Upload folder using huggingface_hub
Browse files- config.json +64 -0
- experiment_cfg/metadata.json +195 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +0 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +3534 -0
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 |
+
}
|
experiment_cfg/metadata.json
ADDED
@@ -0,0 +1,195 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"new_embodiment": {
|
3 |
+
"statistics": {
|
4 |
+
"state": {
|
5 |
+
"single_arm": {
|
6 |
+
"max": [
|
7 |
+
40.341796875,
|
8 |
+
191.42578125,
|
9 |
+
179.47265625,
|
10 |
+
119.00390625,
|
11 |
+
62.314453125
|
12 |
+
],
|
13 |
+
"min": [
|
14 |
+
-41.30859375,
|
15 |
+
14.853515625,
|
16 |
+
-13.359375,
|
17 |
+
-94.921875,
|
18 |
+
-121.81640625
|
19 |
+
],
|
20 |
+
"mean": [
|
21 |
+
2.636470317840576,
|
22 |
+
107.21146392822266,
|
23 |
+
94.0500259399414,
|
24 |
+
60.918453216552734,
|
25 |
+
-37.02033233642578
|
26 |
+
],
|
27 |
+
"std": [
|
28 |
+
11.717199325561523,
|
29 |
+
52.09779739379883,
|
30 |
+
61.32202911376953,
|
31 |
+
33.041255950927734,
|
32 |
+
34.35018539428711
|
33 |
+
],
|
34 |
+
"q01": [
|
35 |
+
-23.37890625,
|
36 |
+
19.072265625,
|
37 |
+
-12.744140625,
|
38 |
+
-35.38212890625,
|
39 |
+
-100.77275390625
|
40 |
+
],
|
41 |
+
"q99": [
|
42 |
+
28.740234375,
|
43 |
+
191.25,
|
44 |
+
179.296875,
|
45 |
+
118.4765625,
|
46 |
+
35.771484375
|
47 |
+
]
|
48 |
+
},
|
49 |
+
"gripper": {
|
50 |
+
"max": [
|
51 |
+
94.76943969726562
|
52 |
+
],
|
53 |
+
"min": [
|
54 |
+
-6.0564351081848145
|
55 |
+
],
|
56 |
+
"mean": [
|
57 |
+
7.397716999053955
|
58 |
+
],
|
59 |
+
"std": [
|
60 |
+
12.042802810668945
|
61 |
+
],
|
62 |
+
"q01": [
|
63 |
+
-6.0564351081848145
|
64 |
+
],
|
65 |
+
"q99": [
|
66 |
+
45.83620071411133
|
67 |
+
]
|
68 |
+
}
|
69 |
+
},
|
70 |
+
"action": {
|
71 |
+
"single_arm": {
|
72 |
+
"max": [
|
73 |
+
40.166015625,
|
74 |
+
191.25,
|
75 |
+
179.82421875,
|
76 |
+
119.00390625,
|
77 |
+
62.578125
|
78 |
+
],
|
79 |
+
"min": [
|
80 |
+
-41.748046875,
|
81 |
+
14.4140625,
|
82 |
+
-14.501953125,
|
83 |
+
-96.416015625,
|
84 |
+
-122.16796875
|
85 |
+
],
|
86 |
+
"mean": [
|
87 |
+
2.526075601577759,
|
88 |
+
108.57428741455078,
|
89 |
+
92.75291442871094,
|
90 |
+
60.634765625,
|
91 |
+
-36.995540618896484
|
92 |
+
],
|
93 |
+
"std": [
|
94 |
+
11.75656509399414,
|
95 |
+
51.38926315307617,
|
96 |
+
61.780426025390625,
|
97 |
+
33.31574630737305,
|
98 |
+
34.43027877807617
|
99 |
+
],
|
100 |
+
"q01": [
|
101 |
+
-23.818359375,
|
102 |
+
19.86328125,
|
103 |
+
-13.974609375,
|
104 |
+
-35.859375,
|
105 |
+
-100.458984375
|
106 |
+
],
|
107 |
+
"q99": [
|
108 |
+
28.65234375,
|
109 |
+
190.986328125,
|
110 |
+
179.384765625,
|
111 |
+
118.828125,
|
112 |
+
35.733691406249974
|
113 |
+
]
|
114 |
+
},
|
115 |
+
"gripper": {
|
116 |
+
"max": [
|
117 |
+
96.82691955566406
|
118 |
+
],
|
119 |
+
"min": [
|
120 |
+
-8.75
|
121 |
+
],
|
122 |
+
"mean": [
|
123 |
+
3.8709464073181152
|
124 |
+
],
|
125 |
+
"std": [
|
126 |
+
14.702107429504395
|
127 |
+
],
|
128 |
+
"q01": [
|
129 |
+
-8.653846740722656
|
130 |
+
],
|
131 |
+
"q99": [
|
132 |
+
46.92307662963867
|
133 |
+
]
|
134 |
+
}
|
135 |
+
}
|
136 |
+
},
|
137 |
+
"modalities": {
|
138 |
+
"video": {
|
139 |
+
"front": {
|
140 |
+
"resolution": [
|
141 |
+
640,
|
142 |
+
480
|
143 |
+
],
|
144 |
+
"channels": 3,
|
145 |
+
"fps": 30.0
|
146 |
+
},
|
147 |
+
"wrist": {
|
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 |
+
}
|
model-00001-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9cd9c87d246bd7ea64e77398a2426cec253df49483dcd04aebaf37aeec65624f
|
3 |
+
size 4999367032
|
model-00002-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a3ae304bdc2b6b22c88725ef4f3802367dd322fab1e447bb6705f37fb7f8f099
|
3 |
+
size 2586508600
|
model.safetensors.index.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b7c1a393c01195efd063e62be5c011c73e6c1f401fc9224044660620017ad43f
|
3 |
+
size 8550325978
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bb983bc645bb469190ac41d016d8bc7f0fab9eab00367ca1502bf1087b26f439
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:50875594b892047c8e46a61b82a18becc1a9e9c9af6ce75ec5a292af0d5a8cd2
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,3534 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 4.476275738585497,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 5000,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.008952551477170993,
|
14 |
+
"grad_norm": 6.379906177520752,
|
15 |
+
"learning_rate": 9e-06,
|
16 |
+
"loss": 0.9258,
|
17 |
+
"step": 10
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.017905102954341987,
|
21 |
+
"grad_norm": 3.932647943496704,
|
22 |
+
"learning_rate": 1.9e-05,
|
23 |
+
"loss": 0.3837,
|
24 |
+
"step": 20
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.02685765443151298,
|
28 |
+
"grad_norm": 2.0510828495025635,
|
29 |
+
"learning_rate": 2.9e-05,
|
30 |
+
"loss": 0.293,
|
31 |
+
"step": 30
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.03581020590868397,
|
35 |
+
"grad_norm": 2.784146785736084,
|
36 |
+
"learning_rate": 3.9000000000000006e-05,
|
37 |
+
"loss": 0.2134,
|
38 |
+
"step": 40
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.044762757385854966,
|
42 |
+
"grad_norm": 1.2497658729553223,
|
43 |
+
"learning_rate": 4.9e-05,
|
44 |
+
"loss": 0.2007,
|
45 |
+
"step": 50
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.05371530886302596,
|
49 |
+
"grad_norm": 1.8504348993301392,
|
50 |
+
"learning_rate": 5.9e-05,
|
51 |
+
"loss": 0.1805,
|
52 |
+
"step": 60
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.06266786034019696,
|
56 |
+
"grad_norm": 0.86327064037323,
|
57 |
+
"learning_rate": 6.9e-05,
|
58 |
+
"loss": 0.1332,
|
59 |
+
"step": 70
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.07162041181736795,
|
63 |
+
"grad_norm": 0.8830410242080688,
|
64 |
+
"learning_rate": 7.900000000000001e-05,
|
65 |
+
"loss": 0.1299,
|
66 |
+
"step": 80
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.08057296329453895,
|
70 |
+
"grad_norm": 1.671918511390686,
|
71 |
+
"learning_rate": 8.900000000000001e-05,
|
72 |
+
"loss": 0.1184,
|
73 |
+
"step": 90
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.08952551477170993,
|
77 |
+
"grad_norm": 0.9165273308753967,
|
78 |
+
"learning_rate": 9.900000000000001e-05,
|
79 |
+
"loss": 0.116,
|
80 |
+
"step": 100
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.09847806624888093,
|
84 |
+
"grad_norm": 0.9304006099700928,
|
85 |
+
"learning_rate": 9.999446382823013e-05,
|
86 |
+
"loss": 0.1115,
|
87 |
+
"step": 110
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.10743061772605192,
|
91 |
+
"grad_norm": 1.2356555461883545,
|
92 |
+
"learning_rate": 9.997532801828658e-05,
|
93 |
+
"loss": 0.0999,
|
94 |
+
"step": 120
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.11638316920322292,
|
98 |
+
"grad_norm": 1.0443627834320068,
|
99 |
+
"learning_rate": 9.99425294526634e-05,
|
100 |
+
"loss": 0.1,
|
101 |
+
"step": 130
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.12533572068039392,
|
105 |
+
"grad_norm": 0.5913493633270264,
|
106 |
+
"learning_rate": 9.989607709816091e-05,
|
107 |
+
"loss": 0.099,
|
108 |
+
"step": 140
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.13428827215756492,
|
112 |
+
"grad_norm": 0.9996615052223206,
|
113 |
+
"learning_rate": 9.983598365438902e-05,
|
114 |
+
"loss": 0.0886,
|
115 |
+
"step": 150
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.1432408236347359,
|
119 |
+
"grad_norm": 0.839759111404419,
|
120 |
+
"learning_rate": 9.976226555029522e-05,
|
121 |
+
"loss": 0.0909,
|
122 |
+
"step": 160
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.1521933751119069,
|
126 |
+
"grad_norm": 1.140505313873291,
|
127 |
+
"learning_rate": 9.967494293967312e-05,
|
128 |
+
"loss": 0.0901,
|
129 |
+
"step": 170
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.1611459265890779,
|
133 |
+
"grad_norm": 0.7392069101333618,
|
134 |
+
"learning_rate": 9.95740396956525e-05,
|
135 |
+
"loss": 0.0861,
|
136 |
+
"step": 180
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.1700984780662489,
|
140 |
+
"grad_norm": 0.45939213037490845,
|
141 |
+
"learning_rate": 9.945958340417283e-05,
|
142 |
+
"loss": 0.0765,
|
143 |
+
"step": 190
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.17905102954341987,
|
147 |
+
"grad_norm": 0.706051766872406,
|
148 |
+
"learning_rate": 9.93316053564413e-05,
|
149 |
+
"loss": 0.0777,
|
150 |
+
"step": 200
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.18800358102059087,
|
154 |
+
"grad_norm": 1.0259908437728882,
|
155 |
+
"learning_rate": 9.919014054037836e-05,
|
156 |
+
"loss": 0.0806,
|
157 |
+
"step": 210
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.19695613249776187,
|
161 |
+
"grad_norm": 0.5664586424827576,
|
162 |
+
"learning_rate": 9.903522763105218e-05,
|
163 |
+
"loss": 0.0721,
|
164 |
+
"step": 220
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.20590868397493287,
|
168 |
+
"grad_norm": 0.6686021685600281,
|
169 |
+
"learning_rate": 9.886690898010535e-05,
|
170 |
+
"loss": 0.0714,
|
171 |
+
"step": 230
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.21486123545210384,
|
175 |
+
"grad_norm": 0.6658005118370056,
|
176 |
+
"learning_rate": 9.868523060417646e-05,
|
177 |
+
"loss": 0.0748,
|
178 |
+
"step": 240
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.22381378692927484,
|
182 |
+
"grad_norm": 0.6819655299186707,
|
183 |
+
"learning_rate": 9.849024217231935e-05,
|
184 |
+
"loss": 0.0666,
|
185 |
+
"step": 250
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.23276633840644584,
|
189 |
+
"grad_norm": 0.8272667527198792,
|
190 |
+
"learning_rate": 9.82819969924244e-05,
|
191 |
+
"loss": 0.0673,
|
192 |
+
"step": 260
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.24171888988361684,
|
196 |
+
"grad_norm": 1.0225334167480469,
|
197 |
+
"learning_rate": 9.806055199664446e-05,
|
198 |
+
"loss": 0.0721,
|
199 |
+
"step": 270
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.25067144136078784,
|
203 |
+
"grad_norm": 0.8607687950134277,
|
204 |
+
"learning_rate": 9.782596772583026e-05,
|
205 |
+
"loss": 0.0693,
|
206 |
+
"step": 280
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.25962399283795884,
|
210 |
+
"grad_norm": 0.7404839992523193,
|
211 |
+
"learning_rate": 9.757830831297914e-05,
|
212 |
+
"loss": 0.071,
|
213 |
+
"step": 290
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.26857654431512984,
|
217 |
+
"grad_norm": 0.7764973640441895,
|
218 |
+
"learning_rate": 9.731764146570173e-05,
|
219 |
+
"loss": 0.0652,
|
220 |
+
"step": 300
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.2775290957923008,
|
224 |
+
"grad_norm": 0.5465092062950134,
|
225 |
+
"learning_rate": 9.704403844771128e-05,
|
226 |
+
"loss": 0.0759,
|
227 |
+
"step": 310
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.2864816472694718,
|
231 |
+
"grad_norm": 0.6784430146217346,
|
232 |
+
"learning_rate": 9.675757405934103e-05,
|
233 |
+
"loss": 0.0638,
|
234 |
+
"step": 320
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.2954341987466428,
|
238 |
+
"grad_norm": 0.7917226552963257,
|
239 |
+
"learning_rate": 9.645832661709444e-05,
|
240 |
+
"loss": 0.0647,
|
241 |
+
"step": 330
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.3043867502238138,
|
245 |
+
"grad_norm": 1.0117071866989136,
|
246 |
+
"learning_rate": 9.614637793223425e-05,
|
247 |
+
"loss": 0.0676,
|
248 |
+
"step": 340
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.3133393017009848,
|
252 |
+
"grad_norm": 0.8718525767326355,
|
253 |
+
"learning_rate": 9.582181328841611e-05,
|
254 |
+
"loss": 0.0632,
|
255 |
+
"step": 350
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.3222918531781558,
|
259 |
+
"grad_norm": 0.7445501089096069,
|
260 |
+
"learning_rate": 9.548472141837286e-05,
|
261 |
+
"loss": 0.0697,
|
262 |
+
"step": 360
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.3312444046553268,
|
266 |
+
"grad_norm": 0.6769773960113525,
|
267 |
+
"learning_rate": 9.513519447965595e-05,
|
268 |
+
"loss": 0.0698,
|
269 |
+
"step": 370
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.3401969561324978,
|
273 |
+
"grad_norm": 0.5535451769828796,
|
274 |
+
"learning_rate": 9.477332802944044e-05,
|
275 |
+
"loss": 0.0634,
|
276 |
+
"step": 380
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.34914950760966873,
|
280 |
+
"grad_norm": 0.5629348158836365,
|
281 |
+
"learning_rate": 9.439922099840054e-05,
|
282 |
+
"loss": 0.0608,
|
283 |
+
"step": 390
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.35810205908683973,
|
287 |
+
"grad_norm": 0.585233211517334,
|
288 |
+
"learning_rate": 9.401297566366318e-05,
|
289 |
+
"loss": 0.0543,
|
290 |
+
"step": 400
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.36705461056401073,
|
294 |
+
"grad_norm": 0.7889137268066406,
|
295 |
+
"learning_rate": 9.36146976208462e-05,
|
296 |
+
"loss": 0.0629,
|
297 |
+
"step": 410
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.37600716204118173,
|
301 |
+
"grad_norm": 0.3033806085586548,
|
302 |
+
"learning_rate": 9.320449575518972e-05,
|
303 |
+
"loss": 0.0557,
|
304 |
+
"step": 420
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.38495971351835273,
|
308 |
+
"grad_norm": 0.47413498163223267,
|
309 |
+
"learning_rate": 9.278248221178798e-05,
|
310 |
+
"loss": 0.06,
|
311 |
+
"step": 430
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.39391226499552373,
|
315 |
+
"grad_norm": 0.444699764251709,
|
316 |
+
"learning_rate": 9.234877236492997e-05,
|
317 |
+
"loss": 0.0546,
|
318 |
+
"step": 440
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.40286481647269473,
|
322 |
+
"grad_norm": 0.5341030955314636,
|
323 |
+
"learning_rate": 9.190348478655724e-05,
|
324 |
+
"loss": 0.052,
|
325 |
+
"step": 450
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.41181736794986573,
|
329 |
+
"grad_norm": 0.7423449754714966,
|
330 |
+
"learning_rate": 9.144674121384757e-05,
|
331 |
+
"loss": 0.0555,
|
332 |
+
"step": 460
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.42076991942703673,
|
336 |
+
"grad_norm": 0.7645367383956909,
|
337 |
+
"learning_rate": 9.097866651593317e-05,
|
338 |
+
"loss": 0.063,
|
339 |
+
"step": 470
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.4297224709042077,
|
343 |
+
"grad_norm": 0.5560976266860962,
|
344 |
+
"learning_rate": 9.049938865976275e-05,
|
345 |
+
"loss": 0.0621,
|
346 |
+
"step": 480
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.4386750223813787,
|
350 |
+
"grad_norm": 0.3317752182483673,
|
351 |
+
"learning_rate": 9.000903867511666e-05,
|
352 |
+
"loss": 0.0527,
|
353 |
+
"step": 490
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.4476275738585497,
|
357 |
+
"grad_norm": 0.5263016819953918,
|
358 |
+
"learning_rate": 8.950775061878453e-05,
|
359 |
+
"loss": 0.0521,
|
360 |
+
"step": 500
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.4565801253357207,
|
364 |
+
"grad_norm": 0.7472353577613831,
|
365 |
+
"learning_rate": 8.899566153791566e-05,
|
366 |
+
"loss": 0.0582,
|
367 |
+
"step": 510
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.4655326768128917,
|
371 |
+
"grad_norm": 0.5456057786941528,
|
372 |
+
"learning_rate": 8.84729114325516e-05,
|
373 |
+
"loss": 0.0516,
|
374 |
+
"step": 520
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.4744852282900627,
|
378 |
+
"grad_norm": 0.41416245698928833,
|
379 |
+
"learning_rate": 8.79396432173515e-05,
|
380 |
+
"loss": 0.0538,
|
381 |
+
"step": 530
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.4834377797672337,
|
385 |
+
"grad_norm": 0.5883777141571045,
|
386 |
+
"learning_rate": 8.739600268252078e-05,
|
387 |
+
"loss": 0.0485,
|
388 |
+
"step": 540
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.4923903312444047,
|
392 |
+
"grad_norm": 0.5608022809028625,
|
393 |
+
"learning_rate": 8.684213845395339e-05,
|
394 |
+
"loss": 0.0447,
|
395 |
+
"step": 550
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.5013428827215757,
|
399 |
+
"grad_norm": 0.447710245847702,
|
400 |
+
"learning_rate": 8.627820195259918e-05,
|
401 |
+
"loss": 0.0471,
|
402 |
+
"step": 560
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.5102954341987467,
|
406 |
+
"grad_norm": 0.6973375082015991,
|
407 |
+
"learning_rate": 8.570434735306671e-05,
|
408 |
+
"loss": 0.047,
|
409 |
+
"step": 570
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.5192479856759177,
|
413 |
+
"grad_norm": 0.44058454036712646,
|
414 |
+
"learning_rate": 8.512073154147362e-05,
|
415 |
+
"loss": 0.0453,
|
416 |
+
"step": 580
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.5282005371530887,
|
420 |
+
"grad_norm": 0.40700763463974,
|
421 |
+
"learning_rate": 8.452751407255541e-05,
|
422 |
+
"loss": 0.0471,
|
423 |
+
"step": 590
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.5371530886302597,
|
427 |
+
"grad_norm": 0.6121862530708313,
|
428 |
+
"learning_rate": 8.392485712604483e-05,
|
429 |
+
"loss": 0.0516,
|
430 |
+
"step": 600
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.5461056401074306,
|
434 |
+
"grad_norm": 0.5500701665878296,
|
435 |
+
"learning_rate": 8.331292546233362e-05,
|
436 |
+
"loss": 0.0558,
|
437 |
+
"step": 610
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.5550581915846016,
|
441 |
+
"grad_norm": 0.6602392196655273,
|
442 |
+
"learning_rate": 8.269188637742846e-05,
|
443 |
+
"loss": 0.0447,
|
444 |
+
"step": 620
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.5640107430617726,
|
448 |
+
"grad_norm": 0.40000760555267334,
|
449 |
+
"learning_rate": 8.206190965721419e-05,
|
450 |
+
"loss": 0.0438,
|
451 |
+
"step": 630
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.5729632945389436,
|
455 |
+
"grad_norm": 0.4768178164958954,
|
456 |
+
"learning_rate": 8.14231675310358e-05,
|
457 |
+
"loss": 0.0422,
|
458 |
+
"step": 640
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.5819158460161146,
|
462 |
+
"grad_norm": 0.3286081552505493,
|
463 |
+
"learning_rate": 8.077583462461283e-05,
|
464 |
+
"loss": 0.0463,
|
465 |
+
"step": 650
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.5908683974932856,
|
469 |
+
"grad_norm": 0.45187002420425415,
|
470 |
+
"learning_rate": 8.012008791229826e-05,
|
471 |
+
"loss": 0.0506,
|
472 |
+
"step": 660
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.5998209489704566,
|
476 |
+
"grad_norm": 0.7616209983825684,
|
477 |
+
"learning_rate": 7.945610666869568e-05,
|
478 |
+
"loss": 0.0578,
|
479 |
+
"step": 670
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.6087735004476276,
|
483 |
+
"grad_norm": 0.5608704686164856,
|
484 |
+
"learning_rate": 7.878407241964729e-05,
|
485 |
+
"loss": 0.0504,
|
486 |
+
"step": 680
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.6177260519247986,
|
490 |
+
"grad_norm": 0.6086702346801758,
|
491 |
+
"learning_rate": 7.810416889260653e-05,
|
492 |
+
"loss": 0.0487,
|
493 |
+
"step": 690
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.6266786034019696,
|
497 |
+
"grad_norm": 0.48586443066596985,
|
498 |
+
"learning_rate": 7.741658196640892e-05,
|
499 |
+
"loss": 0.0518,
|
500 |
+
"step": 700
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.6356311548791406,
|
504 |
+
"grad_norm": 0.9479012489318848,
|
505 |
+
"learning_rate": 7.672149962045457e-05,
|
506 |
+
"loss": 0.056,
|
507 |
+
"step": 710
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.6445837063563116,
|
511 |
+
"grad_norm": 0.5251392722129822,
|
512 |
+
"learning_rate": 7.60191118833165e-05,
|
513 |
+
"loss": 0.05,
|
514 |
+
"step": 720
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.6535362578334826,
|
518 |
+
"grad_norm": 0.716193437576294,
|
519 |
+
"learning_rate": 7.530961078078873e-05,
|
520 |
+
"loss": 0.0479,
|
521 |
+
"step": 730
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.6624888093106536,
|
525 |
+
"grad_norm": 0.7000860571861267,
|
526 |
+
"learning_rate": 7.45931902833884e-05,
|
527 |
+
"loss": 0.0374,
|
528 |
+
"step": 740
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.6714413607878246,
|
532 |
+
"grad_norm": 0.46128469705581665,
|
533 |
+
"learning_rate": 7.387004625332608e-05,
|
534 |
+
"loss": 0.0441,
|
535 |
+
"step": 750
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.6803939122649956,
|
539 |
+
"grad_norm": 0.5693628191947937,
|
540 |
+
"learning_rate": 7.3140376390959e-05,
|
541 |
+
"loss": 0.0391,
|
542 |
+
"step": 760
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.6893464637421666,
|
546 |
+
"grad_norm": 0.415169894695282,
|
547 |
+
"learning_rate": 7.240438018074189e-05,
|
548 |
+
"loss": 0.0378,
|
549 |
+
"step": 770
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.6982990152193375,
|
553 |
+
"grad_norm": 0.5737828016281128,
|
554 |
+
"learning_rate": 7.166225883668969e-05,
|
555 |
+
"loss": 0.0401,
|
556 |
+
"step": 780
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.7072515666965085,
|
560 |
+
"grad_norm": 0.43452468514442444,
|
561 |
+
"learning_rate": 7.091421524736784e-05,
|
562 |
+
"loss": 0.0364,
|
563 |
+
"step": 790
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.7162041181736795,
|
567 |
+
"grad_norm": 0.4742269515991211,
|
568 |
+
"learning_rate": 7.016045392042452e-05,
|
569 |
+
"loss": 0.0388,
|
570 |
+
"step": 800
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.7251566696508505,
|
574 |
+
"grad_norm": 0.4893990159034729,
|
575 |
+
"learning_rate": 6.940118092668022e-05,
|
576 |
+
"loss": 0.0392,
|
577 |
+
"step": 810
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.7341092211280215,
|
581 |
+
"grad_norm": 0.6735272407531738,
|
582 |
+
"learning_rate": 6.863660384379017e-05,
|
583 |
+
"loss": 0.0457,
|
584 |
+
"step": 820
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.7430617726051925,
|
588 |
+
"grad_norm": 0.36594638228416443,
|
589 |
+
"learning_rate": 6.786693169949455e-05,
|
590 |
+
"loss": 0.0423,
|
591 |
+
"step": 830
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.7520143240823635,
|
595 |
+
"grad_norm": 0.46647775173187256,
|
596 |
+
"learning_rate": 6.709237491447249e-05,
|
597 |
+
"loss": 0.0404,
|
598 |
+
"step": 840
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.7609668755595345,
|
602 |
+
"grad_norm": 0.49492594599723816,
|
603 |
+
"learning_rate": 6.631314524481513e-05,
|
604 |
+
"loss": 0.0442,
|
605 |
+
"step": 850
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.7699194270367055,
|
609 |
+
"grad_norm": 0.4176277816295624,
|
610 |
+
"learning_rate": 6.552945572413358e-05,
|
611 |
+
"loss": 0.0422,
|
612 |
+
"step": 860
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.7788719785138765,
|
616 |
+
"grad_norm": 0.4090403616428375,
|
617 |
+
"learning_rate": 6.474152060531768e-05,
|
618 |
+
"loss": 0.041,
|
619 |
+
"step": 870
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.7878245299910475,
|
623 |
+
"grad_norm": 0.3255680203437805,
|
624 |
+
"learning_rate": 6.394955530196147e-05,
|
625 |
+
"loss": 0.0392,
|
626 |
+
"step": 880
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.7967770814682185,
|
630 |
+
"grad_norm": 0.33741018176078796,
|
631 |
+
"learning_rate": 6.315377632947115e-05,
|
632 |
+
"loss": 0.035,
|
633 |
+
"step": 890
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.8057296329453895,
|
637 |
+
"grad_norm": 0.35257619619369507,
|
638 |
+
"learning_rate": 6.235440124587198e-05,
|
639 |
+
"loss": 0.0429,
|
640 |
+
"step": 900
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.8146821844225605,
|
644 |
+
"grad_norm": 0.35682690143585205,
|
645 |
+
"learning_rate": 6.155164859233012e-05,
|
646 |
+
"loss": 0.0332,
|
647 |
+
"step": 910
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.8236347358997315,
|
651 |
+
"grad_norm": 0.49269065260887146,
|
652 |
+
"learning_rate": 6.074573783340562e-05,
|
653 |
+
"loss": 0.039,
|
654 |
+
"step": 920
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 0.8325872873769025,
|
658 |
+
"grad_norm": 0.35709279775619507,
|
659 |
+
"learning_rate": 5.9936889297052986e-05,
|
660 |
+
"loss": 0.0373,
|
661 |
+
"step": 930
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 0.8415398388540735,
|
665 |
+
"grad_norm": 0.38914117217063904,
|
666 |
+
"learning_rate": 5.912532411438576e-05,
|
667 |
+
"loss": 0.033,
|
668 |
+
"step": 940
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 0.8504923903312444,
|
672 |
+
"grad_norm": 0.5345783829689026,
|
673 |
+
"learning_rate": 5.831126415922148e-05,
|
674 |
+
"loss": 0.0464,
|
675 |
+
"step": 950
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 0.8594449418084154,
|
679 |
+
"grad_norm": 0.5637487769126892,
|
680 |
+
"learning_rate": 5.74949319874235e-05,
|
681 |
+
"loss": 0.0353,
|
682 |
+
"step": 960
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 0.8683974932855864,
|
686 |
+
"grad_norm": 0.3370537757873535,
|
687 |
+
"learning_rate": 5.667655077605659e-05,
|
688 |
+
"loss": 0.035,
|
689 |
+
"step": 970
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 0.8773500447627574,
|
693 |
+
"grad_norm": 0.430148184299469,
|
694 |
+
"learning_rate": 5.585634426237246e-05,
|
695 |
+
"loss": 0.0379,
|
696 |
+
"step": 980
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 0.8863025962399284,
|
700 |
+
"grad_norm": 0.19934052228927612,
|
701 |
+
"learning_rate": 5.5034536682642224e-05,
|
702 |
+
"loss": 0.0344,
|
703 |
+
"step": 990
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 0.8952551477170994,
|
707 |
+
"grad_norm": 0.3409677743911743,
|
708 |
+
"learning_rate": 5.4211352710852495e-05,
|
709 |
+
"loss": 0.034,
|
710 |
+
"step": 1000
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 0.9042076991942704,
|
714 |
+
"grad_norm": 0.6933218836784363,
|
715 |
+
"learning_rate": 9.383125572106752e-05,
|
716 |
+
"loss": 0.0337,
|
717 |
+
"step": 1010
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.9131602506714414,
|
721 |
+
"grad_norm": 0.5733363628387451,
|
722 |
+
"learning_rate": 9.367117688664791e-05,
|
723 |
+
"loss": 0.0377,
|
724 |
+
"step": 1020
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 0.9221128021486124,
|
728 |
+
"grad_norm": 0.6528266668319702,
|
729 |
+
"learning_rate": 9.35091877334763e-05,
|
730 |
+
"loss": 0.0437,
|
731 |
+
"step": 1030
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 0.9310653536257834,
|
735 |
+
"grad_norm": 0.3507431149482727,
|
736 |
+
"learning_rate": 9.334529534748297e-05,
|
737 |
+
"loss": 0.0423,
|
738 |
+
"step": 1040
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.9400179051029544,
|
742 |
+
"grad_norm": 0.5123169422149658,
|
743 |
+
"learning_rate": 9.317950689785188e-05,
|
744 |
+
"loss": 0.0349,
|
745 |
+
"step": 1050
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.9489704565801254,
|
749 |
+
"grad_norm": 0.551750123500824,
|
750 |
+
"learning_rate": 9.301182963670688e-05,
|
751 |
+
"loss": 0.0463,
|
752 |
+
"step": 1060
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.9579230080572964,
|
756 |
+
"grad_norm": 0.3931156396865845,
|
757 |
+
"learning_rate": 9.284227089879456e-05,
|
758 |
+
"loss": 0.0406,
|
759 |
+
"step": 1070
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.9668755595344674,
|
763 |
+
"grad_norm": 0.5835707783699036,
|
764 |
+
"learning_rate": 9.26708381011634e-05,
|
765 |
+
"loss": 0.0426,
|
766 |
+
"step": 1080
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.9758281110116384,
|
770 |
+
"grad_norm": 0.3757207989692688,
|
771 |
+
"learning_rate": 9.249753874283937e-05,
|
772 |
+
"loss": 0.0315,
|
773 |
+
"step": 1090
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.9847806624888094,
|
777 |
+
"grad_norm": 0.6495594382286072,
|
778 |
+
"learning_rate": 9.232238040449779e-05,
|
779 |
+
"loss": 0.0419,
|
780 |
+
"step": 1100
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.9937332139659804,
|
784 |
+
"grad_norm": 0.6033176779747009,
|
785 |
+
"learning_rate": 9.214537074813181e-05,
|
786 |
+
"loss": 0.0414,
|
787 |
+
"step": 1110
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 1.0026857654431514,
|
791 |
+
"grad_norm": 0.599250316619873,
|
792 |
+
"learning_rate": 9.196651751671724e-05,
|
793 |
+
"loss": 0.0421,
|
794 |
+
"step": 1120
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 1.0116383169203222,
|
798 |
+
"grad_norm": 0.4811965227127075,
|
799 |
+
"learning_rate": 9.178582853387384e-05,
|
800 |
+
"loss": 0.0477,
|
801 |
+
"step": 1130
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 1.0205908683974934,
|
805 |
+
"grad_norm": 0.6312634944915771,
|
806 |
+
"learning_rate": 9.160331170352304e-05,
|
807 |
+
"loss": 0.0385,
|
808 |
+
"step": 1140
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 1.0295434198746642,
|
812 |
+
"grad_norm": 0.4195096492767334,
|
813 |
+
"learning_rate": 9.141897500954229e-05,
|
814 |
+
"loss": 0.0425,
|
815 |
+
"step": 1150
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 1.0384959713518354,
|
819 |
+
"grad_norm": 0.4496900141239166,
|
820 |
+
"learning_rate": 9.123282651541576e-05,
|
821 |
+
"loss": 0.039,
|
822 |
+
"step": 1160
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 1.0474485228290062,
|
826 |
+
"grad_norm": 0.5114062428474426,
|
827 |
+
"learning_rate": 9.104487436388161e-05,
|
828 |
+
"loss": 0.0362,
|
829 |
+
"step": 1170
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 1.0564010743061774,
|
833 |
+
"grad_norm": 0.46173688769340515,
|
834 |
+
"learning_rate": 9.085512677657582e-05,
|
835 |
+
"loss": 0.04,
|
836 |
+
"step": 1180
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 1.0653536257833482,
|
840 |
+
"grad_norm": 0.5690197348594666,
|
841 |
+
"learning_rate": 9.066359205367258e-05,
|
842 |
+
"loss": 0.0424,
|
843 |
+
"step": 1190
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 1.0743061772605191,
|
847 |
+
"grad_norm": 0.33746033906936646,
|
848 |
+
"learning_rate": 9.047027857352112e-05,
|
849 |
+
"loss": 0.0468,
|
850 |
+
"step": 1200
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 1.0832587287376902,
|
854 |
+
"grad_norm": 0.6568852663040161,
|
855 |
+
"learning_rate": 9.027519479227935e-05,
|
856 |
+
"loss": 0.0406,
|
857 |
+
"step": 1210
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 1.0922112802148611,
|
861 |
+
"grad_norm": 0.4321780502796173,
|
862 |
+
"learning_rate": 9.007834924354383e-05,
|
863 |
+
"loss": 0.0366,
|
864 |
+
"step": 1220
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 1.1011638316920322,
|
868 |
+
"grad_norm": 0.5011163949966431,
|
869 |
+
"learning_rate": 8.987975053797655e-05,
|
870 |
+
"loss": 0.04,
|
871 |
+
"step": 1230
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 1.1101163831692031,
|
875 |
+
"grad_norm": 0.4079154133796692,
|
876 |
+
"learning_rate": 8.967940736292825e-05,
|
877 |
+
"loss": 0.0389,
|
878 |
+
"step": 1240
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 1.1190689346463742,
|
882 |
+
"grad_norm": 0.7240653038024902,
|
883 |
+
"learning_rate": 8.947732848205846e-05,
|
884 |
+
"loss": 0.0448,
|
885 |
+
"step": 1250
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 1.1280214861235451,
|
889 |
+
"grad_norm": 0.6067286729812622,
|
890 |
+
"learning_rate": 8.927352273495204e-05,
|
891 |
+
"loss": 0.0366,
|
892 |
+
"step": 1260
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 1.1369740376007162,
|
896 |
+
"grad_norm": 0.5315434336662292,
|
897 |
+
"learning_rate": 8.906799903673265e-05,
|
898 |
+
"loss": 0.0406,
|
899 |
+
"step": 1270
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 1.1459265890778871,
|
903 |
+
"grad_norm": 0.3086467385292053,
|
904 |
+
"learning_rate": 8.88607663776726e-05,
|
905 |
+
"loss": 0.0386,
|
906 |
+
"step": 1280
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 1.1548791405550582,
|
910 |
+
"grad_norm": 0.3972776234149933,
|
911 |
+
"learning_rate": 8.865183382279978e-05,
|
912 |
+
"loss": 0.0414,
|
913 |
+
"step": 1290
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 1.1638316920322291,
|
917 |
+
"grad_norm": 0.28744080662727356,
|
918 |
+
"learning_rate": 8.844121051150096e-05,
|
919 |
+
"loss": 0.0355,
|
920 |
+
"step": 1300
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 1.1727842435094002,
|
924 |
+
"grad_norm": 0.4855988919734955,
|
925 |
+
"learning_rate": 8.822890565712211e-05,
|
926 |
+
"loss": 0.036,
|
927 |
+
"step": 1310
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 1.1817367949865711,
|
931 |
+
"grad_norm": 0.5339410305023193,
|
932 |
+
"learning_rate": 8.801492854656536e-05,
|
933 |
+
"loss": 0.0401,
|
934 |
+
"step": 1320
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 1.1906893464637422,
|
938 |
+
"grad_norm": 0.3985954225063324,
|
939 |
+
"learning_rate": 8.779928853988268e-05,
|
940 |
+
"loss": 0.0335,
|
941 |
+
"step": 1330
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 1.1996418979409131,
|
945 |
+
"grad_norm": 0.40003374218940735,
|
946 |
+
"learning_rate": 8.758199506986655e-05,
|
947 |
+
"loss": 0.0376,
|
948 |
+
"step": 1340
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 1.2085944494180842,
|
952 |
+
"grad_norm": 0.6960264444351196,
|
953 |
+
"learning_rate": 8.73630576416373e-05,
|
954 |
+
"loss": 0.0342,
|
955 |
+
"step": 1350
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 1.2175470008952551,
|
959 |
+
"grad_norm": 0.49340131878852844,
|
960 |
+
"learning_rate": 8.714248583222726e-05,
|
961 |
+
"loss": 0.0388,
|
962 |
+
"step": 1360
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 1.226499552372426,
|
966 |
+
"grad_norm": 0.43804675340652466,
|
967 |
+
"learning_rate": 8.692028929016196e-05,
|
968 |
+
"loss": 0.0292,
|
969 |
+
"step": 1370
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 1.2354521038495971,
|
973 |
+
"grad_norm": 0.37847551703453064,
|
974 |
+
"learning_rate": 8.669647773503797e-05,
|
975 |
+
"loss": 0.036,
|
976 |
+
"step": 1380
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 1.2444046553267682,
|
980 |
+
"grad_norm": 0.35698914527893066,
|
981 |
+
"learning_rate": 8.647106095709773e-05,
|
982 |
+
"loss": 0.033,
|
983 |
+
"step": 1390
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 1.2533572068039391,
|
987 |
+
"grad_norm": 0.4798845648765564,
|
988 |
+
"learning_rate": 8.624404881680139e-05,
|
989 |
+
"loss": 0.0357,
|
990 |
+
"step": 1400
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 1.26230975828111,
|
994 |
+
"grad_norm": 0.5304312109947205,
|
995 |
+
"learning_rate": 8.601545124439535e-05,
|
996 |
+
"loss": 0.0326,
|
997 |
+
"step": 1410
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 1.2712623097582811,
|
1001 |
+
"grad_norm": 0.4820861220359802,
|
1002 |
+
"learning_rate": 8.5785278239478e-05,
|
1003 |
+
"loss": 0.0392,
|
1004 |
+
"step": 1420
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 1.280214861235452,
|
1008 |
+
"grad_norm": 0.34791550040245056,
|
1009 |
+
"learning_rate": 8.555353987056224e-05,
|
1010 |
+
"loss": 0.0333,
|
1011 |
+
"step": 1430
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 1.2891674127126231,
|
1015 |
+
"grad_norm": 0.47185078263282776,
|
1016 |
+
"learning_rate": 8.532024627463505e-05,
|
1017 |
+
"loss": 0.0357,
|
1018 |
+
"step": 1440
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 1.298119964189794,
|
1022 |
+
"grad_norm": 0.44588857889175415,
|
1023 |
+
"learning_rate": 8.508540765671407e-05,
|
1024 |
+
"loss": 0.0352,
|
1025 |
+
"step": 1450
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 1.3070725156669651,
|
1029 |
+
"grad_norm": 0.4541409909725189,
|
1030 |
+
"learning_rate": 8.484903428940121e-05,
|
1031 |
+
"loss": 0.0376,
|
1032 |
+
"step": 1460
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 1.316025067144136,
|
1036 |
+
"grad_norm": 0.45099449157714844,
|
1037 |
+
"learning_rate": 8.461113651243334e-05,
|
1038 |
+
"loss": 0.0383,
|
1039 |
+
"step": 1470
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 1.3249776186213071,
|
1043 |
+
"grad_norm": 0.5666273832321167,
|
1044 |
+
"learning_rate": 8.437172473222987e-05,
|
1045 |
+
"loss": 0.0372,
|
1046 |
+
"step": 1480
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 1.333930170098478,
|
1050 |
+
"grad_norm": 0.35443374514579773,
|
1051 |
+
"learning_rate": 8.413080942143767e-05,
|
1052 |
+
"loss": 0.0321,
|
1053 |
+
"step": 1490
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 1.3428827215756491,
|
1057 |
+
"grad_norm": 0.4038570821285248,
|
1058 |
+
"learning_rate": 8.388840111847288e-05,
|
1059 |
+
"loss": 0.0391,
|
1060 |
+
"step": 1500
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 1.35183527305282,
|
1064 |
+
"grad_norm": 0.44889935851097107,
|
1065 |
+
"learning_rate": 8.364451042705998e-05,
|
1066 |
+
"loss": 0.0357,
|
1067 |
+
"step": 1510
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 1.3607878245299911,
|
1071 |
+
"grad_norm": 0.24770434200763702,
|
1072 |
+
"learning_rate": 8.33991480157679e-05,
|
1073 |
+
"loss": 0.0348,
|
1074 |
+
"step": 1520
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 1.369740376007162,
|
1078 |
+
"grad_norm": 0.2446165382862091,
|
1079 |
+
"learning_rate": 8.315232461754338e-05,
|
1080 |
+
"loss": 0.0351,
|
1081 |
+
"step": 1530
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 1.378692927484333,
|
1085 |
+
"grad_norm": 0.36377421021461487,
|
1086 |
+
"learning_rate": 8.290405102924144e-05,
|
1087 |
+
"loss": 0.0389,
|
1088 |
+
"step": 1540
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 1.387645478961504,
|
1092 |
+
"grad_norm": 0.38384515047073364,
|
1093 |
+
"learning_rate": 8.265433811115316e-05,
|
1094 |
+
"loss": 0.038,
|
1095 |
+
"step": 1550
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 1.3965980304386751,
|
1099 |
+
"grad_norm": 0.5154732465744019,
|
1100 |
+
"learning_rate": 8.240319678653049e-05,
|
1101 |
+
"loss": 0.037,
|
1102 |
+
"step": 1560
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 1.405550581915846,
|
1106 |
+
"grad_norm": 0.39229637384414673,
|
1107 |
+
"learning_rate": 8.215063804110857e-05,
|
1108 |
+
"loss": 0.0358,
|
1109 |
+
"step": 1570
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 1.414503133393017,
|
1113 |
+
"grad_norm": 0.30531948804855347,
|
1114 |
+
"learning_rate": 8.189667292262512e-05,
|
1115 |
+
"loss": 0.0321,
|
1116 |
+
"step": 1580
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 1.423455684870188,
|
1120 |
+
"grad_norm": 0.3962353765964508,
|
1121 |
+
"learning_rate": 8.164131254033716e-05,
|
1122 |
+
"loss": 0.036,
|
1123 |
+
"step": 1590
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 1.432408236347359,
|
1127 |
+
"grad_norm": 0.24542206525802612,
|
1128 |
+
"learning_rate": 8.138456806453503e-05,
|
1129 |
+
"loss": 0.0346,
|
1130 |
+
"step": 1600
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 1.44136078782453,
|
1134 |
+
"grad_norm": 0.39958980679512024,
|
1135 |
+
"learning_rate": 8.112645072605386e-05,
|
1136 |
+
"loss": 0.0282,
|
1137 |
+
"step": 1610
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 1.450313339301701,
|
1141 |
+
"grad_norm": 0.40440255403518677,
|
1142 |
+
"learning_rate": 8.086697181578222e-05,
|
1143 |
+
"loss": 0.0316,
|
1144 |
+
"step": 1620
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 1.459265890778872,
|
1148 |
+
"grad_norm": 0.4727575182914734,
|
1149 |
+
"learning_rate": 8.060614268416823e-05,
|
1150 |
+
"loss": 0.0331,
|
1151 |
+
"step": 1630
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 1.468218442256043,
|
1155 |
+
"grad_norm": 0.5293134450912476,
|
1156 |
+
"learning_rate": 8.034397474072309e-05,
|
1157 |
+
"loss": 0.0385,
|
1158 |
+
"step": 1640
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 1.477170993733214,
|
1162 |
+
"grad_norm": 0.757794976234436,
|
1163 |
+
"learning_rate": 8.008047945352193e-05,
|
1164 |
+
"loss": 0.0384,
|
1165 |
+
"step": 1650
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 1.486123545210385,
|
1169 |
+
"grad_norm": 0.5270459055900574,
|
1170 |
+
"learning_rate": 7.981566834870225e-05,
|
1171 |
+
"loss": 0.0278,
|
1172 |
+
"step": 1660
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 1.495076096687556,
|
1176 |
+
"grad_norm": 0.44656965136528015,
|
1177 |
+
"learning_rate": 7.954955300995961e-05,
|
1178 |
+
"loss": 0.038,
|
1179 |
+
"step": 1670
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 1.504028648164727,
|
1183 |
+
"grad_norm": 0.3250710070133209,
|
1184 |
+
"learning_rate": 7.928214507804104e-05,
|
1185 |
+
"loss": 0.0318,
|
1186 |
+
"step": 1680
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 1.512981199641898,
|
1190 |
+
"grad_norm": 0.26046040654182434,
|
1191 |
+
"learning_rate": 7.901345625023576e-05,
|
1192 |
+
"loss": 0.0322,
|
1193 |
+
"step": 1690
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 1.521933751119069,
|
1197 |
+
"grad_norm": 0.43262678384780884,
|
1198 |
+
"learning_rate": 7.874349827986354e-05,
|
1199 |
+
"loss": 0.0293,
|
1200 |
+
"step": 1700
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 1.5308863025962398,
|
1204 |
+
"grad_norm": 0.44547683000564575,
|
1205 |
+
"learning_rate": 7.847228297576053e-05,
|
1206 |
+
"loss": 0.0369,
|
1207 |
+
"step": 1710
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 1.539838854073411,
|
1211 |
+
"grad_norm": 0.39528220891952515,
|
1212 |
+
"learning_rate": 7.819982220176276e-05,
|
1213 |
+
"loss": 0.0331,
|
1214 |
+
"step": 1720
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 1.548791405550582,
|
1218 |
+
"grad_norm": 0.522739827632904,
|
1219 |
+
"learning_rate": 7.792612787618714e-05,
|
1220 |
+
"loss": 0.0327,
|
1221 |
+
"step": 1730
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 1.557743957027753,
|
1225 |
+
"grad_norm": 0.3753514289855957,
|
1226 |
+
"learning_rate": 7.765121197131009e-05,
|
1227 |
+
"loss": 0.0303,
|
1228 |
+
"step": 1740
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 1.5666965085049238,
|
1232 |
+
"grad_norm": 0.41809502243995667,
|
1233 |
+
"learning_rate": 7.737508651284391e-05,
|
1234 |
+
"loss": 0.0334,
|
1235 |
+
"step": 1750
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 1.575649059982095,
|
1239 |
+
"grad_norm": 0.43867743015289307,
|
1240 |
+
"learning_rate": 7.709776357941069e-05,
|
1241 |
+
"loss": 0.0351,
|
1242 |
+
"step": 1760
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 1.584601611459266,
|
1246 |
+
"grad_norm": 0.4181774854660034,
|
1247 |
+
"learning_rate": 7.681925530201392e-05,
|
1248 |
+
"loss": 0.0337,
|
1249 |
+
"step": 1770
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 1.593554162936437,
|
1253 |
+
"grad_norm": 0.40942829847335815,
|
1254 |
+
"learning_rate": 7.65395738635079e-05,
|
1255 |
+
"loss": 0.0322,
|
1256 |
+
"step": 1780
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 1.6025067144136078,
|
1260 |
+
"grad_norm": 0.4726716876029968,
|
1261 |
+
"learning_rate": 7.62587314980648e-05,
|
1262 |
+
"loss": 0.0392,
|
1263 |
+
"step": 1790
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 1.611459265890779,
|
1267 |
+
"grad_norm": 0.43389004468917847,
|
1268 |
+
"learning_rate": 7.597674049063947e-05,
|
1269 |
+
"loss": 0.0342,
|
1270 |
+
"step": 1800
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 1.6204118173679498,
|
1274 |
+
"grad_norm": 0.60209721326828,
|
1275 |
+
"learning_rate": 7.569361317643211e-05,
|
1276 |
+
"loss": 0.0307,
|
1277 |
+
"step": 1810
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 1.629364368845121,
|
1281 |
+
"grad_norm": 0.27818477153778076,
|
1282 |
+
"learning_rate": 7.540936194034865e-05,
|
1283 |
+
"loss": 0.0298,
|
1284 |
+
"step": 1820
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.6383169203222918,
|
1288 |
+
"grad_norm": 0.3243078291416168,
|
1289 |
+
"learning_rate": 7.512399921645901e-05,
|
1290 |
+
"loss": 0.0305,
|
1291 |
+
"step": 1830
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.6472694717994627,
|
1295 |
+
"grad_norm": 0.5586011409759521,
|
1296 |
+
"learning_rate": 7.483753748745317e-05,
|
1297 |
+
"loss": 0.0316,
|
1298 |
+
"step": 1840
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 1.6562220232766338,
|
1302 |
+
"grad_norm": 0.3181556165218353,
|
1303 |
+
"learning_rate": 7.454998928409516e-05,
|
1304 |
+
"loss": 0.0342,
|
1305 |
+
"step": 1850
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 1.665174574753805,
|
1309 |
+
"grad_norm": 0.3369368016719818,
|
1310 |
+
"learning_rate": 7.426136718467493e-05,
|
1311 |
+
"loss": 0.0348,
|
1312 |
+
"step": 1860
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 1.6741271262309758,
|
1316 |
+
"grad_norm": 0.42150554060935974,
|
1317 |
+
"learning_rate": 7.397168381445812e-05,
|
1318 |
+
"loss": 0.032,
|
1319 |
+
"step": 1870
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 1.6830796777081467,
|
1323 |
+
"grad_norm": 0.4693697392940521,
|
1324 |
+
"learning_rate": 7.368095184513377e-05,
|
1325 |
+
"loss": 0.0282,
|
1326 |
+
"step": 1880
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 1.6920322291853178,
|
1330 |
+
"grad_norm": 0.33959540724754333,
|
1331 |
+
"learning_rate": 7.338918399426005e-05,
|
1332 |
+
"loss": 0.0253,
|
1333 |
+
"step": 1890
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 1.700984780662489,
|
1337 |
+
"grad_norm": 0.3847089409828186,
|
1338 |
+
"learning_rate": 7.309639302470801e-05,
|
1339 |
+
"loss": 0.0342,
|
1340 |
+
"step": 1900
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 1.7099373321396598,
|
1344 |
+
"grad_norm": 0.49096646904945374,
|
1345 |
+
"learning_rate": 7.280259174410312e-05,
|
1346 |
+
"loss": 0.0326,
|
1347 |
+
"step": 1910
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 1.7188898836168307,
|
1351 |
+
"grad_norm": 0.36844900250434875,
|
1352 |
+
"learning_rate": 7.250779300426517e-05,
|
1353 |
+
"loss": 0.0289,
|
1354 |
+
"step": 1920
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 1.7278424350940018,
|
1358 |
+
"grad_norm": 0.3279619514942169,
|
1359 |
+
"learning_rate": 7.22120097006461e-05,
|
1360 |
+
"loss": 0.0335,
|
1361 |
+
"step": 1930
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 1.736794986571173,
|
1365 |
+
"grad_norm": 0.4344680905342102,
|
1366 |
+
"learning_rate": 7.191525477176577e-05,
|
1367 |
+
"loss": 0.0305,
|
1368 |
+
"step": 1940
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 1.7457475380483438,
|
1372 |
+
"grad_norm": 0.2836930453777313,
|
1373 |
+
"learning_rate": 7.161754119864616e-05,
|
1374 |
+
"loss": 0.0278,
|
1375 |
+
"step": 1950
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 1.7547000895255147,
|
1379 |
+
"grad_norm": 0.49985265731811523,
|
1380 |
+
"learning_rate": 7.131888200424339e-05,
|
1381 |
+
"loss": 0.0261,
|
1382 |
+
"step": 1960
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 1.7636526410026858,
|
1386 |
+
"grad_norm": 0.40372687578201294,
|
1387 |
+
"learning_rate": 7.101929025287816e-05,
|
1388 |
+
"loss": 0.0309,
|
1389 |
+
"step": 1970
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 1.7726051924798567,
|
1393 |
+
"grad_norm": 0.34391847252845764,
|
1394 |
+
"learning_rate": 7.071877904966423e-05,
|
1395 |
+
"loss": 0.0352,
|
1396 |
+
"step": 1980
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 1.7815577439570278,
|
1400 |
+
"grad_norm": 0.3783436417579651,
|
1401 |
+
"learning_rate": 7.04173615399351e-05,
|
1402 |
+
"loss": 0.0275,
|
1403 |
+
"step": 1990
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 1.7905102954341987,
|
1407 |
+
"grad_norm": 0.35148048400878906,
|
1408 |
+
"learning_rate": 7.011505090866913e-05,
|
1409 |
+
"loss": 0.0278,
|
1410 |
+
"step": 2000
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 1.7994628469113696,
|
1414 |
+
"grad_norm": 0.4762101471424103,
|
1415 |
+
"learning_rate": 6.981186037991271e-05,
|
1416 |
+
"loss": 0.0271,
|
1417 |
+
"step": 2010
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 1.8084153983885407,
|
1421 |
+
"grad_norm": 0.28804442286491394,
|
1422 |
+
"learning_rate": 6.950780321620174e-05,
|
1423 |
+
"loss": 0.0299,
|
1424 |
+
"step": 2020
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 1.8173679498657118,
|
1428 |
+
"grad_norm": 0.4394608438014984,
|
1429 |
+
"learning_rate": 6.920289271798157e-05,
|
1430 |
+
"loss": 0.0316,
|
1431 |
+
"step": 2030
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 1.8263205013428827,
|
1435 |
+
"grad_norm": 0.3154217302799225,
|
1436 |
+
"learning_rate": 6.889714222302517e-05,
|
1437 |
+
"loss": 0.0352,
|
1438 |
+
"step": 2040
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 1.8352730528200536,
|
1442 |
+
"grad_norm": 0.390591561794281,
|
1443 |
+
"learning_rate": 6.85905651058497e-05,
|
1444 |
+
"loss": 0.0327,
|
1445 |
+
"step": 2050
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 1.8442256042972247,
|
1449 |
+
"grad_norm": 0.42770206928253174,
|
1450 |
+
"learning_rate": 6.82831747771314e-05,
|
1451 |
+
"loss": 0.0259,
|
1452 |
+
"step": 2060
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 1.8531781557743958,
|
1456 |
+
"grad_norm": 0.28651660680770874,
|
1457 |
+
"learning_rate": 6.797498468311907e-05,
|
1458 |
+
"loss": 0.0258,
|
1459 |
+
"step": 2070
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 1.8621307072515667,
|
1463 |
+
"grad_norm": 0.41473740339279175,
|
1464 |
+
"learning_rate": 6.766600830504585e-05,
|
1465 |
+
"loss": 0.0241,
|
1466 |
+
"step": 2080
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 1.8710832587287376,
|
1470 |
+
"grad_norm": 0.31760624051094055,
|
1471 |
+
"learning_rate": 6.735625915853942e-05,
|
1472 |
+
"loss": 0.0233,
|
1473 |
+
"step": 2090
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 1.8800358102059087,
|
1477 |
+
"grad_norm": 0.41052040457725525,
|
1478 |
+
"learning_rate": 6.70457507930309e-05,
|
1479 |
+
"loss": 0.0253,
|
1480 |
+
"step": 2100
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 1.8889883616830798,
|
1484 |
+
"grad_norm": 0.37952539324760437,
|
1485 |
+
"learning_rate": 6.673449679116215e-05,
|
1486 |
+
"loss": 0.025,
|
1487 |
+
"step": 2110
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 1.8979409131602507,
|
1491 |
+
"grad_norm": 0.33014625310897827,
|
1492 |
+
"learning_rate": 6.642251076819148e-05,
|
1493 |
+
"loss": 0.023,
|
1494 |
+
"step": 2120
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 1.9068934646374216,
|
1498 |
+
"grad_norm": 0.3890449106693268,
|
1499 |
+
"learning_rate": 6.610980637139827e-05,
|
1500 |
+
"loss": 0.0278,
|
1501 |
+
"step": 2130
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 1.9158460161145927,
|
1505 |
+
"grad_norm": 0.34647008776664734,
|
1506 |
+
"learning_rate": 6.579639727948583e-05,
|
1507 |
+
"loss": 0.0289,
|
1508 |
+
"step": 2140
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 1.9247985675917636,
|
1512 |
+
"grad_norm": 0.21908289194107056,
|
1513 |
+
"learning_rate": 6.548229720198315e-05,
|
1514 |
+
"loss": 0.0349,
|
1515 |
+
"step": 2150
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 1.9337511190689347,
|
1519 |
+
"grad_norm": 0.3591015636920929,
|
1520 |
+
"learning_rate": 6.516751987864517e-05,
|
1521 |
+
"loss": 0.0271,
|
1522 |
+
"step": 2160
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 1.9427036705461056,
|
1526 |
+
"grad_norm": 0.4022800922393799,
|
1527 |
+
"learning_rate": 6.485207907885175e-05,
|
1528 |
+
"loss": 0.0233,
|
1529 |
+
"step": 2170
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 1.9516562220232765,
|
1533 |
+
"grad_norm": 0.6193303465843201,
|
1534 |
+
"learning_rate": 6.453598860100536e-05,
|
1535 |
+
"loss": 0.0304,
|
1536 |
+
"step": 2180
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 1.9606087735004476,
|
1540 |
+
"grad_norm": 0.2870684862136841,
|
1541 |
+
"learning_rate": 6.421926227192749e-05,
|
1542 |
+
"loss": 0.0239,
|
1543 |
+
"step": 2190
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 1.9695613249776187,
|
1547 |
+
"grad_norm": 0.4261172413825989,
|
1548 |
+
"learning_rate": 6.390191394625381e-05,
|
1549 |
+
"loss": 0.0294,
|
1550 |
+
"step": 2200
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 1.9785138764547896,
|
1554 |
+
"grad_norm": 0.17389577627182007,
|
1555 |
+
"learning_rate": 6.358395750582817e-05,
|
1556 |
+
"loss": 0.0242,
|
1557 |
+
"step": 2210
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 1.9874664279319605,
|
1561 |
+
"grad_norm": 0.2946527302265167,
|
1562 |
+
"learning_rate": 6.326540685909532e-05,
|
1563 |
+
"loss": 0.0219,
|
1564 |
+
"step": 2220
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 1.9964189794091316,
|
1568 |
+
"grad_norm": 0.35331255197525024,
|
1569 |
+
"learning_rate": 6.294627594049249e-05,
|
1570 |
+
"loss": 0.0291,
|
1571 |
+
"step": 2230
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 2.0053715308863027,
|
1575 |
+
"grad_norm": 0.3004955053329468,
|
1576 |
+
"learning_rate": 6.262657870983989e-05,
|
1577 |
+
"loss": 0.0244,
|
1578 |
+
"step": 2240
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 2.0143240823634736,
|
1582 |
+
"grad_norm": 0.28008395433425903,
|
1583 |
+
"learning_rate": 6.230632915173009e-05,
|
1584 |
+
"loss": 0.0272,
|
1585 |
+
"step": 2250
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 2.0232766338406445,
|
1589 |
+
"grad_norm": 0.2564995586872101,
|
1590 |
+
"learning_rate": 6.198554127491622e-05,
|
1591 |
+
"loss": 0.0264,
|
1592 |
+
"step": 2260
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 2.0322291853178154,
|
1596 |
+
"grad_norm": 0.3245619833469391,
|
1597 |
+
"learning_rate": 6.166422911169923e-05,
|
1598 |
+
"loss": 0.0223,
|
1599 |
+
"step": 2270
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 2.0411817367949867,
|
1603 |
+
"grad_norm": 0.19854438304901123,
|
1604 |
+
"learning_rate": 6.1342406717314e-05,
|
1605 |
+
"loss": 0.0277,
|
1606 |
+
"step": 2280
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 2.0501342882721576,
|
1610 |
+
"grad_norm": 0.1700948178768158,
|
1611 |
+
"learning_rate": 6.102008816931466e-05,
|
1612 |
+
"loss": 0.0248,
|
1613 |
+
"step": 2290
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 2.0590868397493285,
|
1617 |
+
"grad_norm": 0.2796550989151001,
|
1618 |
+
"learning_rate": 6.069728756695866e-05,
|
1619 |
+
"loss": 0.0247,
|
1620 |
+
"step": 2300
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 2.0680393912264994,
|
1624 |
+
"grad_norm": 0.21811795234680176,
|
1625 |
+
"learning_rate": 6.037401903059008e-05,
|
1626 |
+
"loss": 0.0284,
|
1627 |
+
"step": 2310
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 2.0769919427036707,
|
1631 |
+
"grad_norm": 0.24635659158229828,
|
1632 |
+
"learning_rate": 6.005029670102195e-05,
|
1633 |
+
"loss": 0.0218,
|
1634 |
+
"step": 2320
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 2.0859444941808416,
|
1638 |
+
"grad_norm": 0.3100559115409851,
|
1639 |
+
"learning_rate": 5.972613473891766e-05,
|
1640 |
+
"loss": 0.0312,
|
1641 |
+
"step": 2330
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 2.0948970456580125,
|
1645 |
+
"grad_norm": 0.3434387743473053,
|
1646 |
+
"learning_rate": 5.940154732417158e-05,
|
1647 |
+
"loss": 0.0234,
|
1648 |
+
"step": 2340
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 2.1038495971351834,
|
1652 |
+
"grad_norm": 0.4951243996620178,
|
1653 |
+
"learning_rate": 5.907654865528876e-05,
|
1654 |
+
"loss": 0.0305,
|
1655 |
+
"step": 2350
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 2.1128021486123547,
|
1659 |
+
"grad_norm": 0.3198152184486389,
|
1660 |
+
"learning_rate": 5.875115294876381e-05,
|
1661 |
+
"loss": 0.0253,
|
1662 |
+
"step": 2360
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 2.1217547000895256,
|
1666 |
+
"grad_norm": 0.6087833046913147,
|
1667 |
+
"learning_rate": 5.842537443845908e-05,
|
1668 |
+
"loss": 0.0241,
|
1669 |
+
"step": 2370
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 2.1307072515666965,
|
1673 |
+
"grad_norm": 0.30654656887054443,
|
1674 |
+
"learning_rate": 5.809922737498198e-05,
|
1675 |
+
"loss": 0.0211,
|
1676 |
+
"step": 2380
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 2.1396598030438674,
|
1680 |
+
"grad_norm": 0.4575793445110321,
|
1681 |
+
"learning_rate": 5.777272602506165e-05,
|
1682 |
+
"loss": 0.0286,
|
1683 |
+
"step": 2390
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 2.1486123545210383,
|
1687 |
+
"grad_norm": 0.6392210721969604,
|
1688 |
+
"learning_rate": 5.744588467092483e-05,
|
1689 |
+
"loss": 0.0269,
|
1690 |
+
"step": 2400
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 2.1575649059982096,
|
1694 |
+
"grad_norm": 0.2926782965660095,
|
1695 |
+
"learning_rate": 5.7118717609671194e-05,
|
1696 |
+
"loss": 0.0225,
|
1697 |
+
"step": 2410
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 2.1665174574753805,
|
1701 |
+
"grad_norm": 0.3610515296459198,
|
1702 |
+
"learning_rate": 5.679123915264786e-05,
|
1703 |
+
"loss": 0.0249,
|
1704 |
+
"step": 2420
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 2.1754700089525514,
|
1708 |
+
"grad_norm": 0.31393954157829285,
|
1709 |
+
"learning_rate": 5.646346362482342e-05,
|
1710 |
+
"loss": 0.0225,
|
1711 |
+
"step": 2430
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 2.1844225604297223,
|
1715 |
+
"grad_norm": 0.41921770572662354,
|
1716 |
+
"learning_rate": 5.613540536416132e-05,
|
1717 |
+
"loss": 0.0261,
|
1718 |
+
"step": 2440
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 2.1933751119068936,
|
1722 |
+
"grad_norm": 0.42108452320098877,
|
1723 |
+
"learning_rate": 5.5807078720992645e-05,
|
1724 |
+
"loss": 0.0259,
|
1725 |
+
"step": 2450
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 2.2023276633840645,
|
1729 |
+
"grad_norm": 0.43240779638290405,
|
1730 |
+
"learning_rate": 5.547849805738836e-05,
|
1731 |
+
"loss": 0.0274,
|
1732 |
+
"step": 2460
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 2.2112802148612354,
|
1736 |
+
"grad_norm": 0.285217821598053,
|
1737 |
+
"learning_rate": 5.514967774653118e-05,
|
1738 |
+
"loss": 0.0217,
|
1739 |
+
"step": 2470
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 2.2202327663384063,
|
1743 |
+
"grad_norm": 0.3506968021392822,
|
1744 |
+
"learning_rate": 5.482063217208674e-05,
|
1745 |
+
"loss": 0.0224,
|
1746 |
+
"step": 2480
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 2.2291853178155776,
|
1750 |
+
"grad_norm": 0.2749573886394501,
|
1751 |
+
"learning_rate": 5.449137572757439e-05,
|
1752 |
+
"loss": 0.0232,
|
1753 |
+
"step": 2490
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 2.2381378692927485,
|
1757 |
+
"grad_norm": 0.3615213930606842,
|
1758 |
+
"learning_rate": 5.4161922815737696e-05,
|
1759 |
+
"loss": 0.0277,
|
1760 |
+
"step": 2500
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 2.2470904207699194,
|
1764 |
+
"grad_norm": 0.26421698927879333,
|
1765 |
+
"learning_rate": 5.3832287847914276e-05,
|
1766 |
+
"loss": 0.0241,
|
1767 |
+
"step": 2510
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 2.2560429722470903,
|
1771 |
+
"grad_norm": 0.20106279850006104,
|
1772 |
+
"learning_rate": 5.35024852434055e-05,
|
1773 |
+
"loss": 0.0206,
|
1774 |
+
"step": 2520
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 2.2649955237242616,
|
1778 |
+
"grad_norm": 0.2898752987384796,
|
1779 |
+
"learning_rate": 5.317252942884567e-05,
|
1780 |
+
"loss": 0.0204,
|
1781 |
+
"step": 2530
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 2.2739480752014325,
|
1785 |
+
"grad_norm": 0.261635959148407,
|
1786 |
+
"learning_rate": 5.284243483757109e-05,
|
1787 |
+
"loss": 0.0265,
|
1788 |
+
"step": 2540
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 2.2829006266786034,
|
1792 |
+
"grad_norm": 0.3588390052318573,
|
1793 |
+
"learning_rate": 5.2512215908988484e-05,
|
1794 |
+
"loss": 0.0269,
|
1795 |
+
"step": 2550
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 2.2918531781557743,
|
1799 |
+
"grad_norm": 0.36631637811660767,
|
1800 |
+
"learning_rate": 5.218188708794357e-05,
|
1801 |
+
"loss": 0.0241,
|
1802 |
+
"step": 2560
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 2.3008057296329456,
|
1806 |
+
"grad_norm": 0.24271060526371002,
|
1807 |
+
"learning_rate": 5.18514628240891e-05,
|
1808 |
+
"loss": 0.0208,
|
1809 |
+
"step": 2570
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 2.3097582811101165,
|
1813 |
+
"grad_norm": 0.282932311296463,
|
1814 |
+
"learning_rate": 5.1520957571252795e-05,
|
1815 |
+
"loss": 0.0219,
|
1816 |
+
"step": 2580
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 2.3187108325872874,
|
1820 |
+
"grad_norm": 0.39644038677215576,
|
1821 |
+
"learning_rate": 5.1190385786805106e-05,
|
1822 |
+
"loss": 0.0232,
|
1823 |
+
"step": 2590
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 2.3276633840644583,
|
1827 |
+
"grad_norm": 0.3940184712409973,
|
1828 |
+
"learning_rate": 5.085976193102677e-05,
|
1829 |
+
"loss": 0.0261,
|
1830 |
+
"step": 2600
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 2.336615935541629,
|
1834 |
+
"grad_norm": 0.3295639753341675,
|
1835 |
+
"learning_rate": 5.052910046647634e-05,
|
1836 |
+
"loss": 0.023,
|
1837 |
+
"step": 2610
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 2.3455684870188005,
|
1841 |
+
"grad_norm": 0.20834366977214813,
|
1842 |
+
"learning_rate": 5.0198415857357464e-05,
|
1843 |
+
"loss": 0.0232,
|
1844 |
+
"step": 2620
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 2.3545210384959714,
|
1848 |
+
"grad_norm": 0.20106403529644012,
|
1849 |
+
"learning_rate": 4.9867722568886223e-05,
|
1850 |
+
"loss": 0.0205,
|
1851 |
+
"step": 2630
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 2.3634735899731423,
|
1855 |
+
"grad_norm": 0.314100056886673,
|
1856 |
+
"learning_rate": 4.9537035066658314e-05,
|
1857 |
+
"loss": 0.019,
|
1858 |
+
"step": 2640
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 2.372426141450313,
|
1862 |
+
"grad_norm": 0.3410167694091797,
|
1863 |
+
"learning_rate": 4.920636781601638e-05,
|
1864 |
+
"loss": 0.0237,
|
1865 |
+
"step": 2650
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 2.3813786929274845,
|
1869 |
+
"grad_norm": 0.29781830310821533,
|
1870 |
+
"learning_rate": 4.88757352814172e-05,
|
1871 |
+
"loss": 0.0267,
|
1872 |
+
"step": 2660
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 2.3903312444046554,
|
1876 |
+
"grad_norm": 0.26218900084495544,
|
1877 |
+
"learning_rate": 4.8545151925798924e-05,
|
1878 |
+
"loss": 0.0214,
|
1879 |
+
"step": 2670
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 2.3992837958818263,
|
1883 |
+
"grad_norm": 0.24904198944568634,
|
1884 |
+
"learning_rate": 4.821463220994848e-05,
|
1885 |
+
"loss": 0.0217,
|
1886 |
+
"step": 2680
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 2.408236347358997,
|
1890 |
+
"grad_norm": 0.3924921751022339,
|
1891 |
+
"learning_rate": 4.788419059186895e-05,
|
1892 |
+
"loss": 0.0225,
|
1893 |
+
"step": 2690
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 2.4171888988361685,
|
1897 |
+
"grad_norm": 0.3247718811035156,
|
1898 |
+
"learning_rate": 4.7553841526147205e-05,
|
1899 |
+
"loss": 0.0238,
|
1900 |
+
"step": 2700
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 2.4261414503133394,
|
1904 |
+
"grad_norm": 0.2955084443092346,
|
1905 |
+
"learning_rate": 4.722359946332156e-05,
|
1906 |
+
"loss": 0.0227,
|
1907 |
+
"step": 2710
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 2.4350940017905103,
|
1911 |
+
"grad_norm": 0.25664591789245605,
|
1912 |
+
"learning_rate": 4.6893478849249654e-05,
|
1913 |
+
"loss": 0.0217,
|
1914 |
+
"step": 2720
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 2.444046553267681,
|
1918 |
+
"grad_norm": 0.2787065804004669,
|
1919 |
+
"learning_rate": 4.656349412447664e-05,
|
1920 |
+
"loss": 0.0242,
|
1921 |
+
"step": 2730
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 2.452999104744852,
|
1925 |
+
"grad_norm": 0.2122863531112671,
|
1926 |
+
"learning_rate": 4.623365972360337e-05,
|
1927 |
+
"loss": 0.0207,
|
1928 |
+
"step": 2740
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 2.4619516562220234,
|
1932 |
+
"grad_norm": 0.34877288341522217,
|
1933 |
+
"learning_rate": 4.590399007465503e-05,
|
1934 |
+
"loss": 0.0187,
|
1935 |
+
"step": 2750
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 2.4709042076991943,
|
1939 |
+
"grad_norm": 0.31759724020957947,
|
1940 |
+
"learning_rate": 4.557449959845005e-05,
|
1941 |
+
"loss": 0.0215,
|
1942 |
+
"step": 2760
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 2.479856759176365,
|
1946 |
+
"grad_norm": 0.30043768882751465,
|
1947 |
+
"learning_rate": 4.524520270796927e-05,
|
1948 |
+
"loss": 0.0185,
|
1949 |
+
"step": 2770
|
1950 |
+
},
|
1951 |
+
{
|
1952 |
+
"epoch": 2.4888093106535365,
|
1953 |
+
"grad_norm": 0.20219744741916656,
|
1954 |
+
"learning_rate": 4.491611380772545e-05,
|
1955 |
+
"loss": 0.019,
|
1956 |
+
"step": 2780
|
1957 |
+
},
|
1958 |
+
{
|
1959 |
+
"epoch": 2.4977618621307074,
|
1960 |
+
"grad_norm": 0.3866671621799469,
|
1961 |
+
"learning_rate": 4.458724729313318e-05,
|
1962 |
+
"loss": 0.0175,
|
1963 |
+
"step": 2790
|
1964 |
+
},
|
1965 |
+
{
|
1966 |
+
"epoch": 2.5067144136078783,
|
1967 |
+
"grad_norm": 0.23905040323734283,
|
1968 |
+
"learning_rate": 4.42586175498792e-05,
|
1969 |
+
"loss": 0.0215,
|
1970 |
+
"step": 2800
|
1971 |
+
},
|
1972 |
+
{
|
1973 |
+
"epoch": 2.515666965085049,
|
1974 |
+
"grad_norm": 0.3837669789791107,
|
1975 |
+
"learning_rate": 4.3930238953293094e-05,
|
1976 |
+
"loss": 0.0208,
|
1977 |
+
"step": 2810
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 2.52461951656222,
|
1981 |
+
"grad_norm": 0.3023552894592285,
|
1982 |
+
"learning_rate": 4.360212586771847e-05,
|
1983 |
+
"loss": 0.0163,
|
1984 |
+
"step": 2820
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"epoch": 2.5335720680393914,
|
1988 |
+
"grad_norm": 0.25578853487968445,
|
1989 |
+
"learning_rate": 4.327429264588463e-05,
|
1990 |
+
"loss": 0.0231,
|
1991 |
+
"step": 2830
|
1992 |
+
},
|
1993 |
+
{
|
1994 |
+
"epoch": 2.5425246195165623,
|
1995 |
+
"grad_norm": 0.2065793126821518,
|
1996 |
+
"learning_rate": 4.2946753628278725e-05,
|
1997 |
+
"loss": 0.0203,
|
1998 |
+
"step": 2840
|
1999 |
+
},
|
2000 |
+
{
|
2001 |
+
"epoch": 2.551477170993733,
|
2002 |
+
"grad_norm": 0.30230146646499634,
|
2003 |
+
"learning_rate": 4.2619523142518474e-05,
|
2004 |
+
"loss": 0.0196,
|
2005 |
+
"step": 2850
|
2006 |
+
},
|
2007 |
+
{
|
2008 |
+
"epoch": 2.560429722470904,
|
2009 |
+
"grad_norm": 0.336173415184021,
|
2010 |
+
"learning_rate": 4.229261550272539e-05,
|
2011 |
+
"loss": 0.0191,
|
2012 |
+
"step": 2860
|
2013 |
+
},
|
2014 |
+
{
|
2015 |
+
"epoch": 2.569382273948075,
|
2016 |
+
"grad_norm": 0.3281089663505554,
|
2017 |
+
"learning_rate": 4.196604500889868e-05,
|
2018 |
+
"loss": 0.0218,
|
2019 |
+
"step": 2870
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 2.5783348254252463,
|
2023 |
+
"grad_norm": 0.15925340354442596,
|
2024 |
+
"learning_rate": 4.163982594628969e-05,
|
2025 |
+
"loss": 0.0185,
|
2026 |
+
"step": 2880
|
2027 |
+
},
|
2028 |
+
{
|
2029 |
+
"epoch": 2.587287376902417,
|
2030 |
+
"grad_norm": 0.12922883033752441,
|
2031 |
+
"learning_rate": 4.131397258477702e-05,
|
2032 |
+
"loss": 0.0192,
|
2033 |
+
"step": 2890
|
2034 |
+
},
|
2035 |
+
{
|
2036 |
+
"epoch": 2.596239928379588,
|
2037 |
+
"grad_norm": 0.25064435601234436,
|
2038 |
+
"learning_rate": 4.0988499178242315e-05,
|
2039 |
+
"loss": 0.0203,
|
2040 |
+
"step": 2900
|
2041 |
+
},
|
2042 |
+
{
|
2043 |
+
"epoch": 2.6051924798567594,
|
2044 |
+
"grad_norm": 0.42107462882995605,
|
2045 |
+
"learning_rate": 4.066341996394678e-05,
|
2046 |
+
"loss": 0.0281,
|
2047 |
+
"step": 2910
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"epoch": 2.6141450313339303,
|
2051 |
+
"grad_norm": 0.3484732210636139,
|
2052 |
+
"learning_rate": 4.033874916190833e-05,
|
2053 |
+
"loss": 0.0203,
|
2054 |
+
"step": 2920
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"epoch": 2.623097582811101,
|
2058 |
+
"grad_norm": 0.3492477834224701,
|
2059 |
+
"learning_rate": 4.001450097427966e-05,
|
2060 |
+
"loss": 0.0196,
|
2061 |
+
"step": 2930
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 2.632050134288272,
|
2065 |
+
"grad_norm": 0.2821838855743408,
|
2066 |
+
"learning_rate": 3.9690689584726894e-05,
|
2067 |
+
"loss": 0.018,
|
2068 |
+
"step": 2940
|
2069 |
+
},
|
2070 |
+
{
|
2071 |
+
"epoch": 2.641002685765443,
|
2072 |
+
"grad_norm": 0.3413795232772827,
|
2073 |
+
"learning_rate": 3.936732915780923e-05,
|
2074 |
+
"loss": 0.019,
|
2075 |
+
"step": 2950
|
2076 |
+
},
|
2077 |
+
{
|
2078 |
+
"epoch": 2.6499552372426143,
|
2079 |
+
"grad_norm": 0.2790643572807312,
|
2080 |
+
"learning_rate": 3.904443383835929e-05,
|
2081 |
+
"loss": 0.0243,
|
2082 |
+
"step": 2960
|
2083 |
+
},
|
2084 |
+
{
|
2085 |
+
"epoch": 2.658907788719785,
|
2086 |
+
"grad_norm": 0.20392818748950958,
|
2087 |
+
"learning_rate": 3.872201775086437e-05,
|
2088 |
+
"loss": 0.0225,
|
2089 |
+
"step": 2970
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 2.667860340196956,
|
2093 |
+
"grad_norm": 0.2968049943447113,
|
2094 |
+
"learning_rate": 3.8400094998848616e-05,
|
2095 |
+
"loss": 0.0178,
|
2096 |
+
"step": 2980
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 2.6768128916741274,
|
2100 |
+
"grad_norm": 0.26228010654449463,
|
2101 |
+
"learning_rate": 3.807867966425611e-05,
|
2102 |
+
"loss": 0.0182,
|
2103 |
+
"step": 2990
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 2.6857654431512983,
|
2107 |
+
"grad_norm": 0.3043522238731384,
|
2108 |
+
"learning_rate": 3.775778580683481e-05,
|
2109 |
+
"loss": 0.0186,
|
2110 |
+
"step": 3000
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 2.694717994628469,
|
2114 |
+
"grad_norm": 0.2954307794570923,
|
2115 |
+
"learning_rate": 3.743742746352156e-05,
|
2116 |
+
"loss": 0.0227,
|
2117 |
+
"step": 3010
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 2.70367054610564,
|
2121 |
+
"grad_norm": 0.2497381865978241,
|
2122 |
+
"learning_rate": 3.711761864782817e-05,
|
2123 |
+
"loss": 0.0214,
|
2124 |
+
"step": 3020
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 2.712623097582811,
|
2128 |
+
"grad_norm": 0.3127233684062958,
|
2129 |
+
"learning_rate": 3.679837334922825e-05,
|
2130 |
+
"loss": 0.019,
|
2131 |
+
"step": 3030
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 2.7215756490599823,
|
2135 |
+
"grad_norm": 0.18976886570453644,
|
2136 |
+
"learning_rate": 3.647970553254538e-05,
|
2137 |
+
"loss": 0.0159,
|
2138 |
+
"step": 3040
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 2.730528200537153,
|
2142 |
+
"grad_norm": 0.17379194498062134,
|
2143 |
+
"learning_rate": 3.61616291373422e-05,
|
2144 |
+
"loss": 0.0212,
|
2145 |
+
"step": 3050
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 2.739480752014324,
|
2149 |
+
"grad_norm": 0.2456379234790802,
|
2150 |
+
"learning_rate": 3.584415807731065e-05,
|
2151 |
+
"loss": 0.0204,
|
2152 |
+
"step": 3060
|
2153 |
+
},
|
2154 |
+
{
|
2155 |
+
"epoch": 2.748433303491495,
|
2156 |
+
"grad_norm": 0.35977035760879517,
|
2157 |
+
"learning_rate": 3.552730623966337e-05,
|
2158 |
+
"loss": 0.0253,
|
2159 |
+
"step": 3070
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"epoch": 2.757385854968666,
|
2163 |
+
"grad_norm": 0.3271699547767639,
|
2164 |
+
"learning_rate": 3.521108748452617e-05,
|
2165 |
+
"loss": 0.0209,
|
2166 |
+
"step": 3080
|
2167 |
+
},
|
2168 |
+
{
|
2169 |
+
"epoch": 2.766338406445837,
|
2170 |
+
"grad_norm": 0.17938630282878876,
|
2171 |
+
"learning_rate": 3.489551564433186e-05,
|
2172 |
+
"loss": 0.0223,
|
2173 |
+
"step": 3090
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"epoch": 2.775290957923008,
|
2177 |
+
"grad_norm": 0.3276021480560303,
|
2178 |
+
"learning_rate": 3.4580604523215006e-05,
|
2179 |
+
"loss": 0.0217,
|
2180 |
+
"step": 3100
|
2181 |
+
},
|
2182 |
+
{
|
2183 |
+
"epoch": 2.784243509400179,
|
2184 |
+
"grad_norm": 0.25213152170181274,
|
2185 |
+
"learning_rate": 3.4266367896408216e-05,
|
2186 |
+
"loss": 0.0233,
|
2187 |
+
"step": 3110
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 2.7931960608773503,
|
2191 |
+
"grad_norm": 0.19577741622924805,
|
2192 |
+
"learning_rate": 3.3952819509639534e-05,
|
2193 |
+
"loss": 0.0168,
|
2194 |
+
"step": 3120
|
2195 |
+
},
|
2196 |
+
{
|
2197 |
+
"epoch": 2.802148612354521,
|
2198 |
+
"grad_norm": 0.33538851141929626,
|
2199 |
+
"learning_rate": 3.3639973078531165e-05,
|
2200 |
+
"loss": 0.0215,
|
2201 |
+
"step": 3130
|
2202 |
+
},
|
2203 |
+
{
|
2204 |
+
"epoch": 2.811101163831692,
|
2205 |
+
"grad_norm": 0.3436635434627533,
|
2206 |
+
"learning_rate": 3.332784228799947e-05,
|
2207 |
+
"loss": 0.0208,
|
2208 |
+
"step": 3140
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"epoch": 2.820053715308863,
|
2212 |
+
"grad_norm": 0.25322389602661133,
|
2213 |
+
"learning_rate": 3.301644079165638e-05,
|
2214 |
+
"loss": 0.0191,
|
2215 |
+
"step": 3150
|
2216 |
+
},
|
2217 |
+
{
|
2218 |
+
"epoch": 2.829006266786034,
|
2219 |
+
"grad_norm": 0.3786560595035553,
|
2220 |
+
"learning_rate": 3.27057822112122e-05,
|
2221 |
+
"loss": 0.0188,
|
2222 |
+
"step": 3160
|
2223 |
+
},
|
2224 |
+
{
|
2225 |
+
"epoch": 2.837958818263205,
|
2226 |
+
"grad_norm": 0.20764805376529694,
|
2227 |
+
"learning_rate": 3.239588013587958e-05,
|
2228 |
+
"loss": 0.0218,
|
2229 |
+
"step": 3170
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 2.846911369740376,
|
2233 |
+
"grad_norm": 0.17294557392597198,
|
2234 |
+
"learning_rate": 3.208674812177926e-05,
|
2235 |
+
"loss": 0.02,
|
2236 |
+
"step": 3180
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"epoch": 2.855863921217547,
|
2240 |
+
"grad_norm": 0.19973523914813995,
|
2241 |
+
"learning_rate": 3.177839969134698e-05,
|
2242 |
+
"loss": 0.0186,
|
2243 |
+
"step": 3190
|
2244 |
+
},
|
2245 |
+
{
|
2246 |
+
"epoch": 2.864816472694718,
|
2247 |
+
"grad_norm": 0.30084753036499023,
|
2248 |
+
"learning_rate": 3.1470848332742e-05,
|
2249 |
+
"loss": 0.0205,
|
2250 |
+
"step": 3200
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"epoch": 2.8737690241718887,
|
2254 |
+
"grad_norm": 0.371510773897171,
|
2255 |
+
"learning_rate": 3.116410749925708e-05,
|
2256 |
+
"loss": 0.0205,
|
2257 |
+
"step": 3210
|
2258 |
+
},
|
2259 |
+
{
|
2260 |
+
"epoch": 2.88272157564906,
|
2261 |
+
"grad_norm": 0.24790452420711517,
|
2262 |
+
"learning_rate": 3.085819060872995e-05,
|
2263 |
+
"loss": 0.017,
|
2264 |
+
"step": 3220
|
2265 |
+
},
|
2266 |
+
{
|
2267 |
+
"epoch": 2.891674127126231,
|
2268 |
+
"grad_norm": 0.18967312574386597,
|
2269 |
+
"learning_rate": 3.055311104295648e-05,
|
2270 |
+
"loss": 0.018,
|
2271 |
+
"step": 3230
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 2.900626678603402,
|
2275 |
+
"grad_norm": 0.20764872431755066,
|
2276 |
+
"learning_rate": 3.024888214710517e-05,
|
2277 |
+
"loss": 0.0174,
|
2278 |
+
"step": 3240
|
2279 |
+
},
|
2280 |
+
{
|
2281 |
+
"epoch": 2.909579230080573,
|
2282 |
+
"grad_norm": 0.2953355610370636,
|
2283 |
+
"learning_rate": 2.994551722913349e-05,
|
2284 |
+
"loss": 0.0151,
|
2285 |
+
"step": 3250
|
2286 |
+
},
|
2287 |
+
{
|
2288 |
+
"epoch": 2.918531781557744,
|
2289 |
+
"grad_norm": 0.237528458237648,
|
2290 |
+
"learning_rate": 2.9643029559205727e-05,
|
2291 |
+
"loss": 0.023,
|
2292 |
+
"step": 3260
|
2293 |
+
},
|
2294 |
+
{
|
2295 |
+
"epoch": 2.927484333034915,
|
2296 |
+
"grad_norm": 0.225080668926239,
|
2297 |
+
"learning_rate": 2.934143236911248e-05,
|
2298 |
+
"loss": 0.0174,
|
2299 |
+
"step": 3270
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"epoch": 2.936436884512086,
|
2303 |
+
"grad_norm": 0.1836756318807602,
|
2304 |
+
"learning_rate": 2.90407388516919e-05,
|
2305 |
+
"loss": 0.0163,
|
2306 |
+
"step": 3280
|
2307 |
+
},
|
2308 |
+
{
|
2309 |
+
"epoch": 2.9453894359892567,
|
2310 |
+
"grad_norm": 0.1732194870710373,
|
2311 |
+
"learning_rate": 2.8740962160252495e-05,
|
2312 |
+
"loss": 0.0183,
|
2313 |
+
"step": 3290
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 2.954341987466428,
|
2317 |
+
"grad_norm": 0.42215636372566223,
|
2318 |
+
"learning_rate": 2.844211540799797e-05,
|
2319 |
+
"loss": 0.0169,
|
2320 |
+
"step": 3300
|
2321 |
+
},
|
2322 |
+
{
|
2323 |
+
"epoch": 2.963294538943599,
|
2324 |
+
"grad_norm": 0.15775205194950104,
|
2325 |
+
"learning_rate": 2.8144211667453368e-05,
|
2326 |
+
"loss": 0.0162,
|
2327 |
+
"step": 3310
|
2328 |
+
},
|
2329 |
+
{
|
2330 |
+
"epoch": 2.97224709042077,
|
2331 |
+
"grad_norm": 0.2212500423192978,
|
2332 |
+
"learning_rate": 2.7847263969893344e-05,
|
2333 |
+
"loss": 0.0152,
|
2334 |
+
"step": 3320
|
2335 |
+
},
|
2336 |
+
{
|
2337 |
+
"epoch": 2.981199641897941,
|
2338 |
+
"grad_norm": 0.43953937292099,
|
2339 |
+
"learning_rate": 2.7551285304772206e-05,
|
2340 |
+
"loss": 0.0168,
|
2341 |
+
"step": 3330
|
2342 |
+
},
|
2343 |
+
{
|
2344 |
+
"epoch": 2.990152193375112,
|
2345 |
+
"grad_norm": 0.2736935317516327,
|
2346 |
+
"learning_rate": 2.7256288619155567e-05,
|
2347 |
+
"loss": 0.018,
|
2348 |
+
"step": 3340
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"epoch": 2.999104744852283,
|
2352 |
+
"grad_norm": 0.2877485156059265,
|
2353 |
+
"learning_rate": 2.6962286817154158e-05,
|
2354 |
+
"loss": 0.0203,
|
2355 |
+
"step": 3350
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 3.008057296329454,
|
2359 |
+
"grad_norm": 0.21356728672981262,
|
2360 |
+
"learning_rate": 2.6669292759359166e-05,
|
2361 |
+
"loss": 0.0175,
|
2362 |
+
"step": 3360
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"epoch": 3.0170098478066247,
|
2366 |
+
"grad_norm": 0.29351386427879333,
|
2367 |
+
"learning_rate": 2.637731926227993e-05,
|
2368 |
+
"loss": 0.0255,
|
2369 |
+
"step": 3370
|
2370 |
+
},
|
2371 |
+
{
|
2372 |
+
"epoch": 3.025962399283796,
|
2373 |
+
"grad_norm": 0.30802270770072937,
|
2374 |
+
"learning_rate": 2.6086379097783033e-05,
|
2375 |
+
"loss": 0.0185,
|
2376 |
+
"step": 3380
|
2377 |
+
},
|
2378 |
+
{
|
2379 |
+
"epoch": 3.034914950760967,
|
2380 |
+
"grad_norm": 0.23744232952594757,
|
2381 |
+
"learning_rate": 2.579648499253377e-05,
|
2382 |
+
"loss": 0.0172,
|
2383 |
+
"step": 3390
|
2384 |
+
},
|
2385 |
+
{
|
2386 |
+
"epoch": 3.043867502238138,
|
2387 |
+
"grad_norm": 0.33891722559928894,
|
2388 |
+
"learning_rate": 2.5507649627439466e-05,
|
2389 |
+
"loss": 0.0201,
|
2390 |
+
"step": 3400
|
2391 |
+
},
|
2392 |
+
{
|
2393 |
+
"epoch": 3.0528200537153087,
|
2394 |
+
"grad_norm": 0.12712599337100983,
|
2395 |
+
"learning_rate": 2.5219885637094653e-05,
|
2396 |
+
"loss": 0.0181,
|
2397 |
+
"step": 3410
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 3.06177260519248,
|
2401 |
+
"grad_norm": 0.25749579071998596,
|
2402 |
+
"learning_rate": 2.4933205609228533e-05,
|
2403 |
+
"loss": 0.0188,
|
2404 |
+
"step": 3420
|
2405 |
+
},
|
2406 |
+
{
|
2407 |
+
"epoch": 3.070725156669651,
|
2408 |
+
"grad_norm": 0.25427231192588806,
|
2409 |
+
"learning_rate": 2.464762208415419e-05,
|
2410 |
+
"loss": 0.0149,
|
2411 |
+
"step": 3430
|
2412 |
+
},
|
2413 |
+
{
|
2414 |
+
"epoch": 3.079677708146822,
|
2415 |
+
"grad_norm": 0.2582094669342041,
|
2416 |
+
"learning_rate": 2.4363147554220213e-05,
|
2417 |
+
"loss": 0.0184,
|
2418 |
+
"step": 3440
|
2419 |
+
},
|
2420 |
+
{
|
2421 |
+
"epoch": 3.0886302596239927,
|
2422 |
+
"grad_norm": 0.268612265586853,
|
2423 |
+
"learning_rate": 2.407979446326411e-05,
|
2424 |
+
"loss": 0.0147,
|
2425 |
+
"step": 3450
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"epoch": 3.0975828111011636,
|
2429 |
+
"grad_norm": 0.22941374778747559,
|
2430 |
+
"learning_rate": 2.379757520606799e-05,
|
2431 |
+
"loss": 0.0195,
|
2432 |
+
"step": 3460
|
2433 |
+
},
|
2434 |
+
{
|
2435 |
+
"epoch": 3.106535362578335,
|
2436 |
+
"grad_norm": 0.2100794017314911,
|
2437 |
+
"learning_rate": 2.3516502127816455e-05,
|
2438 |
+
"loss": 0.016,
|
2439 |
+
"step": 3470
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 3.115487914055506,
|
2443 |
+
"grad_norm": 0.13210788369178772,
|
2444 |
+
"learning_rate": 2.323658752355647e-05,
|
2445 |
+
"loss": 0.017,
|
2446 |
+
"step": 3480
|
2447 |
+
},
|
2448 |
+
{
|
2449 |
+
"epoch": 3.1244404655326767,
|
2450 |
+
"grad_norm": 0.18918994069099426,
|
2451 |
+
"learning_rate": 2.2957843637659654e-05,
|
2452 |
+
"loss": 0.0165,
|
2453 |
+
"step": 3490
|
2454 |
+
},
|
2455 |
+
{
|
2456 |
+
"epoch": 3.1333930170098476,
|
2457 |
+
"grad_norm": 0.3535359501838684,
|
2458 |
+
"learning_rate": 2.2680282663286552e-05,
|
2459 |
+
"loss": 0.0209,
|
2460 |
+
"step": 3500
|
2461 |
+
},
|
2462 |
+
{
|
2463 |
+
"epoch": 3.142345568487019,
|
2464 |
+
"grad_norm": 0.2309960275888443,
|
2465 |
+
"learning_rate": 2.2403916741853364e-05,
|
2466 |
+
"loss": 0.0167,
|
2467 |
+
"step": 3510
|
2468 |
+
},
|
2469 |
+
{
|
2470 |
+
"epoch": 3.15129811996419,
|
2471 |
+
"grad_norm": 0.21955038607120514,
|
2472 |
+
"learning_rate": 2.2128757962500817e-05,
|
2473 |
+
"loss": 0.0164,
|
2474 |
+
"step": 3520
|
2475 |
+
},
|
2476 |
+
{
|
2477 |
+
"epoch": 3.1602506714413607,
|
2478 |
+
"grad_norm": 0.24366547167301178,
|
2479 |
+
"learning_rate": 2.1854818361565275e-05,
|
2480 |
+
"loss": 0.015,
|
2481 |
+
"step": 3530
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 3.1692032229185316,
|
2485 |
+
"grad_norm": 0.25391441583633423,
|
2486 |
+
"learning_rate": 2.1582109922052364e-05,
|
2487 |
+
"loss": 0.0192,
|
2488 |
+
"step": 3540
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"epoch": 3.178155774395703,
|
2492 |
+
"grad_norm": 0.2967330515384674,
|
2493 |
+
"learning_rate": 2.1310644573112635e-05,
|
2494 |
+
"loss": 0.0168,
|
2495 |
+
"step": 3550
|
2496 |
+
},
|
2497 |
+
{
|
2498 |
+
"epoch": 3.187108325872874,
|
2499 |
+
"grad_norm": 0.2360481321811676,
|
2500 |
+
"learning_rate": 2.1040434189519924e-05,
|
2501 |
+
"loss": 0.0163,
|
2502 |
+
"step": 3560
|
2503 |
+
},
|
2504 |
+
{
|
2505 |
+
"epoch": 3.1960608773500447,
|
2506 |
+
"grad_norm": 0.2346932291984558,
|
2507 |
+
"learning_rate": 2.0771490591151733e-05,
|
2508 |
+
"loss": 0.0181,
|
2509 |
+
"step": 3570
|
2510 |
+
},
|
2511 |
+
{
|
2512 |
+
"epoch": 3.2050134288272156,
|
2513 |
+
"grad_norm": 0.24760273098945618,
|
2514 |
+
"learning_rate": 2.0503825542472317e-05,
|
2515 |
+
"loss": 0.0175,
|
2516 |
+
"step": 3580
|
2517 |
+
},
|
2518 |
+
{
|
2519 |
+
"epoch": 3.213965980304387,
|
2520 |
+
"grad_norm": 0.19240783154964447,
|
2521 |
+
"learning_rate": 2.023745075201805e-05,
|
2522 |
+
"loss": 0.0169,
|
2523 |
+
"step": 3590
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 3.222918531781558,
|
2527 |
+
"grad_norm": 0.2748640477657318,
|
2528 |
+
"learning_rate": 1.9972377871885157e-05,
|
2529 |
+
"loss": 0.0179,
|
2530 |
+
"step": 3600
|
2531 |
+
},
|
2532 |
+
{
|
2533 |
+
"epoch": 3.2318710832587287,
|
2534 |
+
"grad_norm": 0.22245590388774872,
|
2535 |
+
"learning_rate": 1.970861849722017e-05,
|
2536 |
+
"loss": 0.0136,
|
2537 |
+
"step": 3610
|
2538 |
+
},
|
2539 |
+
{
|
2540 |
+
"epoch": 3.2408236347358996,
|
2541 |
+
"grad_norm": 0.24653933942317963,
|
2542 |
+
"learning_rate": 1.9446184165712587e-05,
|
2543 |
+
"loss": 0.0161,
|
2544 |
+
"step": 3620
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 3.2497761862130705,
|
2548 |
+
"grad_norm": 0.21308885514736176,
|
2549 |
+
"learning_rate": 1.9185086357090214e-05,
|
2550 |
+
"loss": 0.0156,
|
2551 |
+
"step": 3630
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 3.258728737690242,
|
2555 |
+
"grad_norm": 0.1842220425605774,
|
2556 |
+
"learning_rate": 1.8925336492617057e-05,
|
2557 |
+
"loss": 0.0198,
|
2558 |
+
"step": 3640
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 3.2676812891674127,
|
2562 |
+
"grad_norm": 0.28430524468421936,
|
2563 |
+
"learning_rate": 1.8666945934593666e-05,
|
2564 |
+
"loss": 0.0174,
|
2565 |
+
"step": 3650
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 3.2766338406445836,
|
2569 |
+
"grad_norm": 0.22021664679050446,
|
2570 |
+
"learning_rate": 1.8409925985860126e-05,
|
2571 |
+
"loss": 0.0143,
|
2572 |
+
"step": 3660
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 3.2855863921217545,
|
2576 |
+
"grad_norm": 0.26781344413757324,
|
2577 |
+
"learning_rate": 1.8154287889301603e-05,
|
2578 |
+
"loss": 0.015,
|
2579 |
+
"step": 3670
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 3.294538943598926,
|
2583 |
+
"grad_norm": 0.19385065138339996,
|
2584 |
+
"learning_rate": 1.7900042827356612e-05,
|
2585 |
+
"loss": 0.0169,
|
2586 |
+
"step": 3680
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 3.3034914950760967,
|
2590 |
+
"grad_norm": 0.1983576864004135,
|
2591 |
+
"learning_rate": 1.76472019215278e-05,
|
2592 |
+
"loss": 0.0164,
|
2593 |
+
"step": 3690
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 3.3124440465532676,
|
2597 |
+
"grad_norm": 0.2234923541545868,
|
2598 |
+
"learning_rate": 1.739577623189545e-05,
|
2599 |
+
"loss": 0.0158,
|
2600 |
+
"step": 3700
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 3.3213965980304385,
|
2604 |
+
"grad_norm": 0.163887158036232,
|
2605 |
+
"learning_rate": 1.7145776756633768e-05,
|
2606 |
+
"loss": 0.0201,
|
2607 |
+
"step": 3710
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 3.33034914950761,
|
2611 |
+
"grad_norm": 0.1897057443857193,
|
2612 |
+
"learning_rate": 1.6897214431529646e-05,
|
2613 |
+
"loss": 0.0175,
|
2614 |
+
"step": 3720
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 3.3393017009847807,
|
2618 |
+
"grad_norm": 0.21445390582084656,
|
2619 |
+
"learning_rate": 1.6650100129504475e-05,
|
2620 |
+
"loss": 0.0138,
|
2621 |
+
"step": 3730
|
2622 |
+
},
|
2623 |
+
{
|
2624 |
+
"epoch": 3.3482542524619516,
|
2625 |
+
"grad_norm": 0.2707969546318054,
|
2626 |
+
"learning_rate": 1.6404444660138335e-05,
|
2627 |
+
"loss": 0.0162,
|
2628 |
+
"step": 3740
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 3.3572068039391225,
|
2632 |
+
"grad_norm": 0.3275142312049866,
|
2633 |
+
"learning_rate": 1.616025876919725e-05,
|
2634 |
+
"loss": 0.0158,
|
2635 |
+
"step": 3750
|
2636 |
+
},
|
2637 |
+
{
|
2638 |
+
"epoch": 3.3661593554162934,
|
2639 |
+
"grad_norm": 0.23695893585681915,
|
2640 |
+
"learning_rate": 1.5917553138163172e-05,
|
2641 |
+
"loss": 0.0174,
|
2642 |
+
"step": 3760
|
2643 |
+
},
|
2644 |
+
{
|
2645 |
+
"epoch": 3.3751119068934647,
|
2646 |
+
"grad_norm": 0.17175161838531494,
|
2647 |
+
"learning_rate": 1.5676338383766632e-05,
|
2648 |
+
"loss": 0.0197,
|
2649 |
+
"step": 3770
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 3.3840644583706356,
|
2653 |
+
"grad_norm": 0.17486472427845,
|
2654 |
+
"learning_rate": 1.5436625057522447e-05,
|
2655 |
+
"loss": 0.0145,
|
2656 |
+
"step": 3780
|
2657 |
+
},
|
2658 |
+
{
|
2659 |
+
"epoch": 3.3930170098478065,
|
2660 |
+
"grad_norm": 0.2262464016675949,
|
2661 |
+
"learning_rate": 1.519842364526804e-05,
|
2662 |
+
"loss": 0.0129,
|
2663 |
+
"step": 3790
|
2664 |
+
},
|
2665 |
+
{
|
2666 |
+
"epoch": 3.401969561324978,
|
2667 |
+
"grad_norm": 0.2933650016784668,
|
2668 |
+
"learning_rate": 1.4961744566704855e-05,
|
2669 |
+
"loss": 0.0163,
|
2670 |
+
"step": 3800
|
2671 |
+
},
|
2672 |
+
{
|
2673 |
+
"epoch": 3.4109221128021487,
|
2674 |
+
"grad_norm": 0.1765672117471695,
|
2675 |
+
"learning_rate": 1.4726598174942551e-05,
|
2676 |
+
"loss": 0.0132,
|
2677 |
+
"step": 3810
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"epoch": 3.4198746642793196,
|
2681 |
+
"grad_norm": 0.22736941277980804,
|
2682 |
+
"learning_rate": 1.4492994756046035e-05,
|
2683 |
+
"loss": 0.0171,
|
2684 |
+
"step": 3820
|
2685 |
+
},
|
2686 |
+
{
|
2687 |
+
"epoch": 3.4288272157564905,
|
2688 |
+
"grad_norm": 0.16248083114624023,
|
2689 |
+
"learning_rate": 1.4260944528585645e-05,
|
2690 |
+
"loss": 0.0167,
|
2691 |
+
"step": 3830
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 3.4377797672336614,
|
2695 |
+
"grad_norm": 0.2661954164505005,
|
2696 |
+
"learning_rate": 1.4030457643190048e-05,
|
2697 |
+
"loss": 0.0141,
|
2698 |
+
"step": 3840
|
2699 |
+
},
|
2700 |
+
{
|
2701 |
+
"epoch": 3.4467323187108327,
|
2702 |
+
"grad_norm": 0.1929498016834259,
|
2703 |
+
"learning_rate": 1.3801544182102311e-05,
|
2704 |
+
"loss": 0.0154,
|
2705 |
+
"step": 3850
|
2706 |
+
},
|
2707 |
+
{
|
2708 |
+
"epoch": 3.4556848701880036,
|
2709 |
+
"grad_norm": 0.21110036969184875,
|
2710 |
+
"learning_rate": 1.3574214158738763e-05,
|
2711 |
+
"loss": 0.0168,
|
2712 |
+
"step": 3860
|
2713 |
+
},
|
2714 |
+
{
|
2715 |
+
"epoch": 3.4646374216651745,
|
2716 |
+
"grad_norm": 0.35853829979896545,
|
2717 |
+
"learning_rate": 1.3348477517251101e-05,
|
2718 |
+
"loss": 0.0158,
|
2719 |
+
"step": 3870
|
2720 |
+
},
|
2721 |
+
{
|
2722 |
+
"epoch": 3.4735899731423454,
|
2723 |
+
"grad_norm": 0.20026364922523499,
|
2724 |
+
"learning_rate": 1.312434413209131e-05,
|
2725 |
+
"loss": 0.0138,
|
2726 |
+
"step": 3880
|
2727 |
+
},
|
2728 |
+
{
|
2729 |
+
"epoch": 3.4825425246195167,
|
2730 |
+
"grad_norm": 0.2420613169670105,
|
2731 |
+
"learning_rate": 1.2901823807579727e-05,
|
2732 |
+
"loss": 0.0146,
|
2733 |
+
"step": 3890
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 3.4914950760966876,
|
2737 |
+
"grad_norm": 0.2799060344696045,
|
2738 |
+
"learning_rate": 1.2680926277476245e-05,
|
2739 |
+
"loss": 0.0159,
|
2740 |
+
"step": 3900
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 3.5004476275738585,
|
2744 |
+
"grad_norm": 0.2773265242576599,
|
2745 |
+
"learning_rate": 1.2461661204554397e-05,
|
2746 |
+
"loss": 0.0151,
|
2747 |
+
"step": 3910
|
2748 |
+
},
|
2749 |
+
{
|
2750 |
+
"epoch": 3.5094001790510294,
|
2751 |
+
"grad_norm": 0.18590183556079865,
|
2752 |
+
"learning_rate": 1.2244038180178835e-05,
|
2753 |
+
"loss": 0.0147,
|
2754 |
+
"step": 3920
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 3.5183527305282007,
|
2758 |
+
"grad_norm": 0.19557642936706543,
|
2759 |
+
"learning_rate": 1.2028066723885612e-05,
|
2760 |
+
"loss": 0.0154,
|
2761 |
+
"step": 3930
|
2762 |
+
},
|
2763 |
+
{
|
2764 |
+
"epoch": 3.5273052820053716,
|
2765 |
+
"grad_norm": 0.22107233107089996,
|
2766 |
+
"learning_rate": 1.1813756282965888e-05,
|
2767 |
+
"loss": 0.0154,
|
2768 |
+
"step": 3940
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 3.5362578334825425,
|
2772 |
+
"grad_norm": 0.18053828179836273,
|
2773 |
+
"learning_rate": 1.1601116232052638e-05,
|
2774 |
+
"loss": 0.0165,
|
2775 |
+
"step": 3950
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 3.5452103849597134,
|
2779 |
+
"grad_norm": 0.21140411496162415,
|
2780 |
+
"learning_rate": 1.1390155872710517e-05,
|
2781 |
+
"loss": 0.0134,
|
2782 |
+
"step": 3960
|
2783 |
+
},
|
2784 |
+
{
|
2785 |
+
"epoch": 3.5541629364368843,
|
2786 |
+
"grad_norm": 0.16849522292613983,
|
2787 |
+
"learning_rate": 1.1180884433029087e-05,
|
2788 |
+
"loss": 0.0165,
|
2789 |
+
"step": 3970
|
2790 |
+
},
|
2791 |
+
{
|
2792 |
+
"epoch": 3.5631154879140556,
|
2793 |
+
"grad_norm": 0.14933831989765167,
|
2794 |
+
"learning_rate": 1.097331106721904e-05,
|
2795 |
+
"loss": 0.0124,
|
2796 |
+
"step": 3980
|
2797 |
+
},
|
2798 |
+
{
|
2799 |
+
"epoch": 3.5720680393912265,
|
2800 |
+
"grad_norm": 0.217430979013443,
|
2801 |
+
"learning_rate": 1.0767444855211862e-05,
|
2802 |
+
"loss": 0.0135,
|
2803 |
+
"step": 3990
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"epoch": 3.5810205908683974,
|
2807 |
+
"grad_norm": 0.21716763079166412,
|
2808 |
+
"learning_rate": 1.0563294802262558e-05,
|
2809 |
+
"loss": 0.0167,
|
2810 |
+
"step": 4000
|
2811 |
+
},
|
2812 |
+
{
|
2813 |
+
"epoch": 3.5899731423455687,
|
2814 |
+
"grad_norm": 0.20996397733688354,
|
2815 |
+
"learning_rate": 1.0360869838555809e-05,
|
2816 |
+
"loss": 0.0131,
|
2817 |
+
"step": 4010
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 3.5989256938227396,
|
2821 |
+
"grad_norm": 0.22892490029335022,
|
2822 |
+
"learning_rate": 1.0160178818815313e-05,
|
2823 |
+
"loss": 0.0134,
|
2824 |
+
"step": 4020
|
2825 |
+
},
|
2826 |
+
{
|
2827 |
+
"epoch": 3.6078782452999105,
|
2828 |
+
"grad_norm": 0.22495168447494507,
|
2829 |
+
"learning_rate": 9.961230521916387e-06,
|
2830 |
+
"loss": 0.0146,
|
2831 |
+
"step": 4030
|
2832 |
+
},
|
2833 |
+
{
|
2834 |
+
"epoch": 3.6168307967770814,
|
2835 |
+
"grad_norm": 0.1184273287653923,
|
2836 |
+
"learning_rate": 9.764033650502074e-06,
|
2837 |
+
"loss": 0.0143,
|
2838 |
+
"step": 4040
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 3.6257833482542523,
|
2842 |
+
"grad_norm": 0.18436811864376068,
|
2843 |
+
"learning_rate": 9.568596830602344e-06,
|
2844 |
+
"loss": 0.0114,
|
2845 |
+
"step": 4050
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 3.6347358997314236,
|
2849 |
+
"grad_norm": 0.2092517912387848,
|
2850 |
+
"learning_rate": 9.37492861125681e-06,
|
2851 |
+
"loss": 0.0164,
|
2852 |
+
"step": 4060
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 3.6436884512085945,
|
2856 |
+
"grad_norm": 0.1160958930850029,
|
2857 |
+
"learning_rate": 9.183037464140804e-06,
|
2858 |
+
"loss": 0.0162,
|
2859 |
+
"step": 4070
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 3.6526410026857654,
|
2863 |
+
"grad_norm": 0.24437487125396729,
|
2864 |
+
"learning_rate": 8.992931783194735e-06,
|
2865 |
+
"loss": 0.0131,
|
2866 |
+
"step": 4080
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 3.6615935541629363,
|
2870 |
+
"grad_norm": 0.185224711894989,
|
2871 |
+
"learning_rate": 8.80461988425696e-06,
|
2872 |
+
"loss": 0.0155,
|
2873 |
+
"step": 4090
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 3.670546105640107,
|
2877 |
+
"grad_norm": 0.2819221019744873,
|
2878 |
+
"learning_rate": 8.618110004699974e-06,
|
2879 |
+
"loss": 0.0148,
|
2880 |
+
"step": 4100
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 3.6794986571172785,
|
2884 |
+
"grad_norm": 0.24776029586791992,
|
2885 |
+
"learning_rate": 8.4334103030701e-06,
|
2886 |
+
"loss": 0.0162,
|
2887 |
+
"step": 4110
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 3.6884512085944494,
|
2891 |
+
"grad_norm": 0.19672740995883942,
|
2892 |
+
"learning_rate": 8.25052885873066e-06,
|
2893 |
+
"loss": 0.0136,
|
2894 |
+
"step": 4120
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 3.6974037600716203,
|
2898 |
+
"grad_norm": 0.26846009492874146,
|
2899 |
+
"learning_rate": 8.06947367150846e-06,
|
2900 |
+
"loss": 0.0139,
|
2901 |
+
"step": 4130
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 3.7063563115487916,
|
2905 |
+
"grad_norm": 0.15493078529834747,
|
2906 |
+
"learning_rate": 7.890252661343938e-06,
|
2907 |
+
"loss": 0.0121,
|
2908 |
+
"step": 4140
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 3.7153088630259625,
|
2912 |
+
"grad_norm": 0.19229039549827576,
|
2913 |
+
"learning_rate": 7.712873667944681e-06,
|
2914 |
+
"loss": 0.0137,
|
2915 |
+
"step": 4150
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 3.7242614145031334,
|
2919 |
+
"grad_norm": 0.2311052680015564,
|
2920 |
+
"learning_rate": 7.537344450442469e-06,
|
2921 |
+
"loss": 0.0104,
|
2922 |
+
"step": 4160
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 3.7332139659803043,
|
2926 |
+
"grad_norm": 0.19902154803276062,
|
2927 |
+
"learning_rate": 7.36367268705393e-06,
|
2928 |
+
"loss": 0.0173,
|
2929 |
+
"step": 4170
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 3.742166517457475,
|
2933 |
+
"grad_norm": 0.2320551574230194,
|
2934 |
+
"learning_rate": 7.1918659747446e-06,
|
2935 |
+
"loss": 0.0144,
|
2936 |
+
"step": 4180
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 3.7511190689346465,
|
2940 |
+
"grad_norm": 0.1548873782157898,
|
2941 |
+
"learning_rate": 7.021931828896666e-06,
|
2942 |
+
"loss": 0.0161,
|
2943 |
+
"step": 4190
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 3.7600716204118174,
|
2947 |
+
"grad_norm": 0.30775901675224304,
|
2948 |
+
"learning_rate": 6.8538776829801584e-06,
|
2949 |
+
"loss": 0.0134,
|
2950 |
+
"step": 4200
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 3.7690241718889883,
|
2954 |
+
"grad_norm": 0.36373430490493774,
|
2955 |
+
"learning_rate": 6.687710888227849e-06,
|
2956 |
+
"loss": 0.0115,
|
2957 |
+
"step": 4210
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 3.777976723366159,
|
2961 |
+
"grad_norm": 0.20033209025859833,
|
2962 |
+
"learning_rate": 6.5234387133136565e-06,
|
2963 |
+
"loss": 0.0109,
|
2964 |
+
"step": 4220
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 3.78692927484333,
|
2968 |
+
"grad_norm": 0.1912829428911209,
|
2969 |
+
"learning_rate": 6.361068344034665e-06,
|
2970 |
+
"loss": 0.0101,
|
2971 |
+
"step": 4230
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 3.7958818263205014,
|
2975 |
+
"grad_norm": 0.2168465107679367,
|
2976 |
+
"learning_rate": 6.200606882996846e-06,
|
2977 |
+
"loss": 0.0117,
|
2978 |
+
"step": 4240
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 3.8048343777976723,
|
2982 |
+
"grad_norm": 0.17642278969287872,
|
2983 |
+
"learning_rate": 6.042061349304312e-06,
|
2984 |
+
"loss": 0.0144,
|
2985 |
+
"step": 4250
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 3.813786929274843,
|
2989 |
+
"grad_norm": 0.1212242841720581,
|
2990 |
+
"learning_rate": 5.885438678252342e-06,
|
2991 |
+
"loss": 0.0172,
|
2992 |
+
"step": 4260
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 3.8227394807520145,
|
2996 |
+
"grad_norm": 0.1348852664232254,
|
2997 |
+
"learning_rate": 5.730745721023939e-06,
|
2998 |
+
"loss": 0.0129,
|
2999 |
+
"step": 4270
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 3.8316920322291854,
|
3003 |
+
"grad_norm": 0.15375061333179474,
|
3004 |
+
"learning_rate": 5.577989244390192e-06,
|
3005 |
+
"loss": 0.0115,
|
3006 |
+
"step": 4280
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 3.8406445837063563,
|
3010 |
+
"grad_norm": 0.16176094114780426,
|
3011 |
+
"learning_rate": 5.4271759304142635e-06,
|
3012 |
+
"loss": 0.0112,
|
3013 |
+
"step": 4290
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 3.849597135183527,
|
3017 |
+
"grad_norm": 0.23048561811447144,
|
3018 |
+
"learning_rate": 5.278312376159051e-06,
|
3019 |
+
"loss": 0.0131,
|
3020 |
+
"step": 4300
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 3.858549686660698,
|
3024 |
+
"grad_norm": 0.1969141662120819,
|
3025 |
+
"learning_rate": 5.1314050933986944e-06,
|
3026 |
+
"loss": 0.0109,
|
3027 |
+
"step": 4310
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 3.8675022381378694,
|
3031 |
+
"grad_norm": 0.15142923593521118,
|
3032 |
+
"learning_rate": 4.986460508333634e-06,
|
3033 |
+
"loss": 0.0117,
|
3034 |
+
"step": 4320
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 3.8764547896150403,
|
3038 |
+
"grad_norm": 0.13823889195919037,
|
3039 |
+
"learning_rate": 4.843484961309597e-06,
|
3040 |
+
"loss": 0.0119,
|
3041 |
+
"step": 4330
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 3.885407341092211,
|
3045 |
+
"grad_norm": 0.2543256878852844,
|
3046 |
+
"learning_rate": 4.702484706540161e-06,
|
3047 |
+
"loss": 0.0126,
|
3048 |
+
"step": 4340
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 3.8943598925693825,
|
3052 |
+
"grad_norm": 0.24175454676151276,
|
3053 |
+
"learning_rate": 4.563465911833259e-06,
|
3054 |
+
"loss": 0.0105,
|
3055 |
+
"step": 4350
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 3.9033124440465534,
|
3059 |
+
"grad_norm": 0.22203406691551208,
|
3060 |
+
"learning_rate": 4.426434658321344e-06,
|
3061 |
+
"loss": 0.0112,
|
3062 |
+
"step": 4360
|
3063 |
+
},
|
3064 |
+
{
|
3065 |
+
"epoch": 3.9122649955237243,
|
3066 |
+
"grad_norm": 0.2877592146396637,
|
3067 |
+
"learning_rate": 4.2913969401953466e-06,
|
3068 |
+
"loss": 0.0124,
|
3069 |
+
"step": 4370
|
3070 |
+
},
|
3071 |
+
{
|
3072 |
+
"epoch": 3.921217547000895,
|
3073 |
+
"grad_norm": 0.15440590679645538,
|
3074 |
+
"learning_rate": 4.15835866444253e-06,
|
3075 |
+
"loss": 0.0116,
|
3076 |
+
"step": 4380
|
3077 |
+
},
|
3078 |
+
{
|
3079 |
+
"epoch": 3.930170098478066,
|
3080 |
+
"grad_norm": 0.2942441999912262,
|
3081 |
+
"learning_rate": 4.027325650588043e-06,
|
3082 |
+
"loss": 0.017,
|
3083 |
+
"step": 4390
|
3084 |
+
},
|
3085 |
+
{
|
3086 |
+
"epoch": 3.9391226499552374,
|
3087 |
+
"grad_norm": 0.3283434212207794,
|
3088 |
+
"learning_rate": 3.898303630440419e-06,
|
3089 |
+
"loss": 0.0166,
|
3090 |
+
"step": 4400
|
3091 |
+
},
|
3092 |
+
{
|
3093 |
+
"epoch": 3.9480752014324083,
|
3094 |
+
"grad_norm": 0.12953047454357147,
|
3095 |
+
"learning_rate": 3.7712982478407877e-06,
|
3096 |
+
"loss": 0.0122,
|
3097 |
+
"step": 4410
|
3098 |
+
},
|
3099 |
+
{
|
3100 |
+
"epoch": 3.957027752909579,
|
3101 |
+
"grad_norm": 0.2435809224843979,
|
3102 |
+
"learning_rate": 3.6463150584160053e-06,
|
3103 |
+
"loss": 0.0137,
|
3104 |
+
"step": 4420
|
3105 |
+
},
|
3106 |
+
{
|
3107 |
+
"epoch": 3.96598030438675,
|
3108 |
+
"grad_norm": 0.3056238293647766,
|
3109 |
+
"learning_rate": 3.5233595293356957e-06,
|
3110 |
+
"loss": 0.0131,
|
3111 |
+
"step": 4430
|
3112 |
+
},
|
3113 |
+
{
|
3114 |
+
"epoch": 3.974932855863921,
|
3115 |
+
"grad_norm": 0.2567388117313385,
|
3116 |
+
"learning_rate": 3.4024370390730033e-06,
|
3117 |
+
"loss": 0.0185,
|
3118 |
+
"step": 4440
|
3119 |
+
},
|
3120 |
+
{
|
3121 |
+
"epoch": 3.9838854073410923,
|
3122 |
+
"grad_norm": 0.22484621405601501,
|
3123 |
+
"learning_rate": 3.2835528771693992e-06,
|
3124 |
+
"loss": 0.0119,
|
3125 |
+
"step": 4450
|
3126 |
+
},
|
3127 |
+
{
|
3128 |
+
"epoch": 3.992837958818263,
|
3129 |
+
"grad_norm": 0.14950992166996002,
|
3130 |
+
"learning_rate": 3.1667122440032505e-06,
|
3131 |
+
"loss": 0.0118,
|
3132 |
+
"step": 4460
|
3133 |
+
},
|
3134 |
+
{
|
3135 |
+
"epoch": 4.001790510295434,
|
3136 |
+
"grad_norm": 0.18850022554397583,
|
3137 |
+
"learning_rate": 3.051920250562351e-06,
|
3138 |
+
"loss": 0.0125,
|
3139 |
+
"step": 4470
|
3140 |
+
},
|
3141 |
+
{
|
3142 |
+
"epoch": 4.010743061772605,
|
3143 |
+
"grad_norm": 0.21999414265155792,
|
3144 |
+
"learning_rate": 2.939181918220385e-06,
|
3145 |
+
"loss": 0.0126,
|
3146 |
+
"step": 4480
|
3147 |
+
},
|
3148 |
+
{
|
3149 |
+
"epoch": 4.019695613249776,
|
3150 |
+
"grad_norm": 0.2078053206205368,
|
3151 |
+
"learning_rate": 2.8285021785172226e-06,
|
3152 |
+
"loss": 0.0134,
|
3153 |
+
"step": 4490
|
3154 |
+
},
|
3155 |
+
{
|
3156 |
+
"epoch": 4.028648164726947,
|
3157 |
+
"grad_norm": 0.11813908815383911,
|
3158 |
+
"learning_rate": 2.7198858729432288e-06,
|
3159 |
+
"loss": 0.0135,
|
3160 |
+
"step": 4500
|
3161 |
+
},
|
3162 |
+
{
|
3163 |
+
"epoch": 4.0376007162041185,
|
3164 |
+
"grad_norm": 0.1995633989572525,
|
3165 |
+
"learning_rate": 2.6133377527274905e-06,
|
3166 |
+
"loss": 0.0155,
|
3167 |
+
"step": 4510
|
3168 |
+
},
|
3169 |
+
{
|
3170 |
+
"epoch": 4.046553267681289,
|
3171 |
+
"grad_norm": 0.22360952198505402,
|
3172 |
+
"learning_rate": 2.5088624786299366e-06,
|
3173 |
+
"loss": 0.0136,
|
3174 |
+
"step": 4520
|
3175 |
+
},
|
3176 |
+
{
|
3177 |
+
"epoch": 4.05550581915846,
|
3178 |
+
"grad_norm": 0.1910884976387024,
|
3179 |
+
"learning_rate": 2.406464620737531e-06,
|
3180 |
+
"loss": 0.0141,
|
3181 |
+
"step": 4530
|
3182 |
+
},
|
3183 |
+
{
|
3184 |
+
"epoch": 4.064458370635631,
|
3185 |
+
"grad_norm": 0.16512851417064667,
|
3186 |
+
"learning_rate": 2.3061486582642734e-06,
|
3187 |
+
"loss": 0.0113,
|
3188 |
+
"step": 4540
|
3189 |
+
},
|
3190 |
+
{
|
3191 |
+
"epoch": 4.073410922112802,
|
3192 |
+
"grad_norm": 0.16644251346588135,
|
3193 |
+
"learning_rate": 2.2079189793553667e-06,
|
3194 |
+
"loss": 0.0103,
|
3195 |
+
"step": 4550
|
3196 |
+
},
|
3197 |
+
{
|
3198 |
+
"epoch": 4.082363473589973,
|
3199 |
+
"grad_norm": 0.19846302270889282,
|
3200 |
+
"learning_rate": 2.111779880895165e-06,
|
3201 |
+
"loss": 0.0102,
|
3202 |
+
"step": 4560
|
3203 |
+
},
|
3204 |
+
{
|
3205 |
+
"epoch": 4.091316025067144,
|
3206 |
+
"grad_norm": 0.11550256609916687,
|
3207 |
+
"learning_rate": 2.01773556831929e-06,
|
3208 |
+
"loss": 0.0135,
|
3209 |
+
"step": 4570
|
3210 |
+
},
|
3211 |
+
{
|
3212 |
+
"epoch": 4.100268576544315,
|
3213 |
+
"grad_norm": 0.1841946244239807,
|
3214 |
+
"learning_rate": 1.9257901554306513e-06,
|
3215 |
+
"loss": 0.0104,
|
3216 |
+
"step": 4580
|
3217 |
+
},
|
3218 |
+
{
|
3219 |
+
"epoch": 4.1092211280214865,
|
3220 |
+
"grad_norm": 0.16650405526161194,
|
3221 |
+
"learning_rate": 1.835947664219445e-06,
|
3222 |
+
"loss": 0.0127,
|
3223 |
+
"step": 4590
|
3224 |
+
},
|
3225 |
+
{
|
3226 |
+
"epoch": 4.118173679498657,
|
3227 |
+
"grad_norm": 0.197437584400177,
|
3228 |
+
"learning_rate": 1.748212024687307e-06,
|
3229 |
+
"loss": 0.0122,
|
3230 |
+
"step": 4600
|
3231 |
+
},
|
3232 |
+
{
|
3233 |
+
"epoch": 4.127126230975828,
|
3234 |
+
"grad_norm": 0.08538883924484253,
|
3235 |
+
"learning_rate": 1.6625870746753147e-06,
|
3236 |
+
"loss": 0.0115,
|
3237 |
+
"step": 4610
|
3238 |
+
},
|
3239 |
+
{
|
3240 |
+
"epoch": 4.136078782452999,
|
3241 |
+
"grad_norm": 0.1866847425699234,
|
3242 |
+
"learning_rate": 1.5790765596961853e-06,
|
3243 |
+
"loss": 0.0125,
|
3244 |
+
"step": 4620
|
3245 |
+
},
|
3246 |
+
{
|
3247 |
+
"epoch": 4.14503133393017,
|
3248 |
+
"grad_norm": 0.19738441705703735,
|
3249 |
+
"learning_rate": 1.4976841327703717e-06,
|
3250 |
+
"loss": 0.0137,
|
3251 |
+
"step": 4630
|
3252 |
+
},
|
3253 |
+
{
|
3254 |
+
"epoch": 4.153983885407341,
|
3255 |
+
"grad_norm": 0.19295786321163177,
|
3256 |
+
"learning_rate": 1.4184133542663014e-06,
|
3257 |
+
"loss": 0.0127,
|
3258 |
+
"step": 4640
|
3259 |
+
},
|
3260 |
+
{
|
3261 |
+
"epoch": 4.162936436884512,
|
3262 |
+
"grad_norm": 0.25274568796157837,
|
3263 |
+
"learning_rate": 1.341267691744641e-06,
|
3264 |
+
"loss": 0.0115,
|
3265 |
+
"step": 4650
|
3266 |
+
},
|
3267 |
+
{
|
3268 |
+
"epoch": 4.171888988361683,
|
3269 |
+
"grad_norm": 0.1629364937543869,
|
3270 |
+
"learning_rate": 1.2662505198065666e-06,
|
3271 |
+
"loss": 0.0172,
|
3272 |
+
"step": 4660
|
3273 |
+
},
|
3274 |
+
{
|
3275 |
+
"epoch": 4.180841539838854,
|
3276 |
+
"grad_norm": 0.22567766904830933,
|
3277 |
+
"learning_rate": 1.193365119946216e-06,
|
3278 |
+
"loss": 0.0126,
|
3279 |
+
"step": 4670
|
3280 |
+
},
|
3281 |
+
{
|
3282 |
+
"epoch": 4.189794091316025,
|
3283 |
+
"grad_norm": 0.14044791460037231,
|
3284 |
+
"learning_rate": 1.1226146804070859e-06,
|
3285 |
+
"loss": 0.0084,
|
3286 |
+
"step": 4680
|
3287 |
+
},
|
3288 |
+
{
|
3289 |
+
"epoch": 4.198746642793196,
|
3290 |
+
"grad_norm": 0.10704066604375839,
|
3291 |
+
"learning_rate": 1.0540022960426111e-06,
|
3292 |
+
"loss": 0.015,
|
3293 |
+
"step": 4690
|
3294 |
+
},
|
3295 |
+
{
|
3296 |
+
"epoch": 4.207699194270367,
|
3297 |
+
"grad_norm": 0.17678844928741455,
|
3298 |
+
"learning_rate": 9.875309681807443e-07,
|
3299 |
+
"loss": 0.0165,
|
3300 |
+
"step": 4700
|
3301 |
+
},
|
3302 |
+
{
|
3303 |
+
"epoch": 4.216651745747538,
|
3304 |
+
"grad_norm": 0.2521614730358124,
|
3305 |
+
"learning_rate": 9.232036044927061e-07,
|
3306 |
+
"loss": 0.0116,
|
3307 |
+
"step": 4710
|
3308 |
+
},
|
3309 |
+
{
|
3310 |
+
"epoch": 4.225604297224709,
|
3311 |
+
"grad_norm": 0.1365845501422882,
|
3312 |
+
"learning_rate": 8.610230188657919e-07,
|
3313 |
+
"loss": 0.0136,
|
3314 |
+
"step": 4720
|
3315 |
+
},
|
3316 |
+
{
|
3317 |
+
"epoch": 4.23455684870188,
|
3318 |
+
"grad_norm": 0.17501398921012878,
|
3319 |
+
"learning_rate": 8.009919312802372e-07,
|
3320 |
+
"loss": 0.0127,
|
3321 |
+
"step": 4730
|
3322 |
+
},
|
3323 |
+
{
|
3324 |
+
"epoch": 4.243509400179051,
|
3325 |
+
"grad_norm": 0.1747271865606308,
|
3326 |
+
"learning_rate": 7.431129676902904e-07,
|
3327 |
+
"loss": 0.0109,
|
3328 |
+
"step": 4740
|
3329 |
+
},
|
3330 |
+
{
|
3331 |
+
"epoch": 4.252461951656222,
|
3332 |
+
"grad_norm": 0.31210875511169434,
|
3333 |
+
"learning_rate": 6.873886599093215e-07,
|
3334 |
+
"loss": 0.0143,
|
3335 |
+
"step": 4750
|
3336 |
+
},
|
3337 |
+
{
|
3338 |
+
"epoch": 4.261414503133393,
|
3339 |
+
"grad_norm": 0.1589571237564087,
|
3340 |
+
"learning_rate": 6.338214454990776e-07,
|
3341 |
+
"loss": 0.0115,
|
3342 |
+
"step": 4760
|
3343 |
+
},
|
3344 |
+
{
|
3345 |
+
"epoch": 4.270367054610564,
|
3346 |
+
"grad_norm": 0.1521824449300766,
|
3347 |
+
"learning_rate": 5.82413667663051e-07,
|
3348 |
+
"loss": 0.0122,
|
3349 |
+
"step": 4770
|
3350 |
+
},
|
3351 |
+
{
|
3352 |
+
"epoch": 4.279319606087735,
|
3353 |
+
"grad_norm": 0.20766985416412354,
|
3354 |
+
"learning_rate": 5.331675751439725e-07,
|
3355 |
+
"loss": 0.0145,
|
3356 |
+
"step": 4780
|
3357 |
+
},
|
3358 |
+
{
|
3359 |
+
"epoch": 4.288272157564906,
|
3360 |
+
"grad_norm": 0.2259196788072586,
|
3361 |
+
"learning_rate": 4.86085322125479e-07,
|
3362 |
+
"loss": 0.0112,
|
3363 |
+
"step": 4790
|
3364 |
+
},
|
3365 |
+
{
|
3366 |
+
"epoch": 4.2972247090420765,
|
3367 |
+
"grad_norm": 0.17303107678890228,
|
3368 |
+
"learning_rate": 4.411689681378284e-07,
|
3369 |
+
"loss": 0.0147,
|
3370 |
+
"step": 4800
|
3371 |
+
},
|
3372 |
+
{
|
3373 |
+
"epoch": 4.306177260519248,
|
3374 |
+
"grad_norm": 0.18465794622898102,
|
3375 |
+
"learning_rate": 3.9842047796786466e-07,
|
3376 |
+
"loss": 0.0114,
|
3377 |
+
"step": 4810
|
3378 |
+
},
|
3379 |
+
{
|
3380 |
+
"epoch": 4.315129811996419,
|
3381 |
+
"grad_norm": 0.1356513351202011,
|
3382 |
+
"learning_rate": 3.578417215730323e-07,
|
3383 |
+
"loss": 0.0089,
|
3384 |
+
"step": 4820
|
3385 |
+
},
|
3386 |
+
{
|
3387 |
+
"epoch": 4.32408236347359,
|
3388 |
+
"grad_norm": 0.2126903235912323,
|
3389 |
+
"learning_rate": 3.1943447399958027e-07,
|
3390 |
+
"loss": 0.0143,
|
3391 |
+
"step": 4830
|
3392 |
+
},
|
3393 |
+
{
|
3394 |
+
"epoch": 4.333034914950761,
|
3395 |
+
"grad_norm": 0.23088884353637695,
|
3396 |
+
"learning_rate": 2.8320041530495724e-07,
|
3397 |
+
"loss": 0.0198,
|
3398 |
+
"step": 4840
|
3399 |
+
},
|
3400 |
+
{
|
3401 |
+
"epoch": 4.341987466427932,
|
3402 |
+
"grad_norm": 0.151603102684021,
|
3403 |
+
"learning_rate": 2.491411304842539e-07,
|
3404 |
+
"loss": 0.0143,
|
3405 |
+
"step": 4850
|
3406 |
+
},
|
3407 |
+
{
|
3408 |
+
"epoch": 4.350940017905103,
|
3409 |
+
"grad_norm": 0.16578730940818787,
|
3410 |
+
"learning_rate": 2.1725810940094183e-07,
|
3411 |
+
"loss": 0.0141,
|
3412 |
+
"step": 4860
|
3413 |
+
},
|
3414 |
+
{
|
3415 |
+
"epoch": 4.359892569382274,
|
3416 |
+
"grad_norm": 0.11706045269966125,
|
3417 |
+
"learning_rate": 1.8755274672164202e-07,
|
3418 |
+
"loss": 0.0123,
|
3419 |
+
"step": 4870
|
3420 |
+
},
|
3421 |
+
{
|
3422 |
+
"epoch": 4.3688451208594445,
|
3423 |
+
"grad_norm": 0.21942616999149323,
|
3424 |
+
"learning_rate": 1.600263418551573e-07,
|
3425 |
+
"loss": 0.0163,
|
3426 |
+
"step": 4880
|
3427 |
+
},
|
3428 |
+
{
|
3429 |
+
"epoch": 4.377797672336616,
|
3430 |
+
"grad_norm": 0.163126140832901,
|
3431 |
+
"learning_rate": 1.346800988955954e-07,
|
3432 |
+
"loss": 0.0121,
|
3433 |
+
"step": 4890
|
3434 |
+
},
|
3435 |
+
{
|
3436 |
+
"epoch": 4.386750223813787,
|
3437 |
+
"grad_norm": 0.18442213535308838,
|
3438 |
+
"learning_rate": 1.1151512656975005e-07,
|
3439 |
+
"loss": 0.0119,
|
3440 |
+
"step": 4900
|
3441 |
+
},
|
3442 |
+
{
|
3443 |
+
"epoch": 4.395702775290958,
|
3444 |
+
"grad_norm": 0.1644607037305832,
|
3445 |
+
"learning_rate": 9.053243818853973e-08,
|
3446 |
+
"loss": 0.0125,
|
3447 |
+
"step": 4910
|
3448 |
+
},
|
3449 |
+
{
|
3450 |
+
"epoch": 4.404655326768129,
|
3451 |
+
"grad_norm": 0.1691417694091797,
|
3452 |
+
"learning_rate": 7.173295160273763e-08,
|
3453 |
+
"loss": 0.0089,
|
3454 |
+
"step": 4920
|
3455 |
+
},
|
3456 |
+
{
|
3457 |
+
"epoch": 4.4136078782453,
|
3458 |
+
"grad_norm": 0.19320355355739594,
|
3459 |
+
"learning_rate": 5.511748916279258e-08,
|
3460 |
+
"loss": 0.013,
|
3461 |
+
"step": 4930
|
3462 |
+
},
|
3463 |
+
{
|
3464 |
+
"epoch": 4.422560429722471,
|
3465 |
+
"grad_norm": 0.30131739377975464,
|
3466 |
+
"learning_rate": 4.068677768285234e-08,
|
3467 |
+
"loss": 0.012,
|
3468 |
+
"step": 4940
|
3469 |
+
},
|
3470 |
+
{
|
3471 |
+
"epoch": 4.431512981199642,
|
3472 |
+
"grad_norm": 0.22268839180469513,
|
3473 |
+
"learning_rate": 2.844144840898344e-08,
|
3474 |
+
"loss": 0.0148,
|
3475 |
+
"step": 4950
|
3476 |
+
},
|
3477 |
+
{
|
3478 |
+
"epoch": 4.4404655326768125,
|
3479 |
+
"grad_norm": 0.15642546117305756,
|
3480 |
+
"learning_rate": 1.8382036991559936e-08,
|
3481 |
+
"loss": 0.0089,
|
3482 |
+
"step": 4960
|
3483 |
+
},
|
3484 |
+
{
|
3485 |
+
"epoch": 4.449418084153984,
|
3486 |
+
"grad_norm": 0.14433293044567108,
|
3487 |
+
"learning_rate": 1.0508983461832156e-08,
|
3488 |
+
"loss": 0.0147,
|
3489 |
+
"step": 4970
|
3490 |
+
},
|
3491 |
+
{
|
3492 |
+
"epoch": 4.458370635631155,
|
3493 |
+
"grad_norm": 0.12535181641578674,
|
3494 |
+
"learning_rate": 4.822632212653222e-09,
|
3495 |
+
"loss": 0.0152,
|
3496 |
+
"step": 4980
|
3497 |
+
},
|
3498 |
+
{
|
3499 |
+
"epoch": 4.467323187108326,
|
3500 |
+
"grad_norm": 0.23488545417785645,
|
3501 |
+
"learning_rate": 1.3232319834632912e-09,
|
3502 |
+
"loss": 0.0127,
|
3503 |
+
"step": 4990
|
3504 |
+
},
|
3505 |
+
{
|
3506 |
+
"epoch": 4.476275738585497,
|
3507 |
+
"grad_norm": 0.22346431016921997,
|
3508 |
+
"learning_rate": 1.0935849353854721e-11,
|
3509 |
+
"loss": 0.0127,
|
3510 |
+
"step": 5000
|
3511 |
+
}
|
3512 |
+
],
|
3513 |
+
"logging_steps": 10,
|
3514 |
+
"max_steps": 5000,
|
3515 |
+
"num_input_tokens_seen": 0,
|
3516 |
+
"num_train_epochs": 5,
|
3517 |
+
"save_steps": 1000,
|
3518 |
+
"stateful_callbacks": {
|
3519 |
+
"TrainerControl": {
|
3520 |
+
"args": {
|
3521 |
+
"should_epoch_stop": false,
|
3522 |
+
"should_evaluate": false,
|
3523 |
+
"should_log": false,
|
3524 |
+
"should_save": true,
|
3525 |
+
"should_training_stop": true
|
3526 |
+
},
|
3527 |
+
"attributes": {}
|
3528 |
+
}
|
3529 |
+
},
|
3530 |
+
"total_flos": 0.0,
|
3531 |
+
"train_batch_size": 32,
|
3532 |
+
"trial_name": null,
|
3533 |
+
"trial_params": null
|
3534 |
+
}
|