Upload folder using huggingface_hub
Browse files- --comit-message=Finetuned model with 2000 steps (gr00t)/config.json +64 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/experiment_cfg/metadata.json +195 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/model-00001-of-00002.safetensors +3 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/model-00002-of-00002.safetensors +3 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/model.safetensors.index.json +0 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/optimizer.pt +3 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/rng_state.pth +3 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/scheduler.pt +3 -0
- --comit-message=Finetuned model with 2000 steps (gr00t)/trainer_state.json +1434 -0
--comit-message=Finetuned model with 2000 steps (gr00t)/config.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"action_dim": 32,
|
| 3 |
+
"action_head_cfg": {
|
| 4 |
+
"action_dim": 32,
|
| 5 |
+
"action_horizon": 16,
|
| 6 |
+
"add_pos_embed": true,
|
| 7 |
+
"backbone_embedding_dim": 2048,
|
| 8 |
+
"diffusion_model_cfg": {
|
| 9 |
+
"attention_head_dim": 48,
|
| 10 |
+
"cross_attention_dim": 2048,
|
| 11 |
+
"dropout": 0.2,
|
| 12 |
+
"final_dropout": true,
|
| 13 |
+
"interleave_self_attention": true,
|
| 14 |
+
"norm_type": "ada_norm",
|
| 15 |
+
"num_attention_heads": 32,
|
| 16 |
+
"num_layers": 16,
|
| 17 |
+
"output_dim": 1024,
|
| 18 |
+
"positional_embeddings": null
|
| 19 |
+
},
|
| 20 |
+
"hidden_size": 1024,
|
| 21 |
+
"input_embedding_dim": 1536,
|
| 22 |
+
"max_action_dim": 32,
|
| 23 |
+
"max_state_dim": 64,
|
| 24 |
+
"model_dtype": "float32",
|
| 25 |
+
"noise_beta_alpha": 1.5,
|
| 26 |
+
"noise_beta_beta": 1.0,
|
| 27 |
+
"noise_s": 0.999,
|
| 28 |
+
"num_inference_timesteps": 4,
|
| 29 |
+
"num_target_vision_tokens": 32,
|
| 30 |
+
"num_timestep_buckets": 1000,
|
| 31 |
+
"tune_diffusion_model": true,
|
| 32 |
+
"tune_projector": true,
|
| 33 |
+
"use_vlln": true,
|
| 34 |
+
"vl_self_attention_cfg": {
|
| 35 |
+
"attention_head_dim": 64,
|
| 36 |
+
"dropout": 0.2,
|
| 37 |
+
"final_dropout": true,
|
| 38 |
+
"num_attention_heads": 32,
|
| 39 |
+
"num_layers": 4,
|
| 40 |
+
"positional_embeddings": null
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"action_horizon": 16,
|
| 44 |
+
"architectures": [
|
| 45 |
+
"GR00T_N1_5"
|
| 46 |
+
],
|
| 47 |
+
"attn_implementation": null,
|
| 48 |
+
"backbone_cfg": {
|
| 49 |
+
"eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops",
|
| 50 |
+
"load_bf16": false,
|
| 51 |
+
"project_to_dim": null,
|
| 52 |
+
"reproject_vision": false,
|
| 53 |
+
"select_layer": 12,
|
| 54 |
+
"tune_llm": false,
|
| 55 |
+
"tune_visual": true,
|
| 56 |
+
"use_flash_attention": true
|
| 57 |
+
},
|
| 58 |
+
"compute_dtype": "bfloat16",
|
| 59 |
+
"hidden_size": 2048,
|
| 60 |
+
"model_dtype": "float32",
|
| 61 |
+
"model_type": "gr00t_n1_5",
|
| 62 |
+
"torch_dtype": "bfloat16",
|
| 63 |
+
"transformers_version": "4.51.3"
|
| 64 |
+
}
|
--comit-message=Finetuned model with 2000 steps (gr00t)/experiment_cfg/metadata.json
ADDED
|
@@ -0,0 +1,195 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"single_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
49.547218322753906,
|
| 8 |
+
61.880340576171875,
|
| 9 |
+
97.15302276611328,
|
| 10 |
+
97.56813049316406,
|
| 11 |
+
33.887123107910156
|
| 12 |
+
],
|
| 13 |
+
"min": [
|
| 14 |
+
-22.639068603515625,
|
| 15 |
+
-97.94871520996094,
|
| 16 |
+
-89.50177764892578,
|
| 17 |
+
39.87421417236328,
|
| 18 |
+
-23.52797508239746
|
| 19 |
+
],
|
| 20 |
+
"mean": [
|
| 21 |
+
9.669540405273438,
|
| 22 |
+
-25.556636810302734,
|
| 23 |
+
27.452985763549805,
|
| 24 |
+
68.78153991699219,
|
| 25 |
+
4.380364418029785
|
| 26 |
+
],
|
| 27 |
+
"std": [
|
| 28 |
+
17.90518569946289,
|
| 29 |
+
54.773555755615234,
|
| 30 |
+
51.100162506103516,
|
| 31 |
+
13.19716739654541,
|
| 32 |
+
8.14745807647705
|
| 33 |
+
],
|
| 34 |
+
"q01": [
|
| 35 |
+
-20.051746368408203,
|
| 36 |
+
-97.77777862548828,
|
| 37 |
+
-64.59074401855469,
|
| 38 |
+
42.641510009765625,
|
| 39 |
+
-19.86318016052246
|
| 40 |
+
],
|
| 41 |
+
"q99": [
|
| 42 |
+
46.01121139526367,
|
| 43 |
+
51.53845977783203,
|
| 44 |
+
96.97509002685547,
|
| 45 |
+
97.56813049316406,
|
| 46 |
+
27.485952377319336
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"gripper": {
|
| 50 |
+
"max": [
|
| 51 |
+
34.90322494506836
|
| 52 |
+
],
|
| 53 |
+
"min": [
|
| 54 |
+
2.1935484409332275
|
| 55 |
+
],
|
| 56 |
+
"mean": [
|
| 57 |
+
10.092109680175781
|
| 58 |
+
],
|
| 59 |
+
"std": [
|
| 60 |
+
9.9507474899292
|
| 61 |
+
],
|
| 62 |
+
"q01": [
|
| 63 |
+
2.1935484409332275
|
| 64 |
+
],
|
| 65 |
+
"q99": [
|
| 66 |
+
32.58064651489258
|
| 67 |
+
]
|
| 68 |
+
}
|
| 69 |
+
},
|
| 70 |
+
"action": {
|
| 71 |
+
"single_arm": {
|
| 72 |
+
"max": [
|
| 73 |
+
50.360015869140625,
|
| 74 |
+
60.167015075683594,
|
| 75 |
+
97.74164581298828,
|
| 76 |
+
99.65694427490234,
|
| 77 |
+
34.145076751708984
|
| 78 |
+
],
|
| 79 |
+
"min": [
|
| 80 |
+
-23.337568283081055,
|
| 81 |
+
-100.0,
|
| 82 |
+
-99.0966567993164,
|
| 83 |
+
38.9365348815918,
|
| 84 |
+
-23.834196090698242
|
| 85 |
+
],
|
| 86 |
+
"mean": [
|
| 87 |
+
9.820510864257812,
|
| 88 |
+
-27.617700576782227,
|
| 89 |
+
24.440105438232422,
|
| 90 |
+
68.3581771850586,
|
| 91 |
+
4.403929710388184
|
| 92 |
+
],
|
| 93 |
+
"std": [
|
| 94 |
+
18.048961639404297,
|
| 95 |
+
53.86724853515625,
|
| 96 |
+
52.31853103637695,
|
| 97 |
+
13.662403106689453,
|
| 98 |
+
8.221379280090332
|
| 99 |
+
],
|
| 100 |
+
"q01": [
|
| 101 |
+
-20.372722625732422,
|
| 102 |
+
-98.9144058227539,
|
| 103 |
+
-66.12466430664062,
|
| 104 |
+
42.02401351928711,
|
| 105 |
+
-20.103626251220703
|
| 106 |
+
],
|
| 107 |
+
"q99": [
|
| 108 |
+
45.95510482788086,
|
| 109 |
+
49.02797512054448,
|
| 110 |
+
97.56097412109375,
|
| 111 |
+
99.57118225097656,
|
| 112 |
+
27.72020721435547
|
| 113 |
+
]
|
| 114 |
+
},
|
| 115 |
+
"gripper": {
|
| 116 |
+
"max": [
|
| 117 |
+
35.219234466552734
|
| 118 |
+
],
|
| 119 |
+
"min": [
|
| 120 |
+
0.07072135806083679
|
| 121 |
+
],
|
| 122 |
+
"mean": [
|
| 123 |
+
8.376781463623047
|
| 124 |
+
],
|
| 125 |
+
"std": [
|
| 126 |
+
10.938384056091309
|
| 127 |
+
],
|
| 128 |
+
"q01": [
|
| 129 |
+
1.2729843854904175
|
| 130 |
+
],
|
| 131 |
+
"q99": [
|
| 132 |
+
32.8147087097168
|
| 133 |
+
]
|
| 134 |
+
}
|
| 135 |
+
}
|
| 136 |
+
},
|
| 137 |
+
"modalities": {
|
| 138 |
+
"video": {
|
| 139 |
+
"wrist": {
|
| 140 |
+
"resolution": [
|
| 141 |
+
640,
|
| 142 |
+
480
|
| 143 |
+
],
|
| 144 |
+
"channels": 3,
|
| 145 |
+
"fps": 30.0
|
| 146 |
+
},
|
| 147 |
+
"side": {
|
| 148 |
+
"resolution": [
|
| 149 |
+
640,
|
| 150 |
+
480
|
| 151 |
+
],
|
| 152 |
+
"channels": 3,
|
| 153 |
+
"fps": 30.0
|
| 154 |
+
}
|
| 155 |
+
},
|
| 156 |
+
"state": {
|
| 157 |
+
"single_arm": {
|
| 158 |
+
"absolute": true,
|
| 159 |
+
"rotation_type": null,
|
| 160 |
+
"shape": [
|
| 161 |
+
5
|
| 162 |
+
],
|
| 163 |
+
"continuous": true
|
| 164 |
+
},
|
| 165 |
+
"gripper": {
|
| 166 |
+
"absolute": true,
|
| 167 |
+
"rotation_type": null,
|
| 168 |
+
"shape": [
|
| 169 |
+
1
|
| 170 |
+
],
|
| 171 |
+
"continuous": true
|
| 172 |
+
}
|
| 173 |
+
},
|
| 174 |
+
"action": {
|
| 175 |
+
"single_arm": {
|
| 176 |
+
"absolute": true,
|
| 177 |
+
"rotation_type": null,
|
| 178 |
+
"shape": [
|
| 179 |
+
5
|
| 180 |
+
],
|
| 181 |
+
"continuous": true
|
| 182 |
+
},
|
| 183 |
+
"gripper": {
|
| 184 |
+
"absolute": true,
|
| 185 |
+
"rotation_type": null,
|
| 186 |
+
"shape": [
|
| 187 |
+
1
|
| 188 |
+
],
|
| 189 |
+
"continuous": true
|
| 190 |
+
}
|
| 191 |
+
}
|
| 192 |
+
},
|
| 193 |
+
"embodiment_tag": "new_embodiment"
|
| 194 |
+
}
|
| 195 |
+
}
|
--comit-message=Finetuned model with 2000 steps (gr00t)/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62d0e3581188632e1ce8c5307ae753654603e5943fe4237c30c10fc8d7be48a3
|
| 3 |
+
size 4999367032
|
--comit-message=Finetuned model with 2000 steps (gr00t)/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3ce80b842cb68c771667f2250bdbae60e8464a0ba46229f6a281aca8f42c444
|
| 3 |
+
size 2586705312
|
--comit-message=Finetuned model with 2000 steps (gr00t)/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
--comit-message=Finetuned model with 2000 steps (gr00t)/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c2b8a3d1bcfb98e68aa361bc283a8ac0326a06ed69c9736a29ff67d310f8f2b
|
| 3 |
+
size 8550720062
|
--comit-message=Finetuned model with 2000 steps (gr00t)/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9324ae866242e523506b335db8fb85287d04edbf8a932bbc246f6328d7bf50c9
|
| 3 |
+
size 14244
|
--comit-message=Finetuned model with 2000 steps (gr00t)/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccf34f8dc96000a297a8a53cfbee51dd50537cf314f74312de55cac0ca70436d
|
| 3 |
+
size 1064
|
--comit-message=Finetuned model with 2000 steps (gr00t)/trainer_state.json
ADDED
|
@@ -0,0 +1,1434 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.5650623885918002,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.017825311942959002,
|
| 14 |
+
"grad_norm": 14.087394714355469,
|
| 15 |
+
"learning_rate": 1.8e-06,
|
| 16 |
+
"loss": 1.0691,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.035650623885918005,
|
| 21 |
+
"grad_norm": 4.053713321685791,
|
| 22 |
+
"learning_rate": 3.8e-06,
|
| 23 |
+
"loss": 0.8746,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.053475935828877004,
|
| 28 |
+
"grad_norm": 4.057157516479492,
|
| 29 |
+
"learning_rate": 5.8e-06,
|
| 30 |
+
"loss": 0.45,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07130124777183601,
|
| 35 |
+
"grad_norm": 3.075242519378662,
|
| 36 |
+
"learning_rate": 7.8e-06,
|
| 37 |
+
"loss": 0.2739,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08912655971479501,
|
| 42 |
+
"grad_norm": 1.5601048469543457,
|
| 43 |
+
"learning_rate": 9.800000000000001e-06,
|
| 44 |
+
"loss": 0.2045,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.10695187165775401,
|
| 49 |
+
"grad_norm": 1.2068922519683838,
|
| 50 |
+
"learning_rate": 1.18e-05,
|
| 51 |
+
"loss": 0.1774,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.12477718360071301,
|
| 56 |
+
"grad_norm": 1.4393501281738281,
|
| 57 |
+
"learning_rate": 1.3800000000000002e-05,
|
| 58 |
+
"loss": 0.146,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.14260249554367202,
|
| 63 |
+
"grad_norm": 1.8730257749557495,
|
| 64 |
+
"learning_rate": 1.58e-05,
|
| 65 |
+
"loss": 0.1489,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16042780748663102,
|
| 70 |
+
"grad_norm": 1.4819114208221436,
|
| 71 |
+
"learning_rate": 1.78e-05,
|
| 72 |
+
"loss": 0.1394,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.17825311942959002,
|
| 77 |
+
"grad_norm": 1.501101016998291,
|
| 78 |
+
"learning_rate": 1.9800000000000004e-05,
|
| 79 |
+
"loss": 0.1231,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.19607843137254902,
|
| 84 |
+
"grad_norm": 0.8101649880409241,
|
| 85 |
+
"learning_rate": 2.18e-05,
|
| 86 |
+
"loss": 0.1124,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.21390374331550802,
|
| 91 |
+
"grad_norm": 1.288984775543213,
|
| 92 |
+
"learning_rate": 2.38e-05,
|
| 93 |
+
"loss": 0.1062,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.23172905525846701,
|
| 98 |
+
"grad_norm": 1.209707260131836,
|
| 99 |
+
"learning_rate": 2.58e-05,
|
| 100 |
+
"loss": 0.1076,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.24955436720142601,
|
| 105 |
+
"grad_norm": 1.1386358737945557,
|
| 106 |
+
"learning_rate": 2.7800000000000005e-05,
|
| 107 |
+
"loss": 0.1014,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.26737967914438504,
|
| 112 |
+
"grad_norm": 1.262028694152832,
|
| 113 |
+
"learning_rate": 2.98e-05,
|
| 114 |
+
"loss": 0.098,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.28520499108734404,
|
| 119 |
+
"grad_norm": 0.5931951999664307,
|
| 120 |
+
"learning_rate": 3.18e-05,
|
| 121 |
+
"loss": 0.089,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.30303030303030304,
|
| 126 |
+
"grad_norm": 1.0128452777862549,
|
| 127 |
+
"learning_rate": 3.38e-05,
|
| 128 |
+
"loss": 0.0807,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.32085561497326204,
|
| 133 |
+
"grad_norm": 0.5979309678077698,
|
| 134 |
+
"learning_rate": 3.58e-05,
|
| 135 |
+
"loss": 0.0889,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.33868092691622104,
|
| 140 |
+
"grad_norm": 0.8895788788795471,
|
| 141 |
+
"learning_rate": 3.7800000000000004e-05,
|
| 142 |
+
"loss": 0.0886,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.35650623885918004,
|
| 147 |
+
"grad_norm": 1.1014182567596436,
|
| 148 |
+
"learning_rate": 3.9800000000000005e-05,
|
| 149 |
+
"loss": 0.0842,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.37433155080213903,
|
| 154 |
+
"grad_norm": 0.5955915451049805,
|
| 155 |
+
"learning_rate": 4.18e-05,
|
| 156 |
+
"loss": 0.0842,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.39215686274509803,
|
| 161 |
+
"grad_norm": 0.9140474796295166,
|
| 162 |
+
"learning_rate": 4.38e-05,
|
| 163 |
+
"loss": 0.0785,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.40998217468805703,
|
| 168 |
+
"grad_norm": 0.7342748641967773,
|
| 169 |
+
"learning_rate": 4.58e-05,
|
| 170 |
+
"loss": 0.067,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.42780748663101603,
|
| 175 |
+
"grad_norm": 0.8894127607345581,
|
| 176 |
+
"learning_rate": 4.78e-05,
|
| 177 |
+
"loss": 0.0692,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.44563279857397503,
|
| 182 |
+
"grad_norm": 0.9929222464561462,
|
| 183 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 184 |
+
"loss": 0.0681,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.46345811051693403,
|
| 189 |
+
"grad_norm": 0.7159563302993774,
|
| 190 |
+
"learning_rate": 5.1800000000000005e-05,
|
| 191 |
+
"loss": 0.0638,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.48128342245989303,
|
| 196 |
+
"grad_norm": 0.7639392018318176,
|
| 197 |
+
"learning_rate": 5.380000000000001e-05,
|
| 198 |
+
"loss": 0.0754,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.49910873440285203,
|
| 203 |
+
"grad_norm": 0.6825873851776123,
|
| 204 |
+
"learning_rate": 5.580000000000001e-05,
|
| 205 |
+
"loss": 0.0723,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5169340463458111,
|
| 210 |
+
"grad_norm": 0.49610263109207153,
|
| 211 |
+
"learning_rate": 5.7799999999999995e-05,
|
| 212 |
+
"loss": 0.0746,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5347593582887701,
|
| 217 |
+
"grad_norm": 0.8349130153656006,
|
| 218 |
+
"learning_rate": 5.9800000000000003e-05,
|
| 219 |
+
"loss": 0.0641,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5525846702317291,
|
| 224 |
+
"grad_norm": 1.0360268354415894,
|
| 225 |
+
"learning_rate": 6.18e-05,
|
| 226 |
+
"loss": 0.0774,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.5704099821746881,
|
| 231 |
+
"grad_norm": 0.879499077796936,
|
| 232 |
+
"learning_rate": 6.38e-05,
|
| 233 |
+
"loss": 0.0679,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5882352941176471,
|
| 238 |
+
"grad_norm": 1.0108102560043335,
|
| 239 |
+
"learning_rate": 6.58e-05,
|
| 240 |
+
"loss": 0.0632,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6060606060606061,
|
| 245 |
+
"grad_norm": 1.0805786848068237,
|
| 246 |
+
"learning_rate": 6.780000000000001e-05,
|
| 247 |
+
"loss": 0.0667,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6238859180035651,
|
| 252 |
+
"grad_norm": 1.0161969661712646,
|
| 253 |
+
"learning_rate": 6.98e-05,
|
| 254 |
+
"loss": 0.0626,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6417112299465241,
|
| 259 |
+
"grad_norm": 0.593278169631958,
|
| 260 |
+
"learning_rate": 7.18e-05,
|
| 261 |
+
"loss": 0.0585,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6595365418894831,
|
| 266 |
+
"grad_norm": 0.8747594952583313,
|
| 267 |
+
"learning_rate": 7.38e-05,
|
| 268 |
+
"loss": 0.0643,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.6773618538324421,
|
| 273 |
+
"grad_norm": 0.8220623135566711,
|
| 274 |
+
"learning_rate": 7.58e-05,
|
| 275 |
+
"loss": 0.0571,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6951871657754011,
|
| 280 |
+
"grad_norm": 0.5637218356132507,
|
| 281 |
+
"learning_rate": 7.780000000000001e-05,
|
| 282 |
+
"loss": 0.0553,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7130124777183601,
|
| 287 |
+
"grad_norm": 0.5024718642234802,
|
| 288 |
+
"learning_rate": 7.98e-05,
|
| 289 |
+
"loss": 0.0545,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7308377896613191,
|
| 294 |
+
"grad_norm": 0.9930310249328613,
|
| 295 |
+
"learning_rate": 8.18e-05,
|
| 296 |
+
"loss": 0.0591,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7486631016042781,
|
| 301 |
+
"grad_norm": 0.6868970394134521,
|
| 302 |
+
"learning_rate": 8.38e-05,
|
| 303 |
+
"loss": 0.0505,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.7664884135472371,
|
| 308 |
+
"grad_norm": 0.5700907111167908,
|
| 309 |
+
"learning_rate": 8.58e-05,
|
| 310 |
+
"loss": 0.0544,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.7843137254901961,
|
| 315 |
+
"grad_norm": 0.9365397691726685,
|
| 316 |
+
"learning_rate": 8.78e-05,
|
| 317 |
+
"loss": 0.057,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8021390374331551,
|
| 322 |
+
"grad_norm": 0.798393189907074,
|
| 323 |
+
"learning_rate": 8.98e-05,
|
| 324 |
+
"loss": 0.0547,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8199643493761141,
|
| 329 |
+
"grad_norm": 1.0544344186782837,
|
| 330 |
+
"learning_rate": 9.180000000000001e-05,
|
| 331 |
+
"loss": 0.0598,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8377896613190731,
|
| 336 |
+
"grad_norm": 0.7224001288414001,
|
| 337 |
+
"learning_rate": 9.38e-05,
|
| 338 |
+
"loss": 0.059,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.8556149732620321,
|
| 343 |
+
"grad_norm": 0.7041344046592712,
|
| 344 |
+
"learning_rate": 9.58e-05,
|
| 345 |
+
"loss": 0.0596,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.8734402852049911,
|
| 350 |
+
"grad_norm": 0.5711538195610046,
|
| 351 |
+
"learning_rate": 9.78e-05,
|
| 352 |
+
"loss": 0.0548,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.8912655971479501,
|
| 357 |
+
"grad_norm": 0.8479695320129395,
|
| 358 |
+
"learning_rate": 9.98e-05,
|
| 359 |
+
"loss": 0.06,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9090909090909091,
|
| 364 |
+
"grad_norm": 0.7224346995353699,
|
| 365 |
+
"learning_rate": 9.9999778549206e-05,
|
| 366 |
+
"loss": 0.0571,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.9269162210338681,
|
| 371 |
+
"grad_norm": 0.9372037053108215,
|
| 372 |
+
"learning_rate": 9.999901304280685e-05,
|
| 373 |
+
"loss": 0.0541,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9447415329768271,
|
| 378 |
+
"grad_norm": 0.7346435785293579,
|
| 379 |
+
"learning_rate": 9.999770075521164e-05,
|
| 380 |
+
"loss": 0.0612,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.9625668449197861,
|
| 385 |
+
"grad_norm": 0.6650673747062683,
|
| 386 |
+
"learning_rate": 9.99958417007713e-05,
|
| 387 |
+
"loss": 0.0508,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.9803921568627451,
|
| 392 |
+
"grad_norm": 0.7607464790344238,
|
| 393 |
+
"learning_rate": 9.999343589981615e-05,
|
| 394 |
+
"loss": 0.0526,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.9982174688057041,
|
| 399 |
+
"grad_norm": 0.5234432220458984,
|
| 400 |
+
"learning_rate": 9.999048337865568e-05,
|
| 401 |
+
"loss": 0.0496,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0160427807486632,
|
| 406 |
+
"grad_norm": 0.5495169758796692,
|
| 407 |
+
"learning_rate": 9.998698416957815e-05,
|
| 408 |
+
"loss": 0.0545,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0338680926916222,
|
| 413 |
+
"grad_norm": 1.1648725271224976,
|
| 414 |
+
"learning_rate": 9.998293831085037e-05,
|
| 415 |
+
"loss": 0.0558,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.0516934046345812,
|
| 420 |
+
"grad_norm": 0.6070398092269897,
|
| 421 |
+
"learning_rate": 9.997834584671719e-05,
|
| 422 |
+
"loss": 0.0519,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.0695187165775402,
|
| 427 |
+
"grad_norm": 0.6913730502128601,
|
| 428 |
+
"learning_rate": 9.997320682740107e-05,
|
| 429 |
+
"loss": 0.0551,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.0873440285204992,
|
| 434 |
+
"grad_norm": 0.6825369000434875,
|
| 435 |
+
"learning_rate": 9.996752130910149e-05,
|
| 436 |
+
"loss": 0.0545,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1051693404634582,
|
| 441 |
+
"grad_norm": 0.5194472074508667,
|
| 442 |
+
"learning_rate": 9.99612893539944e-05,
|
| 443 |
+
"loss": 0.0458,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1229946524064172,
|
| 448 |
+
"grad_norm": 0.5538597106933594,
|
| 449 |
+
"learning_rate": 9.995451103023144e-05,
|
| 450 |
+
"loss": 0.0463,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.1408199643493762,
|
| 455 |
+
"grad_norm": 0.7499181628227234,
|
| 456 |
+
"learning_rate": 9.994718641193928e-05,
|
| 457 |
+
"loss": 0.0436,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.1586452762923352,
|
| 462 |
+
"grad_norm": 0.6800760626792908,
|
| 463 |
+
"learning_rate": 9.993931557921874e-05,
|
| 464 |
+
"loss": 0.0524,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.1764705882352942,
|
| 469 |
+
"grad_norm": 1.2449172735214233,
|
| 470 |
+
"learning_rate": 9.993089861814402e-05,
|
| 471 |
+
"loss": 0.0521,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.1942959001782532,
|
| 476 |
+
"grad_norm": 0.4747939705848694,
|
| 477 |
+
"learning_rate": 9.992193562076166e-05,
|
| 478 |
+
"loss": 0.0601,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2121212121212122,
|
| 483 |
+
"grad_norm": 0.5693209767341614,
|
| 484 |
+
"learning_rate": 9.991242668508954e-05,
|
| 485 |
+
"loss": 0.0529,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.2299465240641712,
|
| 490 |
+
"grad_norm": 0.48745831847190857,
|
| 491 |
+
"learning_rate": 9.990237191511587e-05,
|
| 492 |
+
"loss": 0.0444,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2477718360071302,
|
| 497 |
+
"grad_norm": 0.47945934534072876,
|
| 498 |
+
"learning_rate": 9.989177142079802e-05,
|
| 499 |
+
"loss": 0.058,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.2655971479500892,
|
| 504 |
+
"grad_norm": 0.7471551895141602,
|
| 505 |
+
"learning_rate": 9.988062531806126e-05,
|
| 506 |
+
"loss": 0.0517,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.2834224598930482,
|
| 511 |
+
"grad_norm": 0.313781201839447,
|
| 512 |
+
"learning_rate": 9.986893372879762e-05,
|
| 513 |
+
"loss": 0.0462,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3012477718360071,
|
| 518 |
+
"grad_norm": 0.3812905550003052,
|
| 519 |
+
"learning_rate": 9.985669678086443e-05,
|
| 520 |
+
"loss": 0.0543,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3190730837789661,
|
| 525 |
+
"grad_norm": 0.6839037537574768,
|
| 526 |
+
"learning_rate": 9.984391460808298e-05,
|
| 527 |
+
"loss": 0.0443,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.3368983957219251,
|
| 532 |
+
"grad_norm": 0.5126590728759766,
|
| 533 |
+
"learning_rate": 9.983058735023709e-05,
|
| 534 |
+
"loss": 0.0435,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3547237076648841,
|
| 539 |
+
"grad_norm": 0.48188620805740356,
|
| 540 |
+
"learning_rate": 9.98167151530715e-05,
|
| 541 |
+
"loss": 0.0433,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.3725490196078431,
|
| 546 |
+
"grad_norm": 0.532317042350769,
|
| 547 |
+
"learning_rate": 9.980229816829034e-05,
|
| 548 |
+
"loss": 0.0423,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.3903743315508021,
|
| 553 |
+
"grad_norm": 0.4482864439487457,
|
| 554 |
+
"learning_rate": 9.978733655355544e-05,
|
| 555 |
+
"loss": 0.0434,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4081996434937611,
|
| 560 |
+
"grad_norm": 0.5437975525856018,
|
| 561 |
+
"learning_rate": 9.977183047248464e-05,
|
| 562 |
+
"loss": 0.044,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.4260249554367201,
|
| 567 |
+
"grad_norm": 0.46547406911849976,
|
| 568 |
+
"learning_rate": 9.975578009464992e-05,
|
| 569 |
+
"loss": 0.0487,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.4438502673796791,
|
| 574 |
+
"grad_norm": 0.8020538091659546,
|
| 575 |
+
"learning_rate": 9.97391855955757e-05,
|
| 576 |
+
"loss": 0.0502,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.4616755793226381,
|
| 581 |
+
"grad_norm": 0.5699297189712524,
|
| 582 |
+
"learning_rate": 9.972204715673669e-05,
|
| 583 |
+
"loss": 0.0436,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.4795008912655971,
|
| 588 |
+
"grad_norm": 0.3416242003440857,
|
| 589 |
+
"learning_rate": 9.970436496555617e-05,
|
| 590 |
+
"loss": 0.0493,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.4973262032085561,
|
| 595 |
+
"grad_norm": 0.41045910120010376,
|
| 596 |
+
"learning_rate": 9.968613921540373e-05,
|
| 597 |
+
"loss": 0.0405,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5151515151515151,
|
| 602 |
+
"grad_norm": 0.5546694397926331,
|
| 603 |
+
"learning_rate": 9.966737010559326e-05,
|
| 604 |
+
"loss": 0.052,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.5329768270944741,
|
| 609 |
+
"grad_norm": 0.8029137849807739,
|
| 610 |
+
"learning_rate": 9.964805784138072e-05,
|
| 611 |
+
"loss": 0.0438,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.5508021390374331,
|
| 616 |
+
"grad_norm": 0.45088744163513184,
|
| 617 |
+
"learning_rate": 9.962820263396195e-05,
|
| 618 |
+
"loss": 0.0436,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.5686274509803921,
|
| 623 |
+
"grad_norm": 0.43348464369773865,
|
| 624 |
+
"learning_rate": 9.960780470047033e-05,
|
| 625 |
+
"loss": 0.0426,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.5864527629233511,
|
| 630 |
+
"grad_norm": 0.696536123752594,
|
| 631 |
+
"learning_rate": 9.958686426397437e-05,
|
| 632 |
+
"loss": 0.0434,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.6042780748663101,
|
| 637 |
+
"grad_norm": 0.3600558042526245,
|
| 638 |
+
"learning_rate": 9.956538155347534e-05,
|
| 639 |
+
"loss": 0.0444,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.6221033868092691,
|
| 644 |
+
"grad_norm": 0.28768786787986755,
|
| 645 |
+
"learning_rate": 9.95433568039047e-05,
|
| 646 |
+
"loss": 0.0354,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6399286987522281,
|
| 651 |
+
"grad_norm": 0.626501202583313,
|
| 652 |
+
"learning_rate": 9.952079025612162e-05,
|
| 653 |
+
"loss": 0.0433,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.6577540106951871,
|
| 658 |
+
"grad_norm": 0.41641244292259216,
|
| 659 |
+
"learning_rate": 9.949768215691022e-05,
|
| 660 |
+
"loss": 0.0408,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.6755793226381461,
|
| 665 |
+
"grad_norm": 0.46979671716690063,
|
| 666 |
+
"learning_rate": 9.9474032758977e-05,
|
| 667 |
+
"loss": 0.0386,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.6934046345811051,
|
| 672 |
+
"grad_norm": 0.40619802474975586,
|
| 673 |
+
"learning_rate": 9.944984232094794e-05,
|
| 674 |
+
"loss": 0.0452,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.7112299465240641,
|
| 679 |
+
"grad_norm": 0.5174899101257324,
|
| 680 |
+
"learning_rate": 9.942511110736584e-05,
|
| 681 |
+
"loss": 0.0371,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.7290552584670231,
|
| 686 |
+
"grad_norm": 0.5482277870178223,
|
| 687 |
+
"learning_rate": 9.939983938868726e-05,
|
| 688 |
+
"loss": 0.0382,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.7468805704099821,
|
| 693 |
+
"grad_norm": 0.5827427506446838,
|
| 694 |
+
"learning_rate": 9.93740274412797e-05,
|
| 695 |
+
"loss": 0.0402,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.7647058823529411,
|
| 700 |
+
"grad_norm": 0.5574763417243958,
|
| 701 |
+
"learning_rate": 9.934767554741846e-05,
|
| 702 |
+
"loss": 0.0437,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.7825311942959001,
|
| 707 |
+
"grad_norm": 0.4984002709388733,
|
| 708 |
+
"learning_rate": 9.932078399528361e-05,
|
| 709 |
+
"loss": 0.0417,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8003565062388591,
|
| 714 |
+
"grad_norm": 0.5287971496582031,
|
| 715 |
+
"learning_rate": 9.929335307895689e-05,
|
| 716 |
+
"loss": 0.0462,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.8181818181818183,
|
| 721 |
+
"grad_norm": 0.41133537888526917,
|
| 722 |
+
"learning_rate": 9.926538309841839e-05,
|
| 723 |
+
"loss": 0.0398,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.8360071301247771,
|
| 728 |
+
"grad_norm": 0.5262690186500549,
|
| 729 |
+
"learning_rate": 9.923687435954334e-05,
|
| 730 |
+
"loss": 0.0399,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.8538324420677363,
|
| 735 |
+
"grad_norm": 0.641912579536438,
|
| 736 |
+
"learning_rate": 9.920782717409873e-05,
|
| 737 |
+
"loss": 0.0424,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.8716577540106951,
|
| 742 |
+
"grad_norm": 0.42638346552848816,
|
| 743 |
+
"learning_rate": 9.917824185973994e-05,
|
| 744 |
+
"loss": 0.034,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.8894830659536543,
|
| 749 |
+
"grad_norm": 0.5523088574409485,
|
| 750 |
+
"learning_rate": 9.914811874000723e-05,
|
| 751 |
+
"loss": 0.0439,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.9073083778966131,
|
| 756 |
+
"grad_norm": 0.3323904573917389,
|
| 757 |
+
"learning_rate": 9.911745814432218e-05,
|
| 758 |
+
"loss": 0.0469,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.9251336898395723,
|
| 763 |
+
"grad_norm": 0.44197213649749756,
|
| 764 |
+
"learning_rate": 9.90862604079842e-05,
|
| 765 |
+
"loss": 0.046,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.9429590017825311,
|
| 770 |
+
"grad_norm": 0.587875247001648,
|
| 771 |
+
"learning_rate": 9.90545258721667e-05,
|
| 772 |
+
"loss": 0.0324,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.9607843137254903,
|
| 777 |
+
"grad_norm": 0.40099313855171204,
|
| 778 |
+
"learning_rate": 9.90222548839135e-05,
|
| 779 |
+
"loss": 0.039,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.9786096256684491,
|
| 784 |
+
"grad_norm": 0.5058305263519287,
|
| 785 |
+
"learning_rate": 9.898944779613495e-05,
|
| 786 |
+
"loss": 0.0402,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.9964349376114083,
|
| 791 |
+
"grad_norm": 0.43186384439468384,
|
| 792 |
+
"learning_rate": 9.89561049676041e-05,
|
| 793 |
+
"loss": 0.0397,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.014260249554367,
|
| 798 |
+
"grad_norm": 0.536444902420044,
|
| 799 |
+
"learning_rate": 9.89222267629528e-05,
|
| 800 |
+
"loss": 0.0385,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.0320855614973263,
|
| 805 |
+
"grad_norm": 0.5141022205352783,
|
| 806 |
+
"learning_rate": 9.888781355266763e-05,
|
| 807 |
+
"loss": 0.0424,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.049910873440285,
|
| 812 |
+
"grad_norm": 0.6025420427322388,
|
| 813 |
+
"learning_rate": 9.885286571308598e-05,
|
| 814 |
+
"loss": 0.0452,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.0677361853832443,
|
| 819 |
+
"grad_norm": 0.5754320025444031,
|
| 820 |
+
"learning_rate": 9.881738362639182e-05,
|
| 821 |
+
"loss": 0.041,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.085561497326203,
|
| 826 |
+
"grad_norm": 0.35027605295181274,
|
| 827 |
+
"learning_rate": 9.878136768061154e-05,
|
| 828 |
+
"loss": 0.0361,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.1033868092691623,
|
| 833 |
+
"grad_norm": 0.6212054491043091,
|
| 834 |
+
"learning_rate": 9.874481826960979e-05,
|
| 835 |
+
"loss": 0.0372,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.121212121212121,
|
| 840 |
+
"grad_norm": 0.28216296434402466,
|
| 841 |
+
"learning_rate": 9.870773579308503e-05,
|
| 842 |
+
"loss": 0.0355,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.1390374331550803,
|
| 847 |
+
"grad_norm": 0.41768956184387207,
|
| 848 |
+
"learning_rate": 9.867012065656533e-05,
|
| 849 |
+
"loss": 0.0405,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.156862745098039,
|
| 854 |
+
"grad_norm": 0.46785035729408264,
|
| 855 |
+
"learning_rate": 9.863197327140376e-05,
|
| 856 |
+
"loss": 0.0402,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.1746880570409983,
|
| 861 |
+
"grad_norm": 0.6558189988136292,
|
| 862 |
+
"learning_rate": 9.859329405477403e-05,
|
| 863 |
+
"loss": 0.0322,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.192513368983957,
|
| 868 |
+
"grad_norm": 0.6351569294929504,
|
| 869 |
+
"learning_rate": 9.855408342966585e-05,
|
| 870 |
+
"loss": 0.0423,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.2103386809269163,
|
| 875 |
+
"grad_norm": 0.5595504641532898,
|
| 876 |
+
"learning_rate": 9.851434182488033e-05,
|
| 877 |
+
"loss": 0.0403,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.228163992869875,
|
| 882 |
+
"grad_norm": 0.5321756601333618,
|
| 883 |
+
"learning_rate": 9.84740696750253e-05,
|
| 884 |
+
"loss": 0.0432,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.2459893048128343,
|
| 889 |
+
"grad_norm": 0.7758764028549194,
|
| 890 |
+
"learning_rate": 9.843326742051055e-05,
|
| 891 |
+
"loss": 0.0376,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.263814616755793,
|
| 896 |
+
"grad_norm": 0.5130152106285095,
|
| 897 |
+
"learning_rate": 9.839193550754297e-05,
|
| 898 |
+
"loss": 0.0444,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.2816399286987523,
|
| 903 |
+
"grad_norm": 0.6629639863967896,
|
| 904 |
+
"learning_rate": 9.835007438812177e-05,
|
| 905 |
+
"loss": 0.039,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.299465240641711,
|
| 910 |
+
"grad_norm": 0.3983469605445862,
|
| 911 |
+
"learning_rate": 9.830768452003341e-05,
|
| 912 |
+
"loss": 0.0436,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.3172905525846703,
|
| 917 |
+
"grad_norm": 0.4088013470172882,
|
| 918 |
+
"learning_rate": 9.826476636684671e-05,
|
| 919 |
+
"loss": 0.0392,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.335115864527629,
|
| 924 |
+
"grad_norm": 0.5007789134979248,
|
| 925 |
+
"learning_rate": 9.822132039790773e-05,
|
| 926 |
+
"loss": 0.0368,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.3529411764705883,
|
| 931 |
+
"grad_norm": 0.47945436835289,
|
| 932 |
+
"learning_rate": 9.817734708833461e-05,
|
| 933 |
+
"loss": 0.0384,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.370766488413547,
|
| 938 |
+
"grad_norm": 0.553306519985199,
|
| 939 |
+
"learning_rate": 9.813284691901243e-05,
|
| 940 |
+
"loss": 0.0354,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.3885918003565063,
|
| 945 |
+
"grad_norm": 0.39307186007499695,
|
| 946 |
+
"learning_rate": 9.808782037658792e-05,
|
| 947 |
+
"loss": 0.035,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.406417112299465,
|
| 952 |
+
"grad_norm": 0.4565328061580658,
|
| 953 |
+
"learning_rate": 9.804226795346411e-05,
|
| 954 |
+
"loss": 0.0342,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.4242424242424243,
|
| 959 |
+
"grad_norm": 0.3856677711009979,
|
| 960 |
+
"learning_rate": 9.799619014779503e-05,
|
| 961 |
+
"loss": 0.0344,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.442067736185383,
|
| 966 |
+
"grad_norm": 0.45724722743034363,
|
| 967 |
+
"learning_rate": 9.794958746348013e-05,
|
| 968 |
+
"loss": 0.0341,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.4598930481283423,
|
| 973 |
+
"grad_norm": 0.47785642743110657,
|
| 974 |
+
"learning_rate": 9.790246041015896e-05,
|
| 975 |
+
"loss": 0.0373,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.477718360071301,
|
| 980 |
+
"grad_norm": 0.25403615832328796,
|
| 981 |
+
"learning_rate": 9.785480950320538e-05,
|
| 982 |
+
"loss": 0.0339,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.4955436720142603,
|
| 987 |
+
"grad_norm": 0.5119685530662537,
|
| 988 |
+
"learning_rate": 9.78066352637221e-05,
|
| 989 |
+
"loss": 0.0377,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.5133689839572195,
|
| 994 |
+
"grad_norm": 0.28113624453544617,
|
| 995 |
+
"learning_rate": 9.775793821853488e-05,
|
| 996 |
+
"loss": 0.0356,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.5311942959001783,
|
| 1001 |
+
"grad_norm": 0.44586440920829773,
|
| 1002 |
+
"learning_rate": 9.77087189001868e-05,
|
| 1003 |
+
"loss": 0.0374,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.549019607843137,
|
| 1008 |
+
"grad_norm": 0.4238766133785248,
|
| 1009 |
+
"learning_rate": 9.765897784693243e-05,
|
| 1010 |
+
"loss": 0.0313,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.5668449197860963,
|
| 1015 |
+
"grad_norm": 0.39358988404273987,
|
| 1016 |
+
"learning_rate": 9.760871560273197e-05,
|
| 1017 |
+
"loss": 0.0354,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.5846702317290555,
|
| 1022 |
+
"grad_norm": 0.25771698355674744,
|
| 1023 |
+
"learning_rate": 9.755793271724526e-05,
|
| 1024 |
+
"loss": 0.0347,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.6024955436720143,
|
| 1029 |
+
"grad_norm": 0.42045313119888306,
|
| 1030 |
+
"learning_rate": 9.750662974582584e-05,
|
| 1031 |
+
"loss": 0.0346,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.620320855614973,
|
| 1036 |
+
"grad_norm": 0.3756962716579437,
|
| 1037 |
+
"learning_rate": 9.745480724951473e-05,
|
| 1038 |
+
"loss": 0.0331,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.6381461675579323,
|
| 1043 |
+
"grad_norm": 0.5593686103820801,
|
| 1044 |
+
"learning_rate": 9.740246579503447e-05,
|
| 1045 |
+
"loss": 0.0414,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.6559714795008915,
|
| 1050 |
+
"grad_norm": 0.43153268098831177,
|
| 1051 |
+
"learning_rate": 9.734960595478284e-05,
|
| 1052 |
+
"loss": 0.034,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.6737967914438503,
|
| 1057 |
+
"grad_norm": 0.6576313972473145,
|
| 1058 |
+
"learning_rate": 9.729622830682657e-05,
|
| 1059 |
+
"loss": 0.0427,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.691622103386809,
|
| 1064 |
+
"grad_norm": 0.35079020261764526,
|
| 1065 |
+
"learning_rate": 9.724233343489504e-05,
|
| 1066 |
+
"loss": 0.0401,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.7094474153297683,
|
| 1071 |
+
"grad_norm": 0.49158981442451477,
|
| 1072 |
+
"learning_rate": 9.718792192837396e-05,
|
| 1073 |
+
"loss": 0.0398,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.7272727272727275,
|
| 1078 |
+
"grad_norm": 0.3968684673309326,
|
| 1079 |
+
"learning_rate": 9.713299438229886e-05,
|
| 1080 |
+
"loss": 0.0381,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.7450980392156863,
|
| 1085 |
+
"grad_norm": 0.45859870314598083,
|
| 1086 |
+
"learning_rate": 9.707755139734855e-05,
|
| 1087 |
+
"loss": 0.0427,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.762923351158645,
|
| 1092 |
+
"grad_norm": 0.37702837586402893,
|
| 1093 |
+
"learning_rate": 9.702159357983866e-05,
|
| 1094 |
+
"loss": 0.0356,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.7807486631016043,
|
| 1099 |
+
"grad_norm": 0.3177683651447296,
|
| 1100 |
+
"learning_rate": 9.696512154171492e-05,
|
| 1101 |
+
"loss": 0.0342,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.7985739750445635,
|
| 1106 |
+
"grad_norm": 0.42819589376449585,
|
| 1107 |
+
"learning_rate": 9.690813590054645e-05,
|
| 1108 |
+
"loss": 0.0334,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.8163992869875223,
|
| 1113 |
+
"grad_norm": 0.36810302734375,
|
| 1114 |
+
"learning_rate": 9.685063727951914e-05,
|
| 1115 |
+
"loss": 0.0326,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.834224598930481,
|
| 1120 |
+
"grad_norm": 0.31808367371559143,
|
| 1121 |
+
"learning_rate": 9.679262630742865e-05,
|
| 1122 |
+
"loss": 0.0371,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.8520499108734403,
|
| 1127 |
+
"grad_norm": 0.3115294277667999,
|
| 1128 |
+
"learning_rate": 9.673410361867373e-05,
|
| 1129 |
+
"loss": 0.0323,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.8698752228163995,
|
| 1134 |
+
"grad_norm": 0.31751394271850586,
|
| 1135 |
+
"learning_rate": 9.667506985324909e-05,
|
| 1136 |
+
"loss": 0.0326,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.8877005347593583,
|
| 1141 |
+
"grad_norm": 0.4295711815357208,
|
| 1142 |
+
"learning_rate": 9.661552565673855e-05,
|
| 1143 |
+
"loss": 0.033,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.905525846702317,
|
| 1148 |
+
"grad_norm": 0.5548558831214905,
|
| 1149 |
+
"learning_rate": 9.655547168030789e-05,
|
| 1150 |
+
"loss": 0.0371,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.9233511586452763,
|
| 1155 |
+
"grad_norm": 0.6585544943809509,
|
| 1156 |
+
"learning_rate": 9.649490858069777e-05,
|
| 1157 |
+
"loss": 0.0373,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.9411764705882355,
|
| 1162 |
+
"grad_norm": 0.460356742143631,
|
| 1163 |
+
"learning_rate": 9.643383702021658e-05,
|
| 1164 |
+
"loss": 0.0395,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.9590017825311943,
|
| 1169 |
+
"grad_norm": 0.5966951251029968,
|
| 1170 |
+
"learning_rate": 9.637225766673307e-05,
|
| 1171 |
+
"loss": 0.0307,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.976827094474153,
|
| 1176 |
+
"grad_norm": 0.49974170327186584,
|
| 1177 |
+
"learning_rate": 9.631017119366922e-05,
|
| 1178 |
+
"loss": 0.0345,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.9946524064171123,
|
| 1183 |
+
"grad_norm": 0.350139856338501,
|
| 1184 |
+
"learning_rate": 9.624757827999273e-05,
|
| 1185 |
+
"loss": 0.0337,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 3.0124777183600715,
|
| 1190 |
+
"grad_norm": 0.6313063502311707,
|
| 1191 |
+
"learning_rate": 9.618447961020971e-05,
|
| 1192 |
+
"loss": 0.037,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 3.0303030303030303,
|
| 1197 |
+
"grad_norm": 0.35113832354545593,
|
| 1198 |
+
"learning_rate": 9.612087587435707e-05,
|
| 1199 |
+
"loss": 0.0311,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 3.0481283422459895,
|
| 1204 |
+
"grad_norm": 0.5325800180435181,
|
| 1205 |
+
"learning_rate": 9.605676776799508e-05,
|
| 1206 |
+
"loss": 0.0349,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 3.0659536541889483,
|
| 1211 |
+
"grad_norm": 0.34630802273750305,
|
| 1212 |
+
"learning_rate": 9.599215599219973e-05,
|
| 1213 |
+
"loss": 0.0337,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 3.0837789661319075,
|
| 1218 |
+
"grad_norm": 0.23568764328956604,
|
| 1219 |
+
"learning_rate": 9.592704125355505e-05,
|
| 1220 |
+
"loss": 0.037,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 3.1016042780748663,
|
| 1225 |
+
"grad_norm": 0.37770676612854004,
|
| 1226 |
+
"learning_rate": 9.586142426414538e-05,
|
| 1227 |
+
"loss": 0.0373,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 3.1194295900178255,
|
| 1232 |
+
"grad_norm": 0.4886269271373749,
|
| 1233 |
+
"learning_rate": 9.57953057415476e-05,
|
| 1234 |
+
"loss": 0.0368,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 3.1372549019607843,
|
| 1239 |
+
"grad_norm": 0.30315762758255005,
|
| 1240 |
+
"learning_rate": 9.572868640882328e-05,
|
| 1241 |
+
"loss": 0.0386,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.1550802139037435,
|
| 1246 |
+
"grad_norm": 0.5830493569374084,
|
| 1247 |
+
"learning_rate": 9.56615669945108e-05,
|
| 1248 |
+
"loss": 0.0323,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 3.1729055258467023,
|
| 1253 |
+
"grad_norm": 0.5629374980926514,
|
| 1254 |
+
"learning_rate": 9.55939482326173e-05,
|
| 1255 |
+
"loss": 0.0367,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 3.1907308377896615,
|
| 1260 |
+
"grad_norm": 0.38798198103904724,
|
| 1261 |
+
"learning_rate": 9.552583086261069e-05,
|
| 1262 |
+
"loss": 0.0388,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 3.2085561497326203,
|
| 1267 |
+
"grad_norm": 0.43268340826034546,
|
| 1268 |
+
"learning_rate": 9.545721562941168e-05,
|
| 1269 |
+
"loss": 0.0365,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 3.2263814616755795,
|
| 1274 |
+
"grad_norm": 0.43141889572143555,
|
| 1275 |
+
"learning_rate": 9.538810328338543e-05,
|
| 1276 |
+
"loss": 0.0348,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 3.2442067736185383,
|
| 1281 |
+
"grad_norm": 0.28885987401008606,
|
| 1282 |
+
"learning_rate": 9.531849458033349e-05,
|
| 1283 |
+
"loss": 0.0307,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 3.2620320855614975,
|
| 1288 |
+
"grad_norm": 0.3028181493282318,
|
| 1289 |
+
"learning_rate": 9.524839028148547e-05,
|
| 1290 |
+
"loss": 0.0337,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 3.2798573975044563,
|
| 1295 |
+
"grad_norm": 0.4353356957435608,
|
| 1296 |
+
"learning_rate": 9.517779115349077e-05,
|
| 1297 |
+
"loss": 0.0347,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 3.2976827094474155,
|
| 1302 |
+
"grad_norm": 0.339870810508728,
|
| 1303 |
+
"learning_rate": 9.510669796841014e-05,
|
| 1304 |
+
"loss": 0.033,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 3.3155080213903743,
|
| 1309 |
+
"grad_norm": 0.5104774236679077,
|
| 1310 |
+
"learning_rate": 9.503511150370727e-05,
|
| 1311 |
+
"loss": 0.0368,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 3.3333333333333335,
|
| 1316 |
+
"grad_norm": 0.4100686013698578,
|
| 1317 |
+
"learning_rate": 9.496303254224024e-05,
|
| 1318 |
+
"loss": 0.0391,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.3511586452762923,
|
| 1323 |
+
"grad_norm": 0.5063575506210327,
|
| 1324 |
+
"learning_rate": 9.489046187225306e-05,
|
| 1325 |
+
"loss": 0.0368,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 3.3689839572192515,
|
| 1330 |
+
"grad_norm": 0.2611795663833618,
|
| 1331 |
+
"learning_rate": 9.481740028736692e-05,
|
| 1332 |
+
"loss": 0.0294,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 3.3868092691622103,
|
| 1337 |
+
"grad_norm": 0.6291656494140625,
|
| 1338 |
+
"learning_rate": 9.474384858657164e-05,
|
| 1339 |
+
"loss": 0.0404,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.4046345811051695,
|
| 1344 |
+
"grad_norm": 0.36534208059310913,
|
| 1345 |
+
"learning_rate": 9.466980757421679e-05,
|
| 1346 |
+
"loss": 0.0345,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 3.4224598930481283,
|
| 1351 |
+
"grad_norm": 0.5696530342102051,
|
| 1352 |
+
"learning_rate": 9.459527806000305e-05,
|
| 1353 |
+
"loss": 0.0381,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 3.4402852049910875,
|
| 1358 |
+
"grad_norm": 0.29749760031700134,
|
| 1359 |
+
"learning_rate": 9.452026085897325e-05,
|
| 1360 |
+
"loss": 0.0356,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 3.4581105169340463,
|
| 1365 |
+
"grad_norm": 0.3628501892089844,
|
| 1366 |
+
"learning_rate": 9.444475679150348e-05,
|
| 1367 |
+
"loss": 0.0337,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.4759358288770055,
|
| 1372 |
+
"grad_norm": 0.3700198531150818,
|
| 1373 |
+
"learning_rate": 9.436876668329411e-05,
|
| 1374 |
+
"loss": 0.0415,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 3.4937611408199643,
|
| 1379 |
+
"grad_norm": 0.2461368292570114,
|
| 1380 |
+
"learning_rate": 9.429229136536079e-05,
|
| 1381 |
+
"loss": 0.0297,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 3.5115864527629235,
|
| 1386 |
+
"grad_norm": 0.42685791850090027,
|
| 1387 |
+
"learning_rate": 9.421533167402534e-05,
|
| 1388 |
+
"loss": 0.0387,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 3.5294117647058822,
|
| 1393 |
+
"grad_norm": 0.2974526882171631,
|
| 1394 |
+
"learning_rate": 9.413788845090666e-05,
|
| 1395 |
+
"loss": 0.0352,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 3.5472370766488415,
|
| 1400 |
+
"grad_norm": 0.3983447551727295,
|
| 1401 |
+
"learning_rate": 9.405996254291136e-05,
|
| 1402 |
+
"loss": 0.0345,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 3.5650623885918002,
|
| 1407 |
+
"grad_norm": 0.41947782039642334,
|
| 1408 |
+
"learning_rate": 9.398155480222474e-05,
|
| 1409 |
+
"loss": 0.0294,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
}
|
| 1412 |
+
],
|
| 1413 |
+
"logging_steps": 10,
|
| 1414 |
+
"max_steps": 10000,
|
| 1415 |
+
"num_input_tokens_seen": 0,
|
| 1416 |
+
"num_train_epochs": 18,
|
| 1417 |
+
"save_steps": 1000,
|
| 1418 |
+
"stateful_callbacks": {
|
| 1419 |
+
"TrainerControl": {
|
| 1420 |
+
"args": {
|
| 1421 |
+
"should_epoch_stop": false,
|
| 1422 |
+
"should_evaluate": false,
|
| 1423 |
+
"should_log": false,
|
| 1424 |
+
"should_save": true,
|
| 1425 |
+
"should_training_stop": false
|
| 1426 |
+
},
|
| 1427 |
+
"attributes": {}
|
| 1428 |
+
}
|
| 1429 |
+
},
|
| 1430 |
+
"total_flos": 0.0,
|
| 1431 |
+
"train_batch_size": 32,
|
| 1432 |
+
"trial_name": null,
|
| 1433 |
+
"trial_params": null
|
| 1434 |
+
}
|