
+      "learning_rate": 4.82331667204789e-06,
+      "loss": 0.2801,
+      "step": 186
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.891064465045929,
+      "learning_rate": 4.821434750137619e-06,
+      "loss": 0.4557,
+      "step": 187
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.7352958917617798,
+      "learning_rate": 4.819543229821494e-06,
+      "loss": 0.4963,
+      "step": 188
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.5045980215072632,
+      "learning_rate": 4.8176421189203605e-06,
+      "loss": 0.3175,
+      "step": 189
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.6796278953552246,
+      "learning_rate": 4.815731425294716e-06,
+      "loss": 0.4467,
+      "step": 190
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.6816868185997009,
+      "learning_rate": 4.813811156844681e-06,
+      "loss": 0.4753,
+      "step": 191
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 0.7810450196266174,
+      "learning_rate": 4.811881321509964e-06,
+      "loss": 0.485,
+      "step": 192
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.7930178046226501,
+      "learning_rate": 4.809941927269829e-06,
+      "loss": 0.6113,
+      "step": 193
+    },
+    {
+      "epoch": 1.883495145631068,
+      "grad_norm": 0.7545306086540222,
+      "learning_rate": 4.807992982143064e-06,
+      "loss": 0.4116,
+      "step": 194
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.7625388503074646,
+      "learning_rate": 4.806034494187949e-06,
+      "loss": 0.5885,
+      "step": 195
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.6281082630157471,
+      "learning_rate": 4.804066471502216e-06,
+      "loss": 0.3712,
+      "step": 196
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.685569167137146,
+      "learning_rate": 4.802088922223024e-06,
+      "loss": 0.3116,
+      "step": 197
+    },
+    {
+      "epoch": 1.9223300970873787,
+      "grad_norm": 0.6114542484283447,
+      "learning_rate": 4.80010185452692e-06,
+      "loss": 0.4237,
+      "step": 198
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.5424678921699524,
+      "learning_rate": 4.798105276629806e-06,
+      "loss": 0.2987,
+      "step": 199
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.6528575420379639,
+      "learning_rate": 4.796099196786908e-06,
+      "loss": 0.3472,
+      "step": 200
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.5948964953422546,
+      "learning_rate": 4.794083623292737e-06,
+      "loss": 0.5002,
+      "step": 201
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.6066758036613464,
+      "learning_rate": 4.792058564481058e-06,
+      "loss": 0.3192,
+      "step": 202
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.6982361078262329,
+      "learning_rate": 4.7900240287248554e-06,
+      "loss": 0.4737,
+      "step": 203
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.7239336967468262,
+      "learning_rate": 4.7879800244362975e-06,
+      "loss": 0.5082,
+      "step": 204
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.6995775103569031,
+      "learning_rate": 4.785926560066703e-06,
+      "loss": 0.5026,
+      "step": 205
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5648424625396729,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.3916,
+      "step": 206
+    },
+    {
+      "epoch": 2.0097087378640777,
+      "grad_norm": 0.5701444149017334,
+      "learning_rate": 4.781791285085209e-06,
+      "loss": 0.3145,
+      "step": 207
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.5663194060325623,
+      "learning_rate": 4.779709491571378e-06,
+      "loss": 0.2859,
+      "step": 208
+    },
+    {
+      "epoch": 2.029126213592233,
+      "grad_norm": 0.5743699669837952,
+      "learning_rate": 4.777618272172573e-06,
+      "loss": 0.4217,
+      "step": 209
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.8450823426246643,
+      "learning_rate": 4.775517635535332e-06,
+      "loss": 0.4749,
+      "step": 210
+    },
+    {
+      "epoch": 2.0485436893203883,
+      "grad_norm": 0.6182947158813477,
+      "learning_rate": 4.77340759034513e-06,
+      "loss": 0.4835,
+      "step": 211
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 0.661399245262146,
+      "learning_rate": 4.771288145326343e-06,
+      "loss": 0.4592,
+      "step": 212
+    },
+    {
+      "epoch": 2.0679611650485437,
+      "grad_norm": 0.561326265335083,
+      "learning_rate": 4.769159309242213e-06,
+      "loss": 0.2659,
+      "step": 213
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 0.5948139429092407,
+      "learning_rate": 4.767021090894809e-06,
+      "loss": 0.4182,
+      "step": 214
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.6271581053733826,
+      "learning_rate": 4.764873499124997e-06,
+      "loss": 0.4742,
+      "step": 215
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.5260875821113586,
+      "learning_rate": 4.762716542812395e-06,
+      "loss": 0.3065,
+      "step": 216
+    },
+    {
+      "epoch": 2.1067961165048543,
+      "grad_norm": 0.6065751314163208,
+      "learning_rate": 4.7605502308753415e-06,
+      "loss": 0.3763,
+      "step": 217
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 0.6876965761184692,
+      "learning_rate": 4.758374572270859e-06,
+      "loss": 0.4534,
+      "step": 218
+    },
+    {
+      "epoch": 2.1262135922330097,
+      "grad_norm": 0.48716917634010315,
+      "learning_rate": 4.756189575994614e-06,
+      "loss": 0.1769,
+      "step": 219
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.603748083114624,
+      "learning_rate": 4.753995251080884e-06,
+      "loss": 0.3719,
+      "step": 220
+    },
+    {
+      "epoch": 2.145631067961165,
+      "grad_norm": 0.47758758068084717,
+      "learning_rate": 4.7517916066025126e-06,
+      "loss": 0.2457,
+      "step": 221
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.692158043384552,
+      "learning_rate": 4.7495786516708806e-06,
+      "loss": 0.5067,
+      "step": 222
+    },
+    {
+      "epoch": 2.1650485436893203,
+      "grad_norm": 0.6843806505203247,
+      "learning_rate": 4.747356395435865e-06,
+      "loss": 0.2829,
+      "step": 223
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.8849295377731323,
+      "learning_rate": 4.745124847085799e-06,
+      "loss": 0.3827,
+      "step": 224
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.6735808849334717,
+      "learning_rate": 4.742884015847436e-06,
+      "loss": 0.355,
+      "step": 225
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 0.6396856904029846,
+      "learning_rate": 4.740633910985911e-06,
+      "loss": 0.4024,
+      "step": 226
+    },
+    {
+      "epoch": 2.203883495145631,
+      "grad_norm": 0.45892441272735596,
+      "learning_rate": 4.738374541804704e-06,
+      "loss": 0.2249,
+      "step": 227
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.5224670767784119,
+      "learning_rate": 4.7361059176456e-06,
+      "loss": 0.2975,
+      "step": 228
+    },
+    {
+      "epoch": 2.2233009708737863,
+      "grad_norm": 0.5963131785392761,
+      "learning_rate": 4.733828047888647e-06,
+      "loss": 0.4551,
+      "step": 229
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.49990126490592957,
+      "learning_rate": 4.731540941952126e-06,
+      "loss": 0.2487,
+      "step": 230
+    },
+    {
+      "epoch": 2.2427184466019416,
+      "grad_norm": 0.6114132404327393,
+      "learning_rate": 4.7292446092925016e-06,
+      "loss": 0.323,
+      "step": 231
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.5504849553108215,
+      "learning_rate": 4.726939059404392e-06,
+      "loss": 0.2488,
+      "step": 232
+    },
+    {
+      "epoch": 2.262135922330097,
+      "grad_norm": 0.586751401424408,
+      "learning_rate": 4.724624301820524e-06,
+      "loss": 0.2976,
+      "step": 233
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 0.778872549533844,
+      "learning_rate": 4.722300346111695e-06,
+      "loss": 0.423,
+      "step": 234
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.6887276768684387,
+      "learning_rate": 4.719967201886734e-06,
+      "loss": 0.3558,
+      "step": 235
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.6255162358283997,
+      "learning_rate": 4.717624878792461e-06,
+      "loss": 0.514,
+      "step": 236
+    },
+    {
+      "epoch": 2.3009708737864076,
+      "grad_norm": 0.6514340043067932,
+      "learning_rate": 4.715273386513651e-06,
+      "loss": 0.4811,
+      "step": 237
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 0.5994253754615784,
+      "learning_rate": 4.712912734772988e-06,
+      "loss": 0.3557,
+      "step": 238
+    },
+    {
+      "epoch": 2.320388349514563,
+      "grad_norm": 0.4917253851890564,
+      "learning_rate": 4.710542933331025e-06,
+      "loss": 0.3611,
+      "step": 239
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.6361456513404846,
+      "learning_rate": 4.708163991986152e-06,
+      "loss": 0.4329,
+      "step": 240
+    },
+    {
+      "epoch": 2.3398058252427183,
+      "grad_norm": 0.6043516993522644,
+      "learning_rate": 4.705775920574546e-06,
+      "loss": 0.4122,
+      "step": 241
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 0.6577557325363159,
+      "learning_rate": 4.703378728970134e-06,
+      "loss": 0.3015,
+      "step": 242
+    },
+    {
+      "epoch": 2.3592233009708736,
+      "grad_norm": 0.5667371153831482,
+      "learning_rate": 4.700972427084551e-06,
+      "loss": 0.414,
+      "step": 243
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.6049787402153015,
+      "learning_rate": 4.698557024867105e-06,
+      "loss": 0.2795,
+      "step": 244
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.7228444218635559,
+      "learning_rate": 4.696132532304727e-06,
+      "loss": 0.4881,
+      "step": 245
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 0.5614190101623535,
+      "learning_rate": 4.693698959421935e-06,
+      "loss": 0.3381,
+      "step": 246
+    },
+    {
+      "epoch": 2.3980582524271843,
+      "grad_norm": 0.6152287125587463,
+      "learning_rate": 4.691256316280789e-06,
+      "loss": 0.5593,
+      "step": 247
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.5011581778526306,
+      "learning_rate": 4.688804612980855e-06,
+      "loss": 0.2445,
+      "step": 248
+    },
+    {
+      "epoch": 2.4174757281553396,
+      "grad_norm": 0.5896782279014587,
+      "learning_rate": 4.686343859659158e-06,
+      "loss": 0.473,
+      "step": 249
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.6347712874412537,
+      "learning_rate": 4.683874066490143e-06,
+      "loss": 0.3516,
+      "step": 250
+    },
+    {
+      "epoch": 2.436893203883495,
+      "grad_norm": 0.5596681833267212,
+      "learning_rate": 4.681395243685631e-06,
+      "loss": 0.3678,
+      "step": 251
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.7163019180297852,
+      "learning_rate": 4.67890740149478e-06,
+      "loss": 0.3569,
+      "step": 252
+    },
+    {
+      "epoch": 2.4563106796116507,
+      "grad_norm": 0.6242774128913879,
+      "learning_rate": 4.676410550204036e-06,
+      "loss": 0.3928,
+      "step": 253
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.460670530796051,
+      "learning_rate": 4.673904700137098e-06,
+      "loss": 0.2079,
+      "step": 254
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.5833090543746948,
+      "learning_rate": 4.671389861654873e-06,
+      "loss": 0.3023,
+      "step": 255
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.5938665270805359,
+      "learning_rate": 4.668866045155428e-06,
+      "loss": 0.3848,
+      "step": 256
+    },
+    {
+      "epoch": 2.4951456310679614,
+      "grad_norm": 0.8690244555473328,
+      "learning_rate": 4.666333261073956e-06,
+      "loss": 0.4287,
+      "step": 257
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 0.6412749290466309,
+      "learning_rate": 4.6637915198827265e-06,
+      "loss": 0.466,
+      "step": 258
+    },
+    {
+      "epoch": 2.5145631067961167,
+      "grad_norm": 0.6463085412979126,
+      "learning_rate": 4.661240832091042e-06,
+      "loss": 0.3357,
+      "step": 259
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.6331371665000916,
+      "learning_rate": 4.658681208245198e-06,
+      "loss": 0.3715,
+      "step": 260
+    },
+    {
+      "epoch": 2.533980582524272,
+      "grad_norm": 0.5888141393661499,
+      "learning_rate": 4.65611265892844e-06,
+      "loss": 0.3909,
+      "step": 261
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.6700317859649658,
+      "learning_rate": 4.653535194760912e-06,
+      "loss": 0.3722,
+      "step": 262
+    },
+    {
+      "epoch": 2.5533980582524274,
+      "grad_norm": 0.662674069404602,
+      "learning_rate": 4.650948826399624e-06,
+      "loss": 0.2092,
+      "step": 263
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.7149965167045593,
+      "learning_rate": 4.648353564538397e-06,
+      "loss": 0.2988,
+      "step": 264
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.6281551122665405,
+      "learning_rate": 4.645749419907829e-06,
+      "loss": 0.3102,
+      "step": 265
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 0.5224995017051697,
+      "learning_rate": 4.64313640327524e-06,
+      "loss": 0.2248,
+      "step": 266
+    },
+    {
+      "epoch": 2.592233009708738,
+      "grad_norm": 0.6341948509216309,
+      "learning_rate": 4.640514525444637e-06,
+      "loss": 0.4047,
+      "step": 267
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 0.7885901927947998,
+      "learning_rate": 4.637883797256663e-06,
+      "loss": 0.6537,
+      "step": 268
+    },
+    {
+      "epoch": 2.6116504854368934,
+      "grad_norm": 0.6814199686050415,
+      "learning_rate": 4.635244229588558e-06,
+      "loss": 0.352,
+      "step": 269
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.5105057954788208,
+      "learning_rate": 4.632595833354105e-06,
+      "loss": 0.3706,
+      "step": 270
+    },
+    {
+      "epoch": 2.6310679611650487,
+      "grad_norm": 0.5718193054199219,
+      "learning_rate": 4.629938619503593e-06,
+      "loss": 0.2927,
+      "step": 271
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.5923221707344055,
+      "learning_rate": 4.627272599023772e-06,
+      "loss": 0.4807,
+      "step": 272
+    },
+    {
+      "epoch": 2.650485436893204,
+      "grad_norm": 0.624381422996521,
+      "learning_rate": 4.6245977829378e-06,
+      "loss": 0.5805,
+      "step": 273
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.5855986475944519,
+      "learning_rate": 4.6219141823052035e-06,
+      "loss": 0.2696,
+      "step": 274
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.8095797896385193,
+      "learning_rate": 4.619221808221833e-06,
+      "loss": 0.5238,
+      "step": 275
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.5612257719039917,
+      "learning_rate": 4.616520671819812e-06,
+      "loss": 0.282,
+      "step": 276
+    },
+    {
+      "epoch": 2.6893203883495147,
+      "grad_norm": 0.7121798992156982,
+      "learning_rate": 4.613810784267492e-06,
+      "loss": 0.4285,
+      "step": 277
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.54909747838974,
+      "learning_rate": 4.61109215676941e-06,
+      "loss": 0.2892,
+      "step": 278
+    },
+    {
+      "epoch": 2.70873786407767,
+      "grad_norm": 0.6088219881057739,
+      "learning_rate": 4.608364800566241e-06,
+      "loss": 0.4092,
+      "step": 279
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.548576831817627,
+      "learning_rate": 4.605628726934747e-06,
+      "loss": 0.4547,
+      "step": 280
+    },
+    {
+      "epoch": 2.7281553398058254,
+      "grad_norm": 0.6479307413101196,
+      "learning_rate": 4.602883947187738e-06,
+      "loss": 0.4217,
+      "step": 281
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.5298342704772949,
+      "learning_rate": 4.600130472674017e-06,
+      "loss": 0.2735,
+      "step": 282
+    },
+    {
+      "epoch": 2.7475728155339807,
+      "grad_norm": 0.6714808940887451,
+      "learning_rate": 4.5973683147783405e-06,
+      "loss": 0.549,
+      "step": 283
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.4988473951816559,
+      "learning_rate": 4.594597484921365e-06,
+      "loss": 0.3424,
+      "step": 284
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.6939184665679932,
+      "learning_rate": 4.5918179945596055e-06,
+      "loss": 0.3471,
+      "step": 285
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.49857980012893677,
+      "learning_rate": 4.589029855185384e-06,
+      "loss": 0.2869,
+      "step": 286
+    },
+    {
+      "epoch": 2.7864077669902914,
+      "grad_norm": 0.7681678533554077,
+      "learning_rate": 4.586233078326785e-06,
+      "loss": 0.6411,
+      "step": 287
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.6754655241966248,
+      "learning_rate": 4.583427675547602e-06,
+      "loss": 0.4317,
+      "step": 288
+    },
+    {
+      "epoch": 2.8058252427184467,
+      "grad_norm": 0.7244645357131958,
+      "learning_rate": 4.580613658447301e-06,
+      "loss": 0.4869,
+      "step": 289
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.5869209170341492,
+      "learning_rate": 4.577791038660959e-06,
+      "loss": 0.2967,
+      "step": 290
+    },
+    {
+      "epoch": 2.825242718446602,
+      "grad_norm": 0.6289324164390564,
+      "learning_rate": 4.574959827859226e-06,
+      "loss": 0.268,
+      "step": 291
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.5759575366973877,
+      "learning_rate": 4.572120037748273e-06,
+      "loss": 0.3052,
+      "step": 292
+    },
+    {
+      "epoch": 2.8446601941747574,
+      "grad_norm": 0.6194033622741699,
+      "learning_rate": 4.5692716800697415e-06,
+      "loss": 0.4295,
+      "step": 293
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.5214007496833801,
+      "learning_rate": 4.566414766600698e-06,
+      "loss": 0.3377,
+      "step": 294
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.5058611631393433,
+      "learning_rate": 4.563549309153589e-06,
+      "loss": 0.2281,
+      "step": 295
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.5612390041351318,
+      "learning_rate": 4.56067531957618e-06,
+      "loss": 0.2293,
+      "step": 296
+    },
+    {
+      "epoch": 2.883495145631068,
+      "grad_norm": 0.5222321152687073,
+      "learning_rate": 4.557792809751519e-06,
+      "loss": 0.3842,
+      "step": 297
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.5909943580627441,
+      "learning_rate": 4.554901791597883e-06,
+      "loss": 0.3353,
+      "step": 298
+    },
+    {
+      "epoch": 2.9029126213592233,
+      "grad_norm": 0.6016085743904114,
+      "learning_rate": 4.552002277068725e-06,
+      "loss": 0.341,
+      "step": 299
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.640730082988739,
+      "learning_rate": 4.549094278152631e-06,
+      "loss": 0.4173,
+      "step": 300
+    },
+    {
+      "epoch": 2.9223300970873787,
+      "grad_norm": 0.6297656893730164,
+      "learning_rate": 4.546177806873266e-06,
+      "loss": 0.2969,
+      "step": 301
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 0.7191891074180603,
+      "learning_rate": 4.543252875289326e-06,
+      "loss": 0.3705,
+      "step": 302
+    },
+    {
+      "epoch": 2.941747572815534,
+      "grad_norm": 0.6503893136978149,
+      "learning_rate": 4.540319495494486e-06,
+      "loss": 0.4036,
+      "step": 303
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.6657088994979858,
+      "learning_rate": 4.537377679617353e-06,
+      "loss": 0.3952,
+      "step": 304
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.5238518714904785,
+      "learning_rate": 4.534427439821416e-06,
+      "loss": 0.2522,
+      "step": 305
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 0.6133108735084534,
+      "learning_rate": 4.531468788304992e-06,
+      "loss": 0.2787,
+      "step": 306
+    },
+    {
+      "epoch": 2.9805825242718447,
+      "grad_norm": 0.6255090832710266,
+      "learning_rate": 4.5285017373011784e-06,
+      "loss": 0.301,
+      "step": 307
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.4754847586154938,
+      "learning_rate": 4.5255262990778024e-06,
+      "loss": 0.2602,
+      "step": 308
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.499038428068161,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.272,
+      "step": 309
+    },
+    {
+      "epoch": 3.0097087378640777,
+      "grad_norm": 0.5872356295585632,
+      "learning_rate": 4.519550310217013e-06,
+      "loss": 0.3217,
+      "step": 310
+    },
+    {
+      "epoch": 3.0194174757281553,
+      "grad_norm": 0.5508328676223755,
+      "learning_rate": 4.516549784288442e-06,
+      "loss": 0.2727,
+      "step": 311
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.5548502802848816,
+      "learning_rate": 4.513540920557892e-06,
+      "loss": 0.2955,
+      "step": 312
+    },
+    {
+      "epoch": 3.0388349514563107,
+      "grad_norm": 0.5459408760070801,
+      "learning_rate": 4.510523731466072e-06,
+      "loss": 0.4611,
+      "step": 313
+    },
+    {
+      "epoch": 3.0485436893203883,
+      "grad_norm": 0.40116170048713684,
+      "learning_rate": 4.507498229488116e-06,
+      "loss": 0.1619,
+      "step": 314
+    },
+    {
+      "epoch": 3.058252427184466,
+      "grad_norm": 0.559270977973938,
+      "learning_rate": 4.504464427133527e-06,
+      "loss": 0.2794,
+      "step": 315
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 0.7351297736167908,
+      "learning_rate": 4.501422336946126e-06,
+      "loss": 0.3053,
+      "step": 316
+    },
+    {
+      "epoch": 3.0776699029126213,
+      "grad_norm": 0.623252809047699,
+      "learning_rate": 4.498371971504005e-06,
+      "loss": 0.2301,
+      "step": 317
+    },
+    {
+      "epoch": 3.087378640776699,
+      "grad_norm": 0.5645596385002136,
+      "learning_rate": 4.49531334341947e-06,
+      "loss": 0.2475,
+      "step": 318
+    },
+    {
+      "epoch": 3.0970873786407767,
+      "grad_norm": 0.5398479104042053,
+      "learning_rate": 4.49224646533899e-06,
+      "loss": 0.3074,
+      "step": 319
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 0.5792535543441772,
+      "learning_rate": 4.489171349943144e-06,
+      "loss": 0.2575,
+      "step": 320
+    },
+    {
+      "epoch": 3.116504854368932,
+      "grad_norm": 0.5130180716514587,
+      "learning_rate": 4.486088009946575e-06,
+      "loss": 0.1722,
+      "step": 321
+    },
+    {
+      "epoch": 3.1262135922330097,
+      "grad_norm": 0.6537393927574158,
+      "learning_rate": 4.482996458097926e-06,
+      "loss": 0.3127,
+      "step": 322
+    },
+    {
+      "epoch": 3.1359223300970873,
+      "grad_norm": 0.609282910823822,
+      "learning_rate": 4.479896707179796e-06,
+      "loss": 0.4708,
+      "step": 323
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 0.6748102903366089,
+      "learning_rate": 4.476788770008685e-06,
+      "loss": 0.4804,
+      "step": 324
+    },
+    {
+      "epoch": 3.1553398058252426,
+      "grad_norm": 0.6340557336807251,
+      "learning_rate": 4.473672659434941e-06,
+      "loss": 0.4084,
+      "step": 325
+    },
+    {
+      "epoch": 3.1650485436893203,
+      "grad_norm": 0.6230309009552002,
+      "learning_rate": 4.470548388342704e-06,
+      "loss": 0.3335,
+      "step": 326
+    },
+    {
+      "epoch": 3.174757281553398,
+      "grad_norm": 0.6684495210647583,
+      "learning_rate": 4.467415969649858e-06,
+      "loss": 0.2872,
+      "step": 327
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 0.6560296416282654,
+      "learning_rate": 4.464275416307973e-06,
+      "loss": 0.359,
+      "step": 328
+    },
+    {
+      "epoch": 3.1941747572815533,
+      "grad_norm": 0.6728345155715942,
+      "learning_rate": 4.461126741302253e-06,
+      "loss": 0.3189,
+      "step": 329
+    },
+    {
+      "epoch": 3.203883495145631,
+      "grad_norm": 0.7458807826042175,
+      "learning_rate": 4.457969957651485e-06,
+      "loss": 0.2736,
+      "step": 330
+    },
+    {
+      "epoch": 3.2135922330097086,
+      "grad_norm": 0.7018234133720398,
+      "learning_rate": 4.454805078407979e-06,
+      "loss": 0.4862,
+      "step": 331
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.49450060725212097,
+      "learning_rate": 4.451632116657521e-06,
+      "loss": 0.2543,
+      "step": 332
+    },
+    {
+      "epoch": 3.233009708737864,
+      "grad_norm": 0.48408615589141846,
+      "learning_rate": 4.448451085519314e-06,
+      "loss": 0.212,
+      "step": 333
+    },
+    {
+      "epoch": 3.2427184466019416,
+      "grad_norm": 0.6958401203155518,
+      "learning_rate": 4.445261998145927e-06,
+      "loss": 0.3268,
+      "step": 334
+    },
+    {
+      "epoch": 3.2524271844660193,
+      "grad_norm": 0.5003550052642822,
+      "learning_rate": 4.442064867723236e-06,
+      "loss": 0.2682,
+      "step": 335
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 0.5778501033782959,
+      "learning_rate": 4.438859707470376e-06,
+      "loss": 0.2943,
+      "step": 336
+    },
+    {
+      "epoch": 3.2718446601941746,
+      "grad_norm": 0.7064384818077087,
+      "learning_rate": 4.435646530639679e-06,
+      "loss": 0.1909,
+      "step": 337
+    },
+    {
+      "epoch": 3.2815533980582523,
+      "grad_norm": 0.7310653328895569,
+      "learning_rate": 4.432425350516627e-06,
+      "loss": 0.3328,
+      "step": 338
+    },
+    {
+      "epoch": 3.29126213592233,
+      "grad_norm": 0.6794629693031311,
+      "learning_rate": 4.42919618041979e-06,
+      "loss": 0.1964,
+      "step": 339
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 0.6210153698921204,
+      "learning_rate": 4.425959033700776e-06,
+      "loss": 0.1117,
+      "step": 340
+    },
+    {
+      "epoch": 3.3106796116504853,
+      "grad_norm": 0.6585896015167236,
+      "learning_rate": 4.422713923744174e-06,
+      "loss": 0.3615,
+      "step": 341
+    },
+    {
+      "epoch": 3.320388349514563,
+      "grad_norm": 0.7214643955230713,
+      "learning_rate": 4.419460863967496e-06,
+      "loss": 0.4092,
+      "step": 342
+    },
+    {
+      "epoch": 3.3300970873786406,
+      "grad_norm": 0.5474739670753479,
+      "learning_rate": 4.416199867821126e-06,
+      "loss": 0.3305,
+      "step": 343
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 0.6251519322395325,
+      "learning_rate": 4.412930948788263e-06,
+      "loss": 0.2833,
+      "step": 344
+    },
+    {
+      "epoch": 3.349514563106796,
+      "grad_norm": 0.692952036857605,
+      "learning_rate": 4.409654120384863e-06,
+      "loss": 0.4526,
+      "step": 345
+    },
+    {
+      "epoch": 3.3592233009708736,
+      "grad_norm": 0.5494093894958496,
+      "learning_rate": 4.406369396159585e-06,
+      "loss": 0.3776,
+      "step": 346
+    },
+    {
+      "epoch": 3.3689320388349513,
+      "grad_norm": 0.7002531886100769,
+      "learning_rate": 4.403076789693735e-06,
+      "loss": 0.4925,
+      "step": 347
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 0.5736033320426941,
+      "learning_rate": 4.399776314601212e-06,
+      "loss": 0.2497,
+      "step": 348
+    },
+    {
+      "epoch": 3.3883495145631066,
+      "grad_norm": 0.6323889493942261,
+      "learning_rate": 4.396467984528445e-06,
+      "loss": 0.4034,
+      "step": 349
+    },
+    {
+      "epoch": 3.3980582524271843,
+      "grad_norm": 0.591261088848114,
+      "learning_rate": 4.393151813154345e-06,
+      "loss": 0.2422,
+      "step": 350
+    },
+    {
+      "epoch": 3.407766990291262,
+      "grad_norm": 0.5683301687240601,
+      "learning_rate": 4.3898278141902396e-06,
+      "loss": 0.2869,
+      "step": 351
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 0.6677878499031067,
+      "learning_rate": 4.386496001379826e-06,
+      "loss": 0.3871,
+      "step": 352
+    },
+    {
+      "epoch": 3.4271844660194173,
+      "grad_norm": 0.8581719994544983,
+      "learning_rate": 4.383156388499106e-06,
+      "loss": 0.3608,
+      "step": 353
+    },
+    {
+      "epoch": 3.436893203883495,
+      "grad_norm": 0.5784116387367249,
+      "learning_rate": 4.3798089893563335e-06,
+      "loss": 0.2391,
+      "step": 354
+    },
+    {
+      "epoch": 3.4466019417475726,
+      "grad_norm": 0.6105745434761047,
+      "learning_rate": 4.3764538177919555e-06,
+      "loss": 0.1661,
+      "step": 355
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 0.6270071864128113,
+      "learning_rate": 4.3730908876785574e-06,
+      "loss": 0.2744,
+      "step": 356
+    },
+    {
+      "epoch": 3.466019417475728,
+      "grad_norm": 0.6848212480545044,
+      "learning_rate": 4.3697202129208e-06,
+      "loss": 0.3768,
+      "step": 357
+    },
+    {
+      "epoch": 3.475728155339806,
+      "grad_norm": 0.5955895781517029,
+      "learning_rate": 4.36634180745537e-06,
+      "loss": 0.1898,
+      "step": 358
+    },
+    {
+      "epoch": 3.4854368932038833,
+      "grad_norm": 0.5630469918251038,
+      "learning_rate": 4.3629556852509145e-06,
+      "loss": 0.4752,
+      "step": 359
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.5288087725639343,
+      "learning_rate": 4.35956186030799e-06,
+      "loss": 0.2313,
+      "step": 360
+    },
+    {
+      "epoch": 3.5048543689320386,
+      "grad_norm": 0.525031566619873,
+      "learning_rate": 4.356160346659001e-06,
+      "loss": 0.29,
+      "step": 361
+    },
+    {
+      "epoch": 3.5145631067961167,
+      "grad_norm": 0.7154048681259155,
+      "learning_rate": 4.3527511583681384e-06,
+      "loss": 0.7446,
+      "step": 362
+    },
+    {
+      "epoch": 3.524271844660194,
+      "grad_norm": 0.6375012993812561,
+      "learning_rate": 4.34933430953133e-06,
+      "loss": 0.3629,
+      "step": 363
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 0.7044176459312439,
+      "learning_rate": 4.345909814276177e-06,
+      "loss": 0.4217,
+      "step": 364
+    },
+    {
+      "epoch": 3.5436893203883493,
+      "grad_norm": 0.5310605764389038,
+      "learning_rate": 4.3424776867618935e-06,
+      "loss": 0.3776,
+      "step": 365
+    },
+    {
+      "epoch": 3.5533980582524274,
+      "grad_norm": 0.5908545851707458,
+      "learning_rate": 4.339037941179253e-06,
+      "loss": 0.3704,
+      "step": 366
+    },
+    {
+      "epoch": 3.5631067961165046,
+      "grad_norm": 0.5712844729423523,
+      "learning_rate": 4.335590591750526e-06,
+      "loss": 0.2685,
+      "step": 367
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 0.6057801842689514,
+      "learning_rate": 4.332135652729423e-06,
+      "loss": 0.2823,
+      "step": 368
+    },
+    {
+      "epoch": 3.58252427184466,
+      "grad_norm": 0.5957722067832947,
+      "learning_rate": 4.328673138401036e-06,
+      "loss": 0.401,
+      "step": 369
+    },
+    {
+      "epoch": 3.592233009708738,
+      "grad_norm": 0.690753698348999,
+      "learning_rate": 4.325203063081776e-06,
+      "loss": 0.2192,
+      "step": 370
+    },
+    {
+      "epoch": 3.6019417475728153,
+      "grad_norm": 0.6514263153076172,
+      "learning_rate": 4.32172544111932e-06,
+      "loss": 0.2097,
+      "step": 371
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 0.4971867799758911,
+      "learning_rate": 4.318240286892544e-06,
+      "loss": 0.2443,
+      "step": 372
+    },
+    {
+      "epoch": 3.6213592233009706,
+      "grad_norm": 0.5361030101776123,
+      "learning_rate": 4.314747614811471e-06,
+      "loss": 0.2588,
+      "step": 373
+    },
+    {
+      "epoch": 3.6310679611650487,
+      "grad_norm": 0.5564566850662231,
+      "learning_rate": 4.3112474393172055e-06,
+      "loss": 0.2143,
+      "step": 374
+    },
+    {
+      "epoch": 3.6407766990291264,
+      "grad_norm": 0.5552682876586914,
+      "learning_rate": 4.307739774881878e-06,
+      "loss": 0.2823,
+      "step": 375
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 0.6247363686561584,
+      "learning_rate": 4.304224636008582e-06,
+      "loss": 0.1829,
+      "step": 376
+    },
+    {
+      "epoch": 3.6601941747572817,
+      "grad_norm": 0.6173890829086304,
+      "learning_rate": 4.300702037231318e-06,
+      "loss": 0.413,
+      "step": 377
+    },
+    {
+      "epoch": 3.6699029126213594,
+      "grad_norm": 0.642781138420105,
+      "learning_rate": 4.297171993114927e-06,
+      "loss": 0.145,
+      "step": 378
+    },
+    {
+      "epoch": 3.679611650485437,
+      "grad_norm": 0.6031596660614014,
+      "learning_rate": 4.2936345182550365e-06,
+      "loss": 0.4141,
+      "step": 379
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.5309922695159912,
+      "learning_rate": 4.290089627277998e-06,
+      "loss": 0.2209,
+      "step": 380
+    },
+    {
+      "epoch": 3.6990291262135924,
+      "grad_norm": 0.5616117715835571,
+      "learning_rate": 4.286537334840825e-06,
+      "loss": 0.3514,
+      "step": 381
+    },
+    {
+      "epoch": 3.70873786407767,
+      "grad_norm": 0.6711661219596863,
+      "learning_rate": 4.2829776556311355e-06,
+      "loss": 0.2756,
+      "step": 382
+    },
+    {
+      "epoch": 3.7184466019417477,
+      "grad_norm": 0.9085079431533813,
+      "learning_rate": 4.279410604367088e-06,
+      "loss": 0.5231,
+      "step": 383
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 0.7275105714797974,
+      "learning_rate": 4.275836195797323e-06,
+      "loss": 0.3773,
+      "step": 384
+    },
+    {
+      "epoch": 3.737864077669903,
+      "grad_norm": 0.6976683139801025,
+      "learning_rate": 4.2722544447008995e-06,
+      "loss": 0.4558,
+      "step": 385
+    },
+    {
+      "epoch": 3.7475728155339807,
+      "grad_norm": 0.6676245331764221,
+      "learning_rate": 4.268665365887238e-06,
+      "loss": 0.4503,
+      "step": 386
+    },
+    {
+      "epoch": 3.7572815533980584,
+      "grad_norm": 0.46635928750038147,
+      "learning_rate": 4.265068974196056e-06,
+      "loss": 0.157,
+      "step": 387
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 0.5962404608726501,
+      "learning_rate": 4.261465284497307e-06,
+      "loss": 0.2184,
+      "step": 388
+    },
+    {
+      "epoch": 3.7766990291262137,
+      "grad_norm": 0.7135609984397888,
+      "learning_rate": 4.257854311691118e-06,
+      "loss": 0.4233,
+      "step": 389
+    },
+    {
+      "epoch": 3.7864077669902914,
+      "grad_norm": 0.48697859048843384,
+      "learning_rate": 4.254236070707734e-06,
+      "loss": 0.2134,
+      "step": 390
+    },
+    {
+      "epoch": 3.796116504854369,
+      "grad_norm": 0.7304379940032959,
+      "learning_rate": 4.250610576507445e-06,
+      "loss": 0.216,
+      "step": 391
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 0.6933459043502808,
+      "learning_rate": 4.246977844080537e-06,
+      "loss": 0.3859,
+      "step": 392
+    },
+    {
+      "epoch": 3.8155339805825244,
+      "grad_norm": 0.6276050806045532,
+      "learning_rate": 4.24333788844722e-06,
+      "loss": 0.2899,
+      "step": 393
+    },
+    {
+      "epoch": 3.825242718446602,
+      "grad_norm": 0.8272635340690613,
+      "learning_rate": 4.239690724657571e-06,
+      "loss": 0.324,
+      "step": 394
+    },
+    {
+      "epoch": 3.8349514563106797,
+      "grad_norm": 0.5484209060668945,
+      "learning_rate": 4.236036367791471e-06,
+      "loss": 0.3267,
+      "step": 395
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 0.672440767288208,
+      "learning_rate": 4.23237483295854e-06,
+      "loss": 0.4328,
+      "step": 396
+    },
+    {
+      "epoch": 3.854368932038835,
+      "grad_norm": 0.6158328056335449,
+      "learning_rate": 4.228706135298081e-06,
+      "loss": 0.2575,
+      "step": 397
+    },
+    {
+      "epoch": 3.8640776699029127,
+      "grad_norm": 0.6439457535743713,
+      "learning_rate": 4.225030289979006e-06,
+      "loss": 0.2095,
+      "step": 398
+    },
+    {
+      "epoch": 3.8737864077669903,
+      "grad_norm": 0.5429261922836304,
+      "learning_rate": 4.221347312199788e-06,
+      "loss": 0.2205,
+      "step": 399
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.5270217061042786,
+      "learning_rate": 4.2176572171883865e-06,
+      "loss": 0.3106,
+      "step": 400
+    },
+    {
+      "epoch": 3.8932038834951457,
+      "grad_norm": 0.5904674530029297,
+      "learning_rate": 4.213960020202187e-06,
+      "loss": 0.3903,
+      "step": 401
+    },
+    {
+      "epoch": 3.9029126213592233,
+      "grad_norm": 0.7325130105018616,
+      "learning_rate": 4.2102557365279435e-06,
+      "loss": 0.5404,
+      "step": 402
+    },
+    {
+      "epoch": 3.912621359223301,
+      "grad_norm": 0.6240959763526917,
+      "learning_rate": 4.206544381481708e-06,
+      "loss": 0.399,
+      "step": 403
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 0.632676362991333,
+      "learning_rate": 4.202825970408772e-06,
+      "loss": 0.187,
+      "step": 404
+    },
+    {
+      "epoch": 3.9320388349514563,
+      "grad_norm": 0.5993353128433228,
+      "learning_rate": 4.199100518683601e-06,
+      "loss": 0.2331,
+      "step": 405
+    },
+    {
+      "epoch": 3.941747572815534,
+      "grad_norm": 0.5927851796150208,
+      "learning_rate": 4.195368041709772e-06,
+      "loss": 0.3959,
+      "step": 406
+    },
+    {
+      "epoch": 3.9514563106796117,
+      "grad_norm": 0.494970440864563,
+      "learning_rate": 4.191628554919907e-06,
+      "loss": 0.2533,
+      "step": 407
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 0.5709957480430603,
+      "learning_rate": 4.187882073775615e-06,
+      "loss": 0.267,
+      "step": 408
+    },
+    {
+      "epoch": 3.970873786407767,
+      "grad_norm": 0.5627012848854065,
+      "learning_rate": 4.184128613767422e-06,
+      "loss": 0.2914,
+      "step": 409
+    },
+    {
+      "epoch": 3.9805825242718447,
+      "grad_norm": 0.5530299544334412,
+      "learning_rate": 4.18036819041471e-06,
+      "loss": 0.3071,
+      "step": 410
+    },
+    {
+      "epoch": 3.9902912621359223,
+      "grad_norm": 0.5919513702392578,
+      "learning_rate": 4.17660081926565e-06,
+      "loss": 0.2059,
+      "step": 411
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6598716974258423,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.1953,
+      "step": 412
+    },
+    {
+      "epoch": 4.009708737864078,
+      "grad_norm": 0.5695852041244507,
+      "learning_rate": 4.169045295914757e-06,
+      "loss": 0.3486,
+      "step": 413
+    },
+    {
+      "epoch": 4.019417475728155,
+      "grad_norm": 0.5444216728210449,
+      "learning_rate": 4.165257174952647e-06,
+      "loss": 0.1647,
+      "step": 414
+    },
+    {
+      "epoch": 4.029126213592233,
+      "grad_norm": 0.5701639652252197,
+      "learning_rate": 4.161462168673508e-06,
+      "loss": 0.2867,
+      "step": 415
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 0.5739265084266663,
+      "learning_rate": 4.157660292768502e-06,
+      "loss": 0.3941,
+      "step": 416
+    },
+    {
+      "epoch": 4.048543689320389,
+      "grad_norm": 0.6720489859580994,
+      "learning_rate": 4.1538515629571985e-06,
+      "loss": 0.2837,
+      "step": 417
+    },
+    {
+      "epoch": 4.058252427184466,
+      "grad_norm": 0.5424773097038269,
+      "learning_rate": 4.1500359949875e-06,
+      "loss": 0.3481,
+      "step": 418
+    },
+    {
+      "epoch": 4.067961165048544,
+      "grad_norm": 0.5736377835273743,
+      "learning_rate": 4.1462136046355864e-06,
+      "loss": 0.2342,
+      "step": 419
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 0.5163815021514893,
+      "learning_rate": 4.142384407705846e-06,
+      "loss": 0.447,
+      "step": 420
+    },
+    {
+      "epoch": 4.087378640776699,
+      "grad_norm": 0.5188374519348145,
+      "learning_rate": 4.138548420030808e-06,
+      "loss": 0.1716,
+      "step": 421
+    },
+    {
+      "epoch": 4.097087378640777,
+      "grad_norm": 0.5863322019577026,
+      "learning_rate": 4.13470565747108e-06,
+      "loss": 0.3029,
+      "step": 422
+    },
+    {
+      "epoch": 4.106796116504855,
+      "grad_norm": 0.48825642466545105,
+      "learning_rate": 4.130856135915282e-06,
+      "loss": 0.2099,
+      "step": 423
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 0.6416239738464355,
+      "learning_rate": 4.126999871279982e-06,
+      "loss": 0.2477,
+      "step": 424
+    },
+    {
+      "epoch": 4.12621359223301,
+      "grad_norm": 0.5978772044181824,
+      "learning_rate": 4.123136879509626e-06,
+      "loss": 0.355,
+      "step": 425
+    },
+    {
+      "epoch": 4.135922330097087,
+      "grad_norm": 0.5654548406600952,
+      "learning_rate": 4.119267176576475e-06,
+      "loss": 0.3269,
+      "step": 426
+    },
+    {
+      "epoch": 4.145631067961165,
+      "grad_norm": 0.5691317319869995,
+      "learning_rate": 4.11539077848054e-06,
+      "loss": 0.3658,
+      "step": 427
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 0.6897786855697632,
+      "learning_rate": 4.111507701249513e-06,
+      "loss": 0.3918,
+      "step": 428
+    },
+    {
+      "epoch": 4.165048543689321,
+      "grad_norm": 0.8434731960296631,
+      "learning_rate": 4.107617960938702e-06,
+      "loss": 0.149,
+      "step": 429
+    },
+    {
+      "epoch": 4.174757281553398,
+      "grad_norm": 0.741525411605835,
+      "learning_rate": 4.103721573630965e-06,
+      "loss": 0.2738,
+      "step": 430
+    },
+    {
+      "epoch": 4.184466019417476,
+      "grad_norm": 0.5480707883834839,
+      "learning_rate": 4.099818555436645e-06,
+      "loss": 0.2704,
+      "step": 431
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 0.9266293048858643,
+      "learning_rate": 4.095908922493499e-06,
+      "loss": 0.5329,
+      "step": 432
+    },
+    {
+      "epoch": 4.203883495145631,
+      "grad_norm": 0.7091429829597473,
+      "learning_rate": 4.091992690966636e-06,
+      "loss": 0.1109,
+      "step": 433
+    },
+    {
+      "epoch": 4.213592233009709,
+      "grad_norm": 0.5552862286567688,
+      "learning_rate": 4.088069877048447e-06,
+      "loss": 0.2407,
+      "step": 434
+    },
+    {
+      "epoch": 4.223300970873787,
+      "grad_norm": 0.6244552135467529,
+      "learning_rate": 4.084140496958539e-06,
+      "loss": 0.2108,
+      "step": 435
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 0.7678143978118896,
+      "learning_rate": 4.080204566943668e-06,
+      "loss": 0.2088,
+      "step": 436
+    },
+    {
+      "epoch": 4.242718446601942,
+      "grad_norm": 0.6703317165374756,
+      "learning_rate": 4.076262103277673e-06,
+      "loss": 0.1528,
+      "step": 437
+    },
+    {
+      "epoch": 4.252427184466019,
+      "grad_norm": 0.6737928986549377,
+      "learning_rate": 4.072313122261406e-06,
+      "loss": 0.4616,
+      "step": 438
+    },
+    {
+      "epoch": 4.262135922330097,
+      "grad_norm": 0.5564897656440735,
+      "learning_rate": 4.068357640222668e-06,
+      "loss": 0.1319,
+      "step": 439
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 0.51083904504776,
+      "learning_rate": 4.06439567351614e-06,
+      "loss": 0.1884,
+      "step": 440
+    },
+    {
+      "epoch": 4.281553398058253,
+      "grad_norm": 0.5311152935028076,
+      "learning_rate": 4.0604272385233105e-06,
+      "loss": 0.2328,
+      "step": 441
+    },
+    {
+      "epoch": 4.29126213592233,
+      "grad_norm": 0.6676695346832275,
+      "learning_rate": 4.056452351652418e-06,
+      "loss": 0.145,
+      "step": 442
+    },
+    {
+      "epoch": 4.300970873786408,
+      "grad_norm": 0.5755549073219299,
+      "learning_rate": 4.052471029338375e-06,
+      "loss": 0.3041,
+      "step": 443
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 0.649605929851532,
+      "learning_rate": 4.048483288042703e-06,
+      "loss": 0.2018,
+      "step": 444
+    },
+    {
+      "epoch": 4.320388349514563,
+      "grad_norm": 0.5598784685134888,
+      "learning_rate": 4.0444891442534615e-06,
+      "loss": 0.1736,
+      "step": 445
+    },
+    {
+      "epoch": 4.330097087378641,
+      "grad_norm": 0.6140623092651367,
+      "learning_rate": 4.040488614485187e-06,
+      "loss": 0.3402,
+      "step": 446
+    },
+    {
+      "epoch": 4.339805825242719,
+      "grad_norm": 0.654649019241333,
+      "learning_rate": 4.036481715278818e-06,
+      "loss": 0.322,
+      "step": 447
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 0.5743938684463501,
+      "learning_rate": 4.032468463201626e-06,
+      "loss": 0.2994,
+      "step": 448
+    },
+    {
+      "epoch": 4.359223300970874,
+      "grad_norm": 0.5662089586257935,
+      "learning_rate": 4.028448874847152e-06,
+      "loss": 0.0994,
+      "step": 449
+    },
+    {
+      "epoch": 4.368932038834951,
+      "grad_norm": 0.645409882068634,
+      "learning_rate": 4.024422966835137e-06,
+      "loss": 0.2807,
+      "step": 450
+    },
+    {
+      "epoch": 4.378640776699029,
+      "grad_norm": 0.7538043856620789,
+      "learning_rate": 4.0203907558114475e-06,
+      "loss": 0.2016,
+      "step": 451
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 0.554035484790802,
+      "learning_rate": 4.016352258448016e-06,
+      "loss": 0.1306,
+      "step": 452
+    },
+    {
+      "epoch": 4.398058252427185,
+      "grad_norm": 0.6023773550987244,
+      "learning_rate": 4.0123074914427635e-06,
+      "loss": 0.179,
+      "step": 453
+    },
+    {
+      "epoch": 4.407766990291262,
+      "grad_norm": 0.4744186997413635,
+      "learning_rate": 4.008256471519536e-06,
+      "loss": 0.1312,
+      "step": 454
+    },
+    {
+      "epoch": 4.41747572815534,
+      "grad_norm": 0.6246916651725769,
+      "learning_rate": 4.004199215428032e-06,
+      "loss": 0.279,
+      "step": 455
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 0.8718860745429993,
+      "learning_rate": 4.000135739943735e-06,
+      "loss": 0.4049,
+      "step": 456
+    },
+    {
+      "epoch": 4.436893203883495,
+      "grad_norm": 0.5976299047470093,
+      "learning_rate": 3.996066061867844e-06,
+      "loss": 0.2779,
+      "step": 457
+    },
+    {
+      "epoch": 4.446601941747573,
+      "grad_norm": 0.6921701431274414,
+      "learning_rate": 3.991990198027203e-06,
+      "loss": 0.2268,
+      "step": 458
+    },
+    {
+      "epoch": 4.456310679611651,
+      "grad_norm": 0.5831341743469238,
+      "learning_rate": 3.987908165274233e-06,
+      "loss": 0.2781,
+      "step": 459
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 0.5739938020706177,
+      "learning_rate": 3.9838199804868635e-06,
+      "loss": 0.353,
+      "step": 460
+    },
+    {
+      "epoch": 4.475728155339806,
+      "grad_norm": 0.6788473725318909,
+      "learning_rate": 3.979725660568456e-06,
+      "loss": 0.3235,
+      "step": 461
+    },
+    {
+      "epoch": 4.485436893203883,
+      "grad_norm": 0.6106454133987427,
+      "learning_rate": 3.975625222447742e-06,
+      "loss": 0.1903,
+      "step": 462
+    },
+    {
+      "epoch": 4.495145631067961,
+      "grad_norm": 0.8468668460845947,
+      "learning_rate": 3.97151868307875e-06,
+      "loss": 0.3609,
+      "step": 463
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 0.7000426054000854,
+      "learning_rate": 3.9674060594407345e-06,
+      "loss": 0.2357,
+      "step": 464
+    },
+    {
+      "epoch": 4.514563106796117,
+      "grad_norm": 0.5690450072288513,
+      "learning_rate": 3.963287368538105e-06,
+      "loss": 0.1164,
+      "step": 465
+    },
+    {
+      "epoch": 4.524271844660194,
+      "grad_norm": 0.7158642411231995,
+      "learning_rate": 3.959162627400361e-06,
+      "loss": 0.3587,
+      "step": 466
+    },
+    {
+      "epoch": 4.533980582524272,
+      "grad_norm": 0.8226291537284851,
+      "learning_rate": 3.9550318530820145e-06,
+      "loss": 0.3488,
+      "step": 467
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 0.6773446202278137,
+      "learning_rate": 3.9508950626625244e-06,
+      "loss": 0.1872,
+      "step": 468
+    },
+    {
+      "epoch": 4.553398058252427,
+      "grad_norm": 0.6974344849586487,
+      "learning_rate": 3.946752273246224e-06,
+      "loss": 0.3076,
+      "step": 469
+    },
+    {
+      "epoch": 4.563106796116505,
+      "grad_norm": 0.7240371108055115,
+      "learning_rate": 3.942603501962249e-06,
+      "loss": 0.2194,
+      "step": 470
+    },
+    {
+      "epoch": 4.572815533980583,
+      "grad_norm": 0.8507097363471985,
+      "learning_rate": 3.9384487659644716e-06,
+      "loss": 0.1283,
+      "step": 471
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 0.5208114981651306,
+      "learning_rate": 3.934288082431423e-06,
+      "loss": 0.2337,
+      "step": 472
+    },
+    {
+      "epoch": 4.592233009708738,
+      "grad_norm": 0.659975528717041,
+      "learning_rate": 3.930121468566227e-06,
+      "loss": 0.343,
+      "step": 473
+    },
+    {
+      "epoch": 4.601941747572815,
+      "grad_norm": 0.6047101616859436,
+      "learning_rate": 3.925948941596528e-06,
+      "loss": 0.2477,
+      "step": 474
+    },
+    {
+      "epoch": 4.611650485436893,
+      "grad_norm": 0.591589093208313,
+      "learning_rate": 3.92177051877442e-06,
+      "loss": 0.1969,
+      "step": 475
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 0.5828817486763,
+      "learning_rate": 3.917586217376369e-06,
+      "loss": 0.2927,
+      "step": 476
+    },
+    {
+      "epoch": 4.631067961165049,
+      "grad_norm": 0.5926609635353088,
+      "learning_rate": 3.913396054703155e-06,
+      "loss": 0.3156,
+      "step": 477
+    },
+    {
+      "epoch": 4.640776699029126,
+      "grad_norm": 0.5627316832542419,
+      "learning_rate": 3.909200048079786e-06,
+      "loss": 0.2721,
+      "step": 478
+    },
+    {
+      "epoch": 4.650485436893204,
+      "grad_norm": 0.8213399052619934,
+      "learning_rate": 3.9049982148554384e-06,
+      "loss": 0.4026,
+      "step": 479
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 0.6906929016113281,
+      "learning_rate": 3.900790572403376e-06,
+      "loss": 0.245,
+      "step": 480
+    },
+    {
+      "epoch": 4.669902912621359,
+      "grad_norm": 0.6363279819488525,
+      "learning_rate": 3.896577138120881e-06,
+      "loss": 0.2899,
+      "step": 481
+    },
+    {
+      "epoch": 4.679611650485437,
+      "grad_norm": 0.6629766821861267,
+      "learning_rate": 3.892357929429187e-06,
+      "loss": 0.1948,
+      "step": 482
+    },
+    {
+      "epoch": 4.689320388349515,
+      "grad_norm": 0.6485582590103149,
+      "learning_rate": 3.8881329637734e-06,
+      "loss": 0.1518,
+      "step": 483
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 0.5513966083526611,
+      "learning_rate": 3.883902258622431e-06,
+      "loss": 0.2499,
+      "step": 484
+    },
+    {
+      "epoch": 4.70873786407767,
+      "grad_norm": 0.8740300536155701,
+      "learning_rate": 3.8796658314689205e-06,
+      "loss": 0.3368,
+      "step": 485
+    },
+    {
+      "epoch": 4.718446601941747,
+      "grad_norm": 0.6036941409111023,
+      "learning_rate": 3.875423699829168e-06,
+      "loss": 0.1315,
+      "step": 486
+    },
+    {
+      "epoch": 4.728155339805825,
+      "grad_norm": 0.5299080014228821,
+      "learning_rate": 3.871175881243061e-06,
+      "loss": 0.202,
+      "step": 487
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 0.6407957077026367,
+      "learning_rate": 3.866922393273999e-06,
+      "loss": 0.3696,
+      "step": 488
+    },
+    {
+      "epoch": 4.747572815533981,
+      "grad_norm": 0.5987715721130371,
+      "learning_rate": 3.862663253508822e-06,
+      "loss": 0.2131,
+      "step": 489
+    },
+    {
+      "epoch": 4.757281553398058,
+      "grad_norm": 0.5340495109558105,
+      "learning_rate": 3.858398479557739e-06,
+      "loss": 0.2166,
+      "step": 490
+    },
+    {
+      "epoch": 4.766990291262136,
+      "grad_norm": 0.6883168816566467,
+      "learning_rate": 3.8541280890542565e-06,
+      "loss": 0.2626,
+      "step": 491
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 0.9401558041572571,
+      "learning_rate": 3.849852099655102e-06,
+      "loss": 0.3857,
+      "step": 492
+    },
+    {
+      "epoch": 4.786407766990291,
+      "grad_norm": 0.48749417066574097,
+      "learning_rate": 3.845570529040151e-06,
+      "loss": 0.2171,
+      "step": 493
+    },
+    {
+      "epoch": 4.796116504854369,
+      "grad_norm": 0.6752912402153015,
+      "learning_rate": 3.841283394912361e-06,
+      "loss": 0.2269,
+      "step": 494
+    },
+    {
+      "epoch": 4.805825242718447,
+      "grad_norm": 0.60317063331604,
+      "learning_rate": 3.836990714997686e-06,
+      "loss": 0.3227,
+      "step": 495
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 0.6304256319999695,
+      "learning_rate": 3.832692507045015e-06,
+      "loss": 0.332,
+      "step": 496
+    },
+    {
+      "epoch": 4.825242718446602,
+      "grad_norm": 0.5268096923828125,
+      "learning_rate": 3.828388788826091e-06,
+      "loss": 0.3388,
+      "step": 497
+    },
+    {
+      "epoch": 4.834951456310679,
+      "grad_norm": 0.6119520664215088,
+      "learning_rate": 3.824079578135442e-06,
+      "loss": 0.2921,
+      "step": 498
+    },
+    {
+      "epoch": 4.844660194174757,
+      "grad_norm": 0.5615065097808838,
+      "learning_rate": 3.819764892790307e-06,
+      "loss": 0.2708,
+      "step": 499
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 0.6683458685874939,
+      "learning_rate": 3.815444750630555e-06,
+      "loss": 0.2755,
+      "step": 500
+    },
+    {
+      "epoch": 4.864077669902913,
+      "grad_norm": 0.7224542498588562,
+      "learning_rate": 3.811119169518624e-06,
+      "loss": 0.4511,
+      "step": 501
+    },
+    {
+      "epoch": 4.87378640776699,
+      "grad_norm": 0.7381410598754883,
+      "learning_rate": 3.8067881673394363e-06,
+      "loss": 0.4965,
+      "step": 502
+    },
+    {
+      "epoch": 4.883495145631068,
+      "grad_norm": 0.658463716506958,
+      "learning_rate": 3.802451762000331e-06,
+      "loss": 0.2083,
+      "step": 503
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 0.6669811606407166,
+      "learning_rate": 3.7981099714309856e-06,
+      "loss": 0.1759,
+      "step": 504
+    },
+    {
+      "epoch": 4.902912621359223,
+      "grad_norm": 0.7292122840881348,
+      "learning_rate": 3.7937628135833453e-06,
+      "loss": 0.2642,
+      "step": 505
+    },
+    {
+      "epoch": 4.9126213592233015,
+      "grad_norm": 0.7087453007698059,
+      "learning_rate": 3.7894103064315463e-06,
+      "loss": 0.2894,
+      "step": 506
+    },
+    {
+      "epoch": 4.922330097087379,
+      "grad_norm": 0.6095560789108276,
+      "learning_rate": 3.7850524679718424e-06,
+      "loss": 0.2117,
+      "step": 507
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 0.559994101524353,
+      "learning_rate": 3.7806893162225328e-06,
+      "loss": 0.2267,
+      "step": 508
+    },
+    {
+      "epoch": 4.941747572815534,
+      "grad_norm": 0.5426676869392395,
+      "learning_rate": 3.7763208692238818e-06,
+      "loss": 0.2841,
+      "step": 509
+    },
+    {
+      "epoch": 4.951456310679612,
+      "grad_norm": 0.5384762287139893,
+      "learning_rate": 3.7719471450380518e-06,
+      "loss": 0.1805,
+      "step": 510
+    },
+    {
+      "epoch": 4.961165048543689,
+      "grad_norm": 0.5685274004936218,
+      "learning_rate": 3.7675681617490212e-06,
+      "loss": 0.3862,
+      "step": 511
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 0.5709934830665588,
+      "learning_rate": 3.7631839374625167e-06,
+      "loss": 0.3766,
+      "step": 512
+    },
+    {
+      "epoch": 4.980582524271845,
+      "grad_norm": 2.084059238433838,
+      "learning_rate": 3.758794490305932e-06,
+      "loss": 0.2005,
+      "step": 513
+    },
+    {
+      "epoch": 4.990291262135923,
+      "grad_norm": 0.6240888833999634,
+      "learning_rate": 3.7543998384282565e-06,
+      "loss": 0.2399,
+      "step": 514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.5388016700744629,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3159,
+      "step": 515
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 515,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 57374359822336.0,
+  "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 sha256:8f3c7ddc1db65b3b9f5bab5719ee9184e9c7dd18c206085551877d843e3318d8
+size 7608

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)