Upload folder using huggingface_hub
Browse files- README.md +34 -4
- adapter_config.json +2 -2
- adapter_model.safetensors +1 -1
- checkpoint-1737/adapter_config.json +2 -2
- checkpoint-1737/adapter_model.safetensors +1 -1
- checkpoint-1737/optimizer.pt +1 -1
- checkpoint-1737/rng_state.pth +1 -1
- checkpoint-1737/scheduler.pt +1 -1
- checkpoint-1737/trainer_state.json +212 -212
- checkpoint-1737/training_args.bin +1 -1
- checkpoint-3474/adapter_config.json +2 -2
- checkpoint-3474/adapter_model.safetensors +1 -1
- checkpoint-3474/optimizer.pt +1 -1
- checkpoint-3474/rng_state.pth +1 -1
- checkpoint-3474/scheduler.pt +1 -1
- checkpoint-3474/trainer_state.json +430 -430
- checkpoint-3474/training_args.bin +1 -1
- checkpoint-5211/adapter_config.json +2 -2
- checkpoint-5211/adapter_model.safetensors +1 -1
- checkpoint-5211/optimizer.pt +1 -1
- checkpoint-5211/rng_state.pth +1 -1
- checkpoint-5211/scheduler.pt +1 -1
- checkpoint-5211/trainer_state.json +648 -648
- checkpoint-5211/training_args.bin +1 -1
- checkpoint-6948/adapter_config.json +2 -2
- checkpoint-6948/adapter_model.safetensors +1 -1
- checkpoint-6948/optimizer.pt +1 -1
- checkpoint-6948/rng_state.pth +1 -1
- checkpoint-6948/scheduler.pt +1 -1
- checkpoint-6948/trainer_state.json +861 -861
- checkpoint-6948/training_args.bin +1 -1
- checkpoint-8685/adapter_config.json +3 -3
- checkpoint-8685/adapter_model.safetensors +2 -2
- checkpoint-8685/optimizer.pt +2 -2
- checkpoint-8685/rng_state.pth +1 -1
- checkpoint-8685/scheduler.pt +1 -1
- checkpoint-8685/trainer_state.json +1079 -1079
- checkpoint-8685/training_args.bin +1 -1
- runs/Dec04_11-47-13_129-213-84-8/events.out.tfevents.1764848895.129-213-84-8.25442.0 +3 -0
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -1,9 +1,39 @@
|
|
| 1 |
---
|
| 2 |
base_model: Qwen/Qwen2-VL-2B-Instruct
|
| 3 |
library_name: peft
|
|
|
|
| 4 |
tags:
|
| 5 |
-
-
|
| 6 |
-
-
|
| 7 |
-
-
|
| 8 |
-
-
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
base_model: Qwen/Qwen2-VL-2B-Instruct
|
| 3 |
library_name: peft
|
| 4 |
+
model_name: output
|
| 5 |
tags:
|
| 6 |
+
- adapter
|
| 7 |
+
- lora
|
| 8 |
+
- sft
|
| 9 |
+
- transformers
|
| 10 |
+
- trl
|
| 11 |
+
license: apache-2.0
|
| 12 |
+
pipeline_tag: text-generation
|
| 13 |
---
|
| 14 |
+
|
| 15 |
+
# Model Card for output
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct).
|
| 18 |
+
It has been trained using [TRL](https://github.com/huggingface/trl) and PEFT LoRA.
|
| 19 |
+
|
| 20 |
+
## Quick start
|
| 21 |
+
|
| 22 |
+
```python
|
| 23 |
+
from transformers import pipeline
|
| 24 |
+
|
| 25 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 26 |
+
|
| 27 |
+
generator = pipeline(
|
| 28 |
+
"text-generation",
|
| 29 |
+
model="RetrO21/agrofinetune", # replace with your repo
|
| 30 |
+
device="cuda"
|
| 31 |
+
)
|
| 32 |
+
|
| 33 |
+
output = generator(
|
| 34 |
+
[{"role": "user", "content": question}],
|
| 35 |
+
max_new_tokens=128,
|
| 36 |
+
return_full_text=False
|
| 37 |
+
)[0]
|
| 38 |
+
|
| 39 |
+
print(output["generated_text"])
|
adapter_config.json
CHANGED
|
@@ -29,10 +29,10 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
-
"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26182176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8b4ecb107db701acdc04f96300149f10454a4f22cc800cab0b968eae74c3415
|
| 3 |
size 26182176
|
checkpoint-1737/adapter_config.json
CHANGED
|
@@ -29,10 +29,10 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
-
"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
checkpoint-1737/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26182176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3f567a0aff94e611cf7b87a63d4290b0ee7314a1941c1ad3d0f416afc2fb1f7
|
| 3 |
size 26182176
|
checkpoint-1737/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 52486155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b51eac51c311fe05ae0fb6f075e636ed719de49501d8a85daa67a1976ddb3f2
|
| 3 |
size 52486155
|
checkpoint-1737/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09c74ac7a46536c3808a08a6b9cc111c17592bccdc148cd2300d105708f4cc8b
|
| 3 |
size 14645
|
checkpoint-1737/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ce062a9f59b08a36604b136fb249f7d9f4c575b16c5cc4c39f6833a49683785
|
| 3 |
size 1465
|
checkpoint-1737/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 1737,
|
| 3 |
-
"best_metric": 5.
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-1737",
|
| 5 |
"epoch": 1.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,362 +10,362 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"entropy": 3.
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
-
"grad_norm": 3.
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
-
"loss": 13.
|
| 18 |
-
"mean_token_accuracy": 0.
|
| 19 |
-
"num_tokens":
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"entropy": 3.
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
-
"grad_norm": 3.
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
-
"loss": 14.
|
| 28 |
-
"mean_token_accuracy": 0.
|
| 29 |
-
"num_tokens":
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"entropy": 3.
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
-
"grad_norm": 3.
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
-
"loss":
|
| 38 |
-
"mean_token_accuracy": 0.
|
| 39 |
-
"num_tokens":
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"entropy": 3.
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
-
"grad_norm": 4.
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
-
"loss": 13.
|
| 48 |
-
"mean_token_accuracy": 0.
|
| 49 |
-
"num_tokens":
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"entropy": 3.
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
-
"grad_norm": 4.
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
-
"mean_token_accuracy": 0.
|
| 59 |
-
"num_tokens":
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"entropy":
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
-
"grad_norm": 4.
|
| 66 |
-
"learning_rate": 1.
|
| 67 |
-
"loss": 12.
|
| 68 |
-
"mean_token_accuracy": 0.
|
| 69 |
-
"num_tokens":
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
-
"entropy": 4.
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate": 1.
|
| 77 |
-
"loss": 12.
|
| 78 |
-
"mean_token_accuracy": 0.
|
| 79 |
-
"num_tokens":
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"entropy": 4.
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 1.
|
| 87 |
-
"loss": 11.
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"entropy":
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate": 1.
|
| 97 |
-
"loss":
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"entropy": 5.
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 1.
|
| 107 |
-
"loss": 9.
|
| 108 |
-
"mean_token_accuracy": 0.
|
| 109 |
-
"num_tokens":
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"entropy": 6.
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate": 1.
|
| 117 |
-
"loss":
|
| 118 |
-
"mean_token_accuracy": 0.
|
| 119 |
-
"num_tokens":
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"entropy": 6.
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 1.
|
| 127 |
-
"loss": 7.
|
| 128 |
-
"mean_token_accuracy": 0.
|
| 129 |
-
"num_tokens":
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"entropy": 6.
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 1.
|
| 137 |
-
"loss": 6.
|
| 138 |
-
"mean_token_accuracy": 0.
|
| 139 |
-
"num_tokens":
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"entropy": 6.
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
-
"grad_norm": 4.
|
| 146 |
-
"learning_rate": 1.
|
| 147 |
-
"loss": 6.
|
| 148 |
-
"mean_token_accuracy": 0.
|
| 149 |
-
"num_tokens":
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"entropy": 6.
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 6.
|
| 158 |
-
"mean_token_accuracy": 0.
|
| 159 |
-
"num_tokens":
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"entropy": 6.
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
-
"grad_norm":
|
| 166 |
-
"learning_rate": 1.
|
| 167 |
-
"loss": 6.
|
| 168 |
-
"mean_token_accuracy": 0.
|
| 169 |
-
"num_tokens":
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"entropy": 6.
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
-
"grad_norm": 3.
|
| 176 |
-
"learning_rate": 1.
|
| 177 |
-
"loss": 6.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"entropy": 6.
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 1.
|
| 187 |
-
"loss": 6.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"entropy": 6.
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
-
"grad_norm": 2.
|
| 196 |
-
"learning_rate": 1.
|
| 197 |
-
"loss": 6.
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"entropy": 6.
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
-
"grad_norm": 2.
|
| 206 |
-
"learning_rate": 1.
|
| 207 |
-
"loss": 6.
|
| 208 |
-
"mean_token_accuracy": 0.
|
| 209 |
-
"num_tokens":
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"entropy": 6.
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
-
"grad_norm": 2.
|
| 216 |
-
"learning_rate": 1.
|
| 217 |
-
"loss": 6.
|
| 218 |
-
"mean_token_accuracy": 0.
|
| 219 |
-
"num_tokens":
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"entropy": 6.
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 1.
|
| 227 |
-
"loss": 5.
|
| 228 |
-
"mean_token_accuracy": 0.
|
| 229 |
-
"num_tokens":
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"entropy": 5.
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate": 1.
|
| 237 |
-
"loss": 5.
|
| 238 |
-
"mean_token_accuracy": 0.
|
| 239 |
-
"num_tokens":
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"entropy": 6.
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 1.
|
| 247 |
-
"loss": 5.
|
| 248 |
-
"mean_token_accuracy": 0.
|
| 249 |
-
"num_tokens":
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"entropy": 6.
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
-
"grad_norm": 3.
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 5.
|
| 258 |
-
"mean_token_accuracy": 0.
|
| 259 |
-
"num_tokens":
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"entropy": 6.
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
-
"grad_norm": 2.
|
| 266 |
-
"learning_rate": 1.
|
| 267 |
-
"loss": 5.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"entropy": 5.
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
-
"grad_norm": 2.
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 5.
|
| 278 |
-
"mean_token_accuracy": 0.
|
| 279 |
-
"num_tokens":
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"entropy": 6.
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
-
"grad_norm": 3.
|
| 286 |
-
"learning_rate": 1.
|
| 287 |
-
"loss": 5.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"entropy":
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 1.
|
| 297 |
-
"loss": 5.
|
| 298 |
-
"mean_token_accuracy": 0.
|
| 299 |
-
"num_tokens":
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"entropy":
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 5.
|
| 308 |
-
"mean_token_accuracy": 0.
|
| 309 |
-
"num_tokens":
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"entropy":
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 1.
|
| 317 |
-
"loss": 5.
|
| 318 |
-
"mean_token_accuracy": 0.
|
| 319 |
-
"num_tokens":
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"entropy": 5.
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
-
"grad_norm": 2.
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 5.
|
| 328 |
-
"mean_token_accuracy": 0.
|
| 329 |
-
"num_tokens":
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"entropy":
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
-
"grad_norm": 2.
|
| 336 |
-
"learning_rate": 1.
|
| 337 |
-
"loss": 5.
|
| 338 |
-
"mean_token_accuracy": 0.
|
| 339 |
-
"num_tokens":
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"entropy": 5.
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 1.
|
| 347 |
-
"loss": 5.
|
| 348 |
-
"mean_token_accuracy": 0.
|
| 349 |
-
"num_tokens":
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
-
"eval_entropy":
|
| 355 |
-
"eval_loss": 5.
|
| 356 |
-
"eval_mean_token_accuracy": 0.
|
| 357 |
-
"eval_model_preparation_time": 0.
|
| 358 |
-
"eval_num_tokens":
|
| 359 |
-
"eval_runtime":
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second":
|
| 362 |
"step": 1737
|
| 363 |
}
|
| 364 |
],
|
| 365 |
"logging_steps": 50,
|
| 366 |
-
"max_steps":
|
| 367 |
"num_input_tokens_seen": 0,
|
| 368 |
-
"num_train_epochs":
|
| 369 |
"save_steps": 500,
|
| 370 |
"stateful_callbacks": {
|
| 371 |
"TrainerControl": {
|
|
@@ -379,7 +379,7 @@
|
|
| 379 |
"attributes": {}
|
| 380 |
}
|
| 381 |
},
|
| 382 |
-
"total_flos": 2.
|
| 383 |
"train_batch_size": 2,
|
| 384 |
"trial_name": null,
|
| 385 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 1737,
|
| 3 |
+
"best_metric": 5.737204551696777,
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-1737",
|
| 5 |
"epoch": 1.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"entropy": 3.606692385673523,
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
+
"grad_norm": 3.2999913692474365,
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
+
"loss": 13.6598,
|
| 18 |
+
"mean_token_accuracy": 0.16028020828962325,
|
| 19 |
+
"num_tokens": 53993.0,
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"entropy": 3.618675880432129,
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
+
"grad_norm": 3.101252555847168,
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
+
"loss": 14.0188,
|
| 28 |
+
"mean_token_accuracy": 0.1508466500043869,
|
| 29 |
+
"num_tokens": 110134.0,
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"entropy": 3.5215235900878907,
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
+
"grad_norm": 3.513662815093994,
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
+
"loss": 12.8555,
|
| 38 |
+
"mean_token_accuracy": 0.18527640983462335,
|
| 39 |
+
"num_tokens": 160191.0,
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"entropy": 3.667909698486328,
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
+
"grad_norm": 4.327610492706299,
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
+
"loss": 13.5394,
|
| 48 |
+
"mean_token_accuracy": 0.157139780074358,
|
| 49 |
+
"num_tokens": 214993.0,
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"entropy": 3.768263258934021,
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
+
"grad_norm": 4.290107250213623,
|
| 56 |
+
"learning_rate": 1.988450206246317e-06,
|
| 57 |
+
"loss": 12.8912,
|
| 58 |
+
"mean_token_accuracy": 0.17374794125556947,
|
| 59 |
+
"num_tokens": 268184.0,
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"entropy": 3.990619196891785,
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
+
"grad_norm": 4.444278717041016,
|
| 66 |
+
"learning_rate": 1.976664702416028e-06,
|
| 67 |
+
"loss": 12.455,
|
| 68 |
+
"mean_token_accuracy": 0.17780130118131637,
|
| 69 |
+
"num_tokens": 319458.0,
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"entropy": 4.162646284103394,
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
+
"grad_norm": 5.615262508392334,
|
| 76 |
+
"learning_rate": 1.9648791985857395e-06,
|
| 77 |
+
"loss": 12.0893,
|
| 78 |
+
"mean_token_accuracy": 0.18191319867968558,
|
| 79 |
+
"num_tokens": 373337.0,
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"entropy": 4.532100868225098,
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
+
"grad_norm": 10.074016571044922,
|
| 86 |
+
"learning_rate": 1.9530936947554507e-06,
|
| 87 |
+
"loss": 11.9261,
|
| 88 |
+
"mean_token_accuracy": 0.169477596282959,
|
| 89 |
+
"num_tokens": 427526.0,
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"entropy": 4.923871030807495,
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
+
"grad_norm": 16.220163345336914,
|
| 96 |
+
"learning_rate": 1.9413081909251622e-06,
|
| 97 |
+
"loss": 11.0048,
|
| 98 |
+
"mean_token_accuracy": 0.1704501649737358,
|
| 99 |
+
"num_tokens": 480528.0,
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"entropy": 5.521005854606629,
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
+
"grad_norm": 29.904008865356445,
|
| 106 |
+
"learning_rate": 1.9295226870948733e-06,
|
| 107 |
+
"loss": 9.6524,
|
| 108 |
+
"mean_token_accuracy": 0.16450899541378022,
|
| 109 |
+
"num_tokens": 535314.0,
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"entropy": 6.092623329162597,
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
+
"grad_norm": 17.821575164794922,
|
| 116 |
+
"learning_rate": 1.9177371832645845e-06,
|
| 117 |
+
"loss": 8.1054,
|
| 118 |
+
"mean_token_accuracy": 0.17205011785030366,
|
| 119 |
+
"num_tokens": 588410.0,
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"entropy": 6.385262680053711,
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
+
"grad_norm": 5.502202987670898,
|
| 126 |
+
"learning_rate": 1.9059516794342958e-06,
|
| 127 |
+
"loss": 7.4313,
|
| 128 |
+
"mean_token_accuracy": 0.1734227080643177,
|
| 129 |
+
"num_tokens": 641736.0,
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"entropy": 6.278562617301941,
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
+
"grad_norm": 5.4657697677612305,
|
| 136 |
+
"learning_rate": 1.8941661756040071e-06,
|
| 137 |
+
"loss": 6.9266,
|
| 138 |
+
"mean_token_accuracy": 0.18680249139666558,
|
| 139 |
+
"num_tokens": 692200.0,
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"entropy": 6.553266277313233,
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
+
"grad_norm": 4.955812931060791,
|
| 146 |
+
"learning_rate": 1.8823806717737183e-06,
|
| 147 |
+
"loss": 6.9847,
|
| 148 |
+
"mean_token_accuracy": 0.16679802387952805,
|
| 149 |
+
"num_tokens": 745830.0,
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"entropy": 6.470935583114624,
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
+
"grad_norm": 4.198381423950195,
|
| 156 |
+
"learning_rate": 1.8705951679434296e-06,
|
| 157 |
+
"loss": 6.7277,
|
| 158 |
+
"mean_token_accuracy": 0.17847734570503235,
|
| 159 |
+
"num_tokens": 798872.0,
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"entropy": 6.5620588779449465,
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
+
"grad_norm": 3.1793746948242188,
|
| 166 |
+
"learning_rate": 1.8588096641131407e-06,
|
| 167 |
+
"loss": 6.7032,
|
| 168 |
+
"mean_token_accuracy": 0.17336134731769562,
|
| 169 |
+
"num_tokens": 853045.0,
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"entropy": 6.532204885482788,
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
+
"grad_norm": 3.824537515640259,
|
| 176 |
+
"learning_rate": 1.847024160282852e-06,
|
| 177 |
+
"loss": 6.5762,
|
| 178 |
+
"mean_token_accuracy": 0.1805124071240425,
|
| 179 |
+
"num_tokens": 907679.0,
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"entropy": 6.535988225936889,
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
+
"grad_norm": 4.350001811981201,
|
| 186 |
+
"learning_rate": 1.8352386564525632e-06,
|
| 187 |
+
"loss": 6.505,
|
| 188 |
+
"mean_token_accuracy": 0.1842605724930763,
|
| 189 |
+
"num_tokens": 964170.0,
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"entropy": 6.204533562660218,
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
+
"grad_norm": 2.193660020828247,
|
| 196 |
+
"learning_rate": 1.8234531526222745e-06,
|
| 197 |
+
"loss": 6.1211,
|
| 198 |
+
"mean_token_accuracy": 0.21968430042266845,
|
| 199 |
+
"num_tokens": 1015909.0,
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"entropy": 6.308737449645996,
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
+
"grad_norm": 2.325622320175171,
|
| 206 |
+
"learning_rate": 1.8116676487919857e-06,
|
| 207 |
+
"loss": 6.1653,
|
| 208 |
+
"mean_token_accuracy": 0.21636426240205764,
|
| 209 |
+
"num_tokens": 1068859.0,
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"entropy": 6.332560749053955,
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
+
"grad_norm": 2.0439090728759766,
|
| 216 |
+
"learning_rate": 1.799882144961697e-06,
|
| 217 |
+
"loss": 6.1559,
|
| 218 |
+
"mean_token_accuracy": 0.21859725564718246,
|
| 219 |
+
"num_tokens": 1123202.0,
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"entropy": 6.042124252319336,
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
+
"grad_norm": 3.621903657913208,
|
| 226 |
+
"learning_rate": 1.7880966411314081e-06,
|
| 227 |
+
"loss": 5.8441,
|
| 228 |
+
"mean_token_accuracy": 0.24906315237283708,
|
| 229 |
+
"num_tokens": 1173403.0,
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"entropy": 5.921343173980713,
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
+
"grad_norm": 5.658033847808838,
|
| 236 |
+
"learning_rate": 1.7763111373011195e-06,
|
| 237 |
+
"loss": 5.7104,
|
| 238 |
+
"mean_token_accuracy": 0.2625067520141602,
|
| 239 |
+
"num_tokens": 1225026.0,
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"entropy": 6.093586492538452,
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
+
"grad_norm": 2.4292995929718018,
|
| 246 |
+
"learning_rate": 1.7645256334708308e-06,
|
| 247 |
+
"loss": 5.8658,
|
| 248 |
+
"mean_token_accuracy": 0.24842385441064835,
|
| 249 |
+
"num_tokens": 1279013.0,
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"entropy": 6.119112596511841,
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
+
"grad_norm": 3.369384288787842,
|
| 256 |
+
"learning_rate": 1.752740129640542e-06,
|
| 257 |
+
"loss": 5.8784,
|
| 258 |
+
"mean_token_accuracy": 0.24857850253582,
|
| 259 |
+
"num_tokens": 1332547.0,
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"entropy": 6.025163550376892,
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
+
"grad_norm": 2.5110116004943848,
|
| 266 |
+
"learning_rate": 1.7409546258102533e-06,
|
| 267 |
+
"loss": 5.7769,
|
| 268 |
+
"mean_token_accuracy": 0.25835376888513567,
|
| 269 |
+
"num_tokens": 1385192.0,
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"entropy": 5.877259612083435,
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
+
"grad_norm": 2.4179303646087646,
|
| 276 |
+
"learning_rate": 1.7291691219799646e-06,
|
| 277 |
+
"loss": 5.6284,
|
| 278 |
+
"mean_token_accuracy": 0.2756252554059029,
|
| 279 |
+
"num_tokens": 1437071.0,
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"entropy": 6.002246947288513,
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
+
"grad_norm": 3.494359016418457,
|
| 286 |
+
"learning_rate": 1.717383618149676e-06,
|
| 287 |
+
"loss": 5.747,
|
| 288 |
+
"mean_token_accuracy": 0.26462210685014725,
|
| 289 |
+
"num_tokens": 1490818.0,
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"entropy": 5.991955623626709,
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
+
"grad_norm": 2.340975761413574,
|
| 296 |
+
"learning_rate": 1.705598114319387e-06,
|
| 297 |
+
"loss": 5.7379,
|
| 298 |
+
"mean_token_accuracy": 0.26444981098175047,
|
| 299 |
+
"num_tokens": 1544997.0,
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"entropy": 5.91768889427185,
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
+
"grad_norm": 2.2394514083862305,
|
| 306 |
+
"learning_rate": 1.6938126104890984e-06,
|
| 307 |
+
"loss": 5.6564,
|
| 308 |
+
"mean_token_accuracy": 0.2730415526032448,
|
| 309 |
+
"num_tokens": 1598302.0,
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"entropy": 5.982716989517212,
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
+
"grad_norm": 1.876839518547058,
|
| 316 |
+
"learning_rate": 1.6820271066588098e-06,
|
| 317 |
+
"loss": 5.7215,
|
| 318 |
+
"mean_token_accuracy": 0.26642445534467696,
|
| 319 |
+
"num_tokens": 1655267.0,
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"entropy": 5.820467872619629,
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
+
"grad_norm": 2.219966173171997,
|
| 326 |
+
"learning_rate": 1.6702416028285209e-06,
|
| 327 |
+
"loss": 5.5555,
|
| 328 |
+
"mean_token_accuracy": 0.2856418335437775,
|
| 329 |
+
"num_tokens": 1709199.0,
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"entropy": 5.996349005699158,
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
+
"grad_norm": 2.247213840484619,
|
| 336 |
+
"learning_rate": 1.6584560989982322e-06,
|
| 337 |
+
"loss": 5.7283,
|
| 338 |
+
"mean_token_accuracy": 0.2696125540137291,
|
| 339 |
+
"num_tokens": 1765443.0,
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"entropy": 5.696683068275451,
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
+
"grad_norm": 2.8499979972839355,
|
| 346 |
+
"learning_rate": 1.6466705951679433e-06,
|
| 347 |
+
"loss": 5.4335,
|
| 348 |
+
"mean_token_accuracy": 0.29918427973985673,
|
| 349 |
+
"num_tokens": 1817494.0,
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
+
"eval_entropy": 5.993559589034401,
|
| 355 |
+
"eval_loss": 5.737204551696777,
|
| 356 |
+
"eval_mean_token_accuracy": 0.2618687468739699,
|
| 357 |
+
"eval_model_preparation_time": 0.0045,
|
| 358 |
+
"eval_num_tokens": 1856362.0,
|
| 359 |
+
"eval_runtime": 50.5332,
|
| 360 |
+
"eval_samples_per_second": 8.588,
|
| 361 |
+
"eval_steps_per_second": 4.294,
|
| 362 |
"step": 1737
|
| 363 |
}
|
| 364 |
],
|
| 365 |
"logging_steps": 50,
|
| 366 |
+
"max_steps": 8685,
|
| 367 |
"num_input_tokens_seen": 0,
|
| 368 |
+
"num_train_epochs": 5,
|
| 369 |
"save_steps": 500,
|
| 370 |
"stateful_callbacks": {
|
| 371 |
"TrainerControl": {
|
|
|
|
| 379 |
"attributes": {}
|
| 380 |
}
|
| 381 |
},
|
| 382 |
+
"total_flos": 2.546183130710016e+16,
|
| 383 |
"train_batch_size": 2,
|
| 384 |
"trial_name": null,
|
| 385 |
"trial_params": null
|
checkpoint-1737/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|
checkpoint-3474/adapter_config.json
CHANGED
|
@@ -29,10 +29,10 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
-
"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
checkpoint-3474/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26182176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df1785d18603767be800e1f55b15fb6ca91bfb92d13b41c606ed125306990ad0
|
| 3 |
size 26182176
|
checkpoint-3474/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 52486155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:351f85f1592228a689e3684e2c19ad9a864153045233b07024283e9a19837ffc
|
| 3 |
size 52486155
|
checkpoint-3474/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62a6bc0e3b9744793642655fb166aa4f5a9fb6952bc69941d500e104fd082ebd
|
| 3 |
size 14645
|
checkpoint-3474/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c06cf7ae7d08ea484734082c3c87adfab434e5715a578ee4ab7e0ffcbea54c0b
|
| 3 |
size 1465
|
checkpoint-3474/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3474,
|
| 3 |
-
"best_metric": 5.
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-3474",
|
| 5 |
"epoch": 2.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,724 +10,724 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"entropy": 3.
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
-
"grad_norm": 3.
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
-
"loss": 13.
|
| 18 |
-
"mean_token_accuracy": 0.
|
| 19 |
-
"num_tokens":
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"entropy": 3.
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
-
"grad_norm": 3.
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
-
"loss": 14.
|
| 28 |
-
"mean_token_accuracy": 0.
|
| 29 |
-
"num_tokens":
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"entropy": 3.
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
-
"grad_norm": 3.
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
-
"loss":
|
| 38 |
-
"mean_token_accuracy": 0.
|
| 39 |
-
"num_tokens":
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"entropy": 3.
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
-
"grad_norm": 4.
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
-
"loss": 13.
|
| 48 |
-
"mean_token_accuracy": 0.
|
| 49 |
-
"num_tokens":
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"entropy": 3.
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
-
"grad_norm": 4.
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
-
"mean_token_accuracy": 0.
|
| 59 |
-
"num_tokens":
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"entropy":
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
-
"grad_norm": 4.
|
| 66 |
-
"learning_rate": 1.
|
| 67 |
-
"loss": 12.
|
| 68 |
-
"mean_token_accuracy": 0.
|
| 69 |
-
"num_tokens":
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
-
"entropy": 4.
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate": 1.
|
| 77 |
-
"loss": 12.
|
| 78 |
-
"mean_token_accuracy": 0.
|
| 79 |
-
"num_tokens":
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"entropy": 4.
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 1.
|
| 87 |
-
"loss": 11.
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"entropy":
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate": 1.
|
| 97 |
-
"loss":
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"entropy": 5.
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 1.
|
| 107 |
-
"loss": 9.
|
| 108 |
-
"mean_token_accuracy": 0.
|
| 109 |
-
"num_tokens":
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"entropy": 6.
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate": 1.
|
| 117 |
-
"loss":
|
| 118 |
-
"mean_token_accuracy": 0.
|
| 119 |
-
"num_tokens":
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"entropy": 6.
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 1.
|
| 127 |
-
"loss": 7.
|
| 128 |
-
"mean_token_accuracy": 0.
|
| 129 |
-
"num_tokens":
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"entropy": 6.
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 1.
|
| 137 |
-
"loss": 6.
|
| 138 |
-
"mean_token_accuracy": 0.
|
| 139 |
-
"num_tokens":
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"entropy": 6.
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
-
"grad_norm": 4.
|
| 146 |
-
"learning_rate": 1.
|
| 147 |
-
"loss": 6.
|
| 148 |
-
"mean_token_accuracy": 0.
|
| 149 |
-
"num_tokens":
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"entropy": 6.
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 6.
|
| 158 |
-
"mean_token_accuracy": 0.
|
| 159 |
-
"num_tokens":
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"entropy": 6.
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
-
"grad_norm":
|
| 166 |
-
"learning_rate": 1.
|
| 167 |
-
"loss": 6.
|
| 168 |
-
"mean_token_accuracy": 0.
|
| 169 |
-
"num_tokens":
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"entropy": 6.
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
-
"grad_norm": 3.
|
| 176 |
-
"learning_rate": 1.
|
| 177 |
-
"loss": 6.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"entropy": 6.
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 1.
|
| 187 |
-
"loss": 6.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"entropy": 6.
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
-
"grad_norm": 2.
|
| 196 |
-
"learning_rate": 1.
|
| 197 |
-
"loss": 6.
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"entropy": 6.
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
-
"grad_norm": 2.
|
| 206 |
-
"learning_rate": 1.
|
| 207 |
-
"loss": 6.
|
| 208 |
-
"mean_token_accuracy": 0.
|
| 209 |
-
"num_tokens":
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"entropy": 6.
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
-
"grad_norm": 2.
|
| 216 |
-
"learning_rate": 1.
|
| 217 |
-
"loss": 6.
|
| 218 |
-
"mean_token_accuracy": 0.
|
| 219 |
-
"num_tokens":
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"entropy": 6.
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 1.
|
| 227 |
-
"loss": 5.
|
| 228 |
-
"mean_token_accuracy": 0.
|
| 229 |
-
"num_tokens":
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"entropy": 5.
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate": 1.
|
| 237 |
-
"loss": 5.
|
| 238 |
-
"mean_token_accuracy": 0.
|
| 239 |
-
"num_tokens":
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"entropy": 6.
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 1.
|
| 247 |
-
"loss": 5.
|
| 248 |
-
"mean_token_accuracy": 0.
|
| 249 |
-
"num_tokens":
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"entropy": 6.
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
-
"grad_norm": 3.
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 5.
|
| 258 |
-
"mean_token_accuracy": 0.
|
| 259 |
-
"num_tokens":
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"entropy": 6.
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
-
"grad_norm": 2.
|
| 266 |
-
"learning_rate": 1.
|
| 267 |
-
"loss": 5.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"entropy": 5.
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
-
"grad_norm": 2.
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 5.
|
| 278 |
-
"mean_token_accuracy": 0.
|
| 279 |
-
"num_tokens":
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"entropy": 6.
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
-
"grad_norm": 3.
|
| 286 |
-
"learning_rate": 1.
|
| 287 |
-
"loss": 5.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"entropy":
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 1.
|
| 297 |
-
"loss": 5.
|
| 298 |
-
"mean_token_accuracy": 0.
|
| 299 |
-
"num_tokens":
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"entropy":
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 5.
|
| 308 |
-
"mean_token_accuracy": 0.
|
| 309 |
-
"num_tokens":
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"entropy":
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 1.
|
| 317 |
-
"loss": 5.
|
| 318 |
-
"mean_token_accuracy": 0.
|
| 319 |
-
"num_tokens":
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"entropy": 5.
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
-
"grad_norm": 2.
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 5.
|
| 328 |
-
"mean_token_accuracy": 0.
|
| 329 |
-
"num_tokens":
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"entropy":
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
-
"grad_norm": 2.
|
| 336 |
-
"learning_rate": 1.
|
| 337 |
-
"loss": 5.
|
| 338 |
-
"mean_token_accuracy": 0.
|
| 339 |
-
"num_tokens":
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"entropy": 5.
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 1.
|
| 347 |
-
"loss": 5.
|
| 348 |
-
"mean_token_accuracy": 0.
|
| 349 |
-
"num_tokens":
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
-
"eval_entropy":
|
| 355 |
-
"eval_loss": 5.
|
| 356 |
-
"eval_mean_token_accuracy": 0.
|
| 357 |
-
"eval_model_preparation_time": 0.
|
| 358 |
-
"eval_num_tokens":
|
| 359 |
-
"eval_runtime":
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second":
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"entropy": 5.
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
-
"grad_norm": 2.
|
| 368 |
-
"learning_rate": 1.
|
| 369 |
-
"loss": 5.
|
| 370 |
-
"mean_token_accuracy": 0.
|
| 371 |
-
"num_tokens":
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"entropy":
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 1.
|
| 379 |
-
"loss": 5.
|
| 380 |
-
"mean_token_accuracy": 0.
|
| 381 |
-
"num_tokens":
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"entropy": 5.
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 1.
|
| 389 |
-
"loss": 5.
|
| 390 |
-
"mean_token_accuracy": 0.
|
| 391 |
-
"num_tokens":
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"entropy": 5.
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
-
"grad_norm":
|
| 398 |
-
"learning_rate": 1.
|
| 399 |
-
"loss": 5.
|
| 400 |
-
"mean_token_accuracy": 0.
|
| 401 |
-
"num_tokens":
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"entropy": 5.
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 1.
|
| 409 |
-
"loss": 5.
|
| 410 |
-
"mean_token_accuracy": 0.
|
| 411 |
-
"num_tokens":
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"entropy": 5.
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
-
"grad_norm":
|
| 418 |
-
"learning_rate": 1.
|
| 419 |
-
"loss": 5.
|
| 420 |
-
"mean_token_accuracy": 0.
|
| 421 |
-
"num_tokens":
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"entropy": 5.
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 1.
|
| 429 |
-
"loss": 5.
|
| 430 |
-
"mean_token_accuracy": 0.
|
| 431 |
-
"num_tokens":
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"entropy": 5.
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss": 5.
|
| 440 |
-
"mean_token_accuracy": 0.
|
| 441 |
-
"num_tokens":
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
-
"entropy": 5.
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
-
"grad_norm": 2.
|
| 448 |
-
"learning_rate": 1.
|
| 449 |
-
"loss": 5.
|
| 450 |
-
"mean_token_accuracy": 0.
|
| 451 |
-
"num_tokens":
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"entropy": 5.
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
-
"grad_norm":
|
| 458 |
-
"learning_rate": 1.
|
| 459 |
-
"loss": 5.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
-
"entropy": 5.
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
-
"grad_norm":
|
| 468 |
-
"learning_rate": 1.
|
| 469 |
-
"loss": 5.
|
| 470 |
-
"mean_token_accuracy": 0.
|
| 471 |
-
"num_tokens":
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"entropy": 5.
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 1.
|
| 479 |
-
"loss": 5.
|
| 480 |
-
"mean_token_accuracy": 0.
|
| 481 |
-
"num_tokens":
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"entropy": 5.
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
-
"grad_norm":
|
| 488 |
-
"learning_rate": 1.
|
| 489 |
-
"loss": 5.
|
| 490 |
-
"mean_token_accuracy": 0.
|
| 491 |
-
"num_tokens":
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"entropy": 5.
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 1.
|
| 499 |
-
"loss": 5.
|
| 500 |
-
"mean_token_accuracy": 0.
|
| 501 |
-
"num_tokens":
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"entropy": 5.
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate": 1.
|
| 509 |
-
"loss": 5.
|
| 510 |
-
"mean_token_accuracy": 0.
|
| 511 |
-
"num_tokens":
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"entropy": 5.
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
-
"grad_norm": 1.
|
| 518 |
-
"learning_rate": 1.
|
| 519 |
-
"loss": 5.
|
| 520 |
-
"mean_token_accuracy": 0.
|
| 521 |
-
"num_tokens":
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"entropy": 5.
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
-
"grad_norm": 3.
|
| 528 |
-
"learning_rate": 1.
|
| 529 |
-
"loss": 5.
|
| 530 |
-
"mean_token_accuracy": 0.
|
| 531 |
-
"num_tokens":
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"entropy": 5.
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
-
"grad_norm": 1.
|
| 538 |
-
"learning_rate": 1.
|
| 539 |
-
"loss": 5.
|
| 540 |
-
"mean_token_accuracy": 0.
|
| 541 |
-
"num_tokens":
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"entropy": 5.
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
-
"grad_norm":
|
| 548 |
-
"learning_rate": 1.
|
| 549 |
-
"loss": 5.
|
| 550 |
-
"mean_token_accuracy": 0.
|
| 551 |
-
"num_tokens":
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"entropy": 5.
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
-
"grad_norm":
|
| 558 |
-
"learning_rate": 1.
|
| 559 |
-
"loss": 5.
|
| 560 |
-
"mean_token_accuracy": 0.
|
| 561 |
-
"num_tokens":
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"entropy": 5.
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
-
"grad_norm": 2.
|
| 568 |
-
"learning_rate": 1.
|
| 569 |
-
"loss": 5.
|
| 570 |
-
"mean_token_accuracy": 0.
|
| 571 |
-
"num_tokens":
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"entropy": 5.
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate": 1.
|
| 579 |
-
"loss": 5.
|
| 580 |
-
"mean_token_accuracy": 0.
|
| 581 |
-
"num_tokens":
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"entropy": 5.
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
-
"grad_norm": 1.
|
| 588 |
-
"learning_rate": 1.
|
| 589 |
-
"loss": 5.
|
| 590 |
-
"mean_token_accuracy": 0.
|
| 591 |
-
"num_tokens":
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
-
"entropy": 5.
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
-
"grad_norm": 1.
|
| 598 |
-
"learning_rate": 1.
|
| 599 |
-
"loss": 5.
|
| 600 |
-
"mean_token_accuracy": 0.
|
| 601 |
-
"num_tokens":
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"entropy": 5.
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
-
"grad_norm": 1.
|
| 608 |
-
"learning_rate": 1.
|
| 609 |
-
"loss": 5.
|
| 610 |
-
"mean_token_accuracy": 0.
|
| 611 |
-
"num_tokens":
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"entropy": 5.
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
-
"grad_norm":
|
| 618 |
-
"learning_rate": 1.
|
| 619 |
-
"loss": 5.
|
| 620 |
-
"mean_token_accuracy": 0.
|
| 621 |
-
"num_tokens":
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
-
"entropy": 5.
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
-
"grad_norm": 1.
|
| 628 |
-
"learning_rate": 1.
|
| 629 |
-
"loss": 5.
|
| 630 |
-
"mean_token_accuracy": 0.
|
| 631 |
-
"num_tokens":
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"entropy": 5.
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate": 1.
|
| 639 |
-
"loss": 5.
|
| 640 |
-
"mean_token_accuracy": 0.
|
| 641 |
-
"num_tokens":
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"entropy": 5.
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate": 1.
|
| 649 |
-
"loss":
|
| 650 |
-
"mean_token_accuracy": 0.
|
| 651 |
-
"num_tokens":
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
-
"entropy": 5.
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
-
"grad_norm": 1.
|
| 658 |
-
"learning_rate": 1.
|
| 659 |
-
"loss": 5.
|
| 660 |
-
"mean_token_accuracy": 0.
|
| 661 |
-
"num_tokens":
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"entropy": 5.
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
-
"grad_norm": 1.
|
| 668 |
-
"learning_rate": 1.
|
| 669 |
-
"loss": 5.
|
| 670 |
-
"mean_token_accuracy": 0.
|
| 671 |
-
"num_tokens":
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"entropy": 5.
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
-
"grad_norm":
|
| 678 |
-
"learning_rate": 1.
|
| 679 |
-
"loss": 5.
|
| 680 |
-
"mean_token_accuracy": 0.
|
| 681 |
-
"num_tokens":
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"entropy": 5.
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
-
"grad_norm":
|
| 688 |
-
"learning_rate": 1.
|
| 689 |
-
"loss": 5.
|
| 690 |
-
"mean_token_accuracy": 0.
|
| 691 |
-
"num_tokens":
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"entropy": 5.
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
-
"grad_norm": 1.
|
| 698 |
-
"learning_rate": 1.
|
| 699 |
-
"loss": 5.
|
| 700 |
-
"mean_token_accuracy": 0.
|
| 701 |
-
"num_tokens":
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"entropy": 5.
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
-
"grad_norm": 1.
|
| 708 |
-
"learning_rate": 1.
|
| 709 |
-
"loss": 5.
|
| 710 |
-
"mean_token_accuracy": 0.
|
| 711 |
-
"num_tokens":
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
-
"eval_entropy": 5.
|
| 717 |
-
"eval_loss": 5.
|
| 718 |
-
"eval_mean_token_accuracy": 0.
|
| 719 |
-
"eval_model_preparation_time": 0.
|
| 720 |
-
"eval_num_tokens":
|
| 721 |
-
"eval_runtime":
|
| 722 |
-
"eval_samples_per_second":
|
| 723 |
-
"eval_steps_per_second":
|
| 724 |
"step": 3474
|
| 725 |
}
|
| 726 |
],
|
| 727 |
"logging_steps": 50,
|
| 728 |
-
"max_steps":
|
| 729 |
"num_input_tokens_seen": 0,
|
| 730 |
-
"num_train_epochs":
|
| 731 |
"save_steps": 500,
|
| 732 |
"stateful_callbacks": {
|
| 733 |
"TrainerControl": {
|
|
@@ -741,7 +741,7 @@
|
|
| 741 |
"attributes": {}
|
| 742 |
}
|
| 743 |
},
|
| 744 |
-
"total_flos": 5.
|
| 745 |
"train_batch_size": 2,
|
| 746 |
"trial_name": null,
|
| 747 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3474,
|
| 3 |
+
"best_metric": 5.55628776550293,
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-3474",
|
| 5 |
"epoch": 2.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"entropy": 3.606692385673523,
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
+
"grad_norm": 3.2999913692474365,
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
+
"loss": 13.6598,
|
| 18 |
+
"mean_token_accuracy": 0.16028020828962325,
|
| 19 |
+
"num_tokens": 53993.0,
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"entropy": 3.618675880432129,
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
+
"grad_norm": 3.101252555847168,
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
+
"loss": 14.0188,
|
| 28 |
+
"mean_token_accuracy": 0.1508466500043869,
|
| 29 |
+
"num_tokens": 110134.0,
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"entropy": 3.5215235900878907,
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
+
"grad_norm": 3.513662815093994,
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
+
"loss": 12.8555,
|
| 38 |
+
"mean_token_accuracy": 0.18527640983462335,
|
| 39 |
+
"num_tokens": 160191.0,
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"entropy": 3.667909698486328,
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
+
"grad_norm": 4.327610492706299,
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
+
"loss": 13.5394,
|
| 48 |
+
"mean_token_accuracy": 0.157139780074358,
|
| 49 |
+
"num_tokens": 214993.0,
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"entropy": 3.768263258934021,
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
+
"grad_norm": 4.290107250213623,
|
| 56 |
+
"learning_rate": 1.988450206246317e-06,
|
| 57 |
+
"loss": 12.8912,
|
| 58 |
+
"mean_token_accuracy": 0.17374794125556947,
|
| 59 |
+
"num_tokens": 268184.0,
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"entropy": 3.990619196891785,
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
+
"grad_norm": 4.444278717041016,
|
| 66 |
+
"learning_rate": 1.976664702416028e-06,
|
| 67 |
+
"loss": 12.455,
|
| 68 |
+
"mean_token_accuracy": 0.17780130118131637,
|
| 69 |
+
"num_tokens": 319458.0,
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"entropy": 4.162646284103394,
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
+
"grad_norm": 5.615262508392334,
|
| 76 |
+
"learning_rate": 1.9648791985857395e-06,
|
| 77 |
+
"loss": 12.0893,
|
| 78 |
+
"mean_token_accuracy": 0.18191319867968558,
|
| 79 |
+
"num_tokens": 373337.0,
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"entropy": 4.532100868225098,
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
+
"grad_norm": 10.074016571044922,
|
| 86 |
+
"learning_rate": 1.9530936947554507e-06,
|
| 87 |
+
"loss": 11.9261,
|
| 88 |
+
"mean_token_accuracy": 0.169477596282959,
|
| 89 |
+
"num_tokens": 427526.0,
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"entropy": 4.923871030807495,
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
+
"grad_norm": 16.220163345336914,
|
| 96 |
+
"learning_rate": 1.9413081909251622e-06,
|
| 97 |
+
"loss": 11.0048,
|
| 98 |
+
"mean_token_accuracy": 0.1704501649737358,
|
| 99 |
+
"num_tokens": 480528.0,
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"entropy": 5.521005854606629,
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
+
"grad_norm": 29.904008865356445,
|
| 106 |
+
"learning_rate": 1.9295226870948733e-06,
|
| 107 |
+
"loss": 9.6524,
|
| 108 |
+
"mean_token_accuracy": 0.16450899541378022,
|
| 109 |
+
"num_tokens": 535314.0,
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"entropy": 6.092623329162597,
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
+
"grad_norm": 17.821575164794922,
|
| 116 |
+
"learning_rate": 1.9177371832645845e-06,
|
| 117 |
+
"loss": 8.1054,
|
| 118 |
+
"mean_token_accuracy": 0.17205011785030366,
|
| 119 |
+
"num_tokens": 588410.0,
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"entropy": 6.385262680053711,
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
+
"grad_norm": 5.502202987670898,
|
| 126 |
+
"learning_rate": 1.9059516794342958e-06,
|
| 127 |
+
"loss": 7.4313,
|
| 128 |
+
"mean_token_accuracy": 0.1734227080643177,
|
| 129 |
+
"num_tokens": 641736.0,
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"entropy": 6.278562617301941,
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
+
"grad_norm": 5.4657697677612305,
|
| 136 |
+
"learning_rate": 1.8941661756040071e-06,
|
| 137 |
+
"loss": 6.9266,
|
| 138 |
+
"mean_token_accuracy": 0.18680249139666558,
|
| 139 |
+
"num_tokens": 692200.0,
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"entropy": 6.553266277313233,
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
+
"grad_norm": 4.955812931060791,
|
| 146 |
+
"learning_rate": 1.8823806717737183e-06,
|
| 147 |
+
"loss": 6.9847,
|
| 148 |
+
"mean_token_accuracy": 0.16679802387952805,
|
| 149 |
+
"num_tokens": 745830.0,
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"entropy": 6.470935583114624,
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
+
"grad_norm": 4.198381423950195,
|
| 156 |
+
"learning_rate": 1.8705951679434296e-06,
|
| 157 |
+
"loss": 6.7277,
|
| 158 |
+
"mean_token_accuracy": 0.17847734570503235,
|
| 159 |
+
"num_tokens": 798872.0,
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"entropy": 6.5620588779449465,
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
+
"grad_norm": 3.1793746948242188,
|
| 166 |
+
"learning_rate": 1.8588096641131407e-06,
|
| 167 |
+
"loss": 6.7032,
|
| 168 |
+
"mean_token_accuracy": 0.17336134731769562,
|
| 169 |
+
"num_tokens": 853045.0,
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"entropy": 6.532204885482788,
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
+
"grad_norm": 3.824537515640259,
|
| 176 |
+
"learning_rate": 1.847024160282852e-06,
|
| 177 |
+
"loss": 6.5762,
|
| 178 |
+
"mean_token_accuracy": 0.1805124071240425,
|
| 179 |
+
"num_tokens": 907679.0,
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"entropy": 6.535988225936889,
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
+
"grad_norm": 4.350001811981201,
|
| 186 |
+
"learning_rate": 1.8352386564525632e-06,
|
| 187 |
+
"loss": 6.505,
|
| 188 |
+
"mean_token_accuracy": 0.1842605724930763,
|
| 189 |
+
"num_tokens": 964170.0,
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"entropy": 6.204533562660218,
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
+
"grad_norm": 2.193660020828247,
|
| 196 |
+
"learning_rate": 1.8234531526222745e-06,
|
| 197 |
+
"loss": 6.1211,
|
| 198 |
+
"mean_token_accuracy": 0.21968430042266845,
|
| 199 |
+
"num_tokens": 1015909.0,
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"entropy": 6.308737449645996,
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
+
"grad_norm": 2.325622320175171,
|
| 206 |
+
"learning_rate": 1.8116676487919857e-06,
|
| 207 |
+
"loss": 6.1653,
|
| 208 |
+
"mean_token_accuracy": 0.21636426240205764,
|
| 209 |
+
"num_tokens": 1068859.0,
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"entropy": 6.332560749053955,
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
+
"grad_norm": 2.0439090728759766,
|
| 216 |
+
"learning_rate": 1.799882144961697e-06,
|
| 217 |
+
"loss": 6.1559,
|
| 218 |
+
"mean_token_accuracy": 0.21859725564718246,
|
| 219 |
+
"num_tokens": 1123202.0,
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"entropy": 6.042124252319336,
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
+
"grad_norm": 3.621903657913208,
|
| 226 |
+
"learning_rate": 1.7880966411314081e-06,
|
| 227 |
+
"loss": 5.8441,
|
| 228 |
+
"mean_token_accuracy": 0.24906315237283708,
|
| 229 |
+
"num_tokens": 1173403.0,
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"entropy": 5.921343173980713,
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
+
"grad_norm": 5.658033847808838,
|
| 236 |
+
"learning_rate": 1.7763111373011195e-06,
|
| 237 |
+
"loss": 5.7104,
|
| 238 |
+
"mean_token_accuracy": 0.2625067520141602,
|
| 239 |
+
"num_tokens": 1225026.0,
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"entropy": 6.093586492538452,
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
+
"grad_norm": 2.4292995929718018,
|
| 246 |
+
"learning_rate": 1.7645256334708308e-06,
|
| 247 |
+
"loss": 5.8658,
|
| 248 |
+
"mean_token_accuracy": 0.24842385441064835,
|
| 249 |
+
"num_tokens": 1279013.0,
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"entropy": 6.119112596511841,
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
+
"grad_norm": 3.369384288787842,
|
| 256 |
+
"learning_rate": 1.752740129640542e-06,
|
| 257 |
+
"loss": 5.8784,
|
| 258 |
+
"mean_token_accuracy": 0.24857850253582,
|
| 259 |
+
"num_tokens": 1332547.0,
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"entropy": 6.025163550376892,
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
+
"grad_norm": 2.5110116004943848,
|
| 266 |
+
"learning_rate": 1.7409546258102533e-06,
|
| 267 |
+
"loss": 5.7769,
|
| 268 |
+
"mean_token_accuracy": 0.25835376888513567,
|
| 269 |
+
"num_tokens": 1385192.0,
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"entropy": 5.877259612083435,
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
+
"grad_norm": 2.4179303646087646,
|
| 276 |
+
"learning_rate": 1.7291691219799646e-06,
|
| 277 |
+
"loss": 5.6284,
|
| 278 |
+
"mean_token_accuracy": 0.2756252554059029,
|
| 279 |
+
"num_tokens": 1437071.0,
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"entropy": 6.002246947288513,
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
+
"grad_norm": 3.494359016418457,
|
| 286 |
+
"learning_rate": 1.717383618149676e-06,
|
| 287 |
+
"loss": 5.747,
|
| 288 |
+
"mean_token_accuracy": 0.26462210685014725,
|
| 289 |
+
"num_tokens": 1490818.0,
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"entropy": 5.991955623626709,
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
+
"grad_norm": 2.340975761413574,
|
| 296 |
+
"learning_rate": 1.705598114319387e-06,
|
| 297 |
+
"loss": 5.7379,
|
| 298 |
+
"mean_token_accuracy": 0.26444981098175047,
|
| 299 |
+
"num_tokens": 1544997.0,
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"entropy": 5.91768889427185,
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
+
"grad_norm": 2.2394514083862305,
|
| 306 |
+
"learning_rate": 1.6938126104890984e-06,
|
| 307 |
+
"loss": 5.6564,
|
| 308 |
+
"mean_token_accuracy": 0.2730415526032448,
|
| 309 |
+
"num_tokens": 1598302.0,
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"entropy": 5.982716989517212,
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
+
"grad_norm": 1.876839518547058,
|
| 316 |
+
"learning_rate": 1.6820271066588098e-06,
|
| 317 |
+
"loss": 5.7215,
|
| 318 |
+
"mean_token_accuracy": 0.26642445534467696,
|
| 319 |
+
"num_tokens": 1655267.0,
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"entropy": 5.820467872619629,
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
+
"grad_norm": 2.219966173171997,
|
| 326 |
+
"learning_rate": 1.6702416028285209e-06,
|
| 327 |
+
"loss": 5.5555,
|
| 328 |
+
"mean_token_accuracy": 0.2856418335437775,
|
| 329 |
+
"num_tokens": 1709199.0,
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"entropy": 5.996349005699158,
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
+
"grad_norm": 2.247213840484619,
|
| 336 |
+
"learning_rate": 1.6584560989982322e-06,
|
| 337 |
+
"loss": 5.7283,
|
| 338 |
+
"mean_token_accuracy": 0.2696125540137291,
|
| 339 |
+
"num_tokens": 1765443.0,
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"entropy": 5.696683068275451,
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
+
"grad_norm": 2.8499979972839355,
|
| 346 |
+
"learning_rate": 1.6466705951679433e-06,
|
| 347 |
+
"loss": 5.4335,
|
| 348 |
+
"mean_token_accuracy": 0.29918427973985673,
|
| 349 |
+
"num_tokens": 1817494.0,
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
+
"eval_entropy": 5.993559589034401,
|
| 355 |
+
"eval_loss": 5.737204551696777,
|
| 356 |
+
"eval_mean_token_accuracy": 0.2618687468739699,
|
| 357 |
+
"eval_model_preparation_time": 0.0045,
|
| 358 |
+
"eval_num_tokens": 1856362.0,
|
| 359 |
+
"eval_runtime": 50.5332,
|
| 360 |
+
"eval_samples_per_second": 8.588,
|
| 361 |
+
"eval_steps_per_second": 4.294,
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"entropy": 5.746842083930969,
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
+
"grad_norm": 2.33052921295166,
|
| 368 |
+
"learning_rate": 1.6348850913376547e-06,
|
| 369 |
+
"loss": 5.4796,
|
| 370 |
+
"mean_token_accuracy": 0.2966849410533905,
|
| 371 |
+
"num_tokens": 1870353.0,
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"entropy": 5.859029049873352,
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
+
"grad_norm": 1.6248886585235596,
|
| 378 |
+
"learning_rate": 1.6230995875073658e-06,
|
| 379 |
+
"loss": 5.5975,
|
| 380 |
+
"mean_token_accuracy": 0.2838129925727844,
|
| 381 |
+
"num_tokens": 1926205.0,
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"entropy": 5.731445336341858,
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
+
"grad_norm": 1.6941566467285156,
|
| 388 |
+
"learning_rate": 1.6113140836770771e-06,
|
| 389 |
+
"loss": 5.476,
|
| 390 |
+
"mean_token_accuracy": 0.2992346465587616,
|
| 391 |
+
"num_tokens": 1979821.0,
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"entropy": 5.6993954515457155,
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
+
"grad_norm": 1.1746597290039062,
|
| 398 |
+
"learning_rate": 1.5995285798467883e-06,
|
| 399 |
+
"loss": 5.4608,
|
| 400 |
+
"mean_token_accuracy": 0.3000726142525673,
|
| 401 |
+
"num_tokens": 2034373.0,
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"entropy": 5.668873124122619,
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
+
"grad_norm": 1.728211760520935,
|
| 408 |
+
"learning_rate": 1.5877430760164996e-06,
|
| 409 |
+
"loss": 5.4347,
|
| 410 |
+
"mean_token_accuracy": 0.3033922725915909,
|
| 411 |
+
"num_tokens": 2087339.0,
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
+
"entropy": 5.624621086120605,
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
+
"grad_norm": 1.4078539609909058,
|
| 418 |
+
"learning_rate": 1.5759575721862107e-06,
|
| 419 |
+
"loss": 5.3954,
|
| 420 |
+
"mean_token_accuracy": 0.30784171640872954,
|
| 421 |
+
"num_tokens": 2139520.0,
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"entropy": 5.7141213130950925,
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
+
"grad_norm": 2.186459541320801,
|
| 428 |
+
"learning_rate": 1.564172068355922e-06,
|
| 429 |
+
"loss": 5.4847,
|
| 430 |
+
"mean_token_accuracy": 0.29594049394130706,
|
| 431 |
+
"num_tokens": 2193987.0,
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"entropy": 5.632415266036987,
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
+
"grad_norm": 1.3601349592208862,
|
| 438 |
+
"learning_rate": 1.5523865645256334e-06,
|
| 439 |
+
"loss": 5.4135,
|
| 440 |
+
"mean_token_accuracy": 0.30366597563028336,
|
| 441 |
+
"num_tokens": 2249616.0,
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
+
"entropy": 5.510904269218445,
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
+
"grad_norm": 2.065760612487793,
|
| 448 |
+
"learning_rate": 1.5406010606953445e-06,
|
| 449 |
+
"loss": 5.2904,
|
| 450 |
+
"mean_token_accuracy": 0.3211754837632179,
|
| 451 |
+
"num_tokens": 2300863.0,
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"entropy": 5.703383626937867,
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
+
"grad_norm": 1.1172698736190796,
|
| 458 |
+
"learning_rate": 1.5288155568650559e-06,
|
| 459 |
+
"loss": 5.4802,
|
| 460 |
+
"mean_token_accuracy": 0.29713701367378237,
|
| 461 |
+
"num_tokens": 2356029.0,
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
+
"entropy": 5.565930342674255,
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
+
"grad_norm": 1.7528513669967651,
|
| 468 |
+
"learning_rate": 1.5170300530347672e-06,
|
| 469 |
+
"loss": 5.3518,
|
| 470 |
+
"mean_token_accuracy": 0.31301232606172563,
|
| 471 |
+
"num_tokens": 2408957.0,
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"entropy": 5.496430187225342,
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
+
"grad_norm": 1.892640233039856,
|
| 478 |
+
"learning_rate": 1.5052445492044786e-06,
|
| 479 |
+
"loss": 5.2967,
|
| 480 |
+
"mean_token_accuracy": 0.3181899458169937,
|
| 481 |
+
"num_tokens": 2462569.0,
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"entropy": 5.725150098800659,
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
+
"grad_norm": 1.774940848350525,
|
| 488 |
+
"learning_rate": 1.4934590453741897e-06,
|
| 489 |
+
"loss": 5.5215,
|
| 490 |
+
"mean_token_accuracy": 0.29055028676986694,
|
| 491 |
+
"num_tokens": 2518544.0,
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"entropy": 5.4884827613830565,
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
+
"grad_norm": 2.2167599201202393,
|
| 498 |
+
"learning_rate": 1.481673541543901e-06,
|
| 499 |
+
"loss": 5.2917,
|
| 500 |
+
"mean_token_accuracy": 0.31803421139717103,
|
| 501 |
+
"num_tokens": 2570863.0,
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"entropy": 5.697079472541809,
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
+
"grad_norm": 1.6489030122756958,
|
| 508 |
+
"learning_rate": 1.4698880377136124e-06,
|
| 509 |
+
"loss": 5.4982,
|
| 510 |
+
"mean_token_accuracy": 0.2925163987278938,
|
| 511 |
+
"num_tokens": 2626998.0,
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"entropy": 5.46209939956665,
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
+
"grad_norm": 1.153914451599121,
|
| 518 |
+
"learning_rate": 1.4581025338833235e-06,
|
| 519 |
+
"loss": 5.2736,
|
| 520 |
+
"mean_token_accuracy": 0.3182168474793434,
|
| 521 |
+
"num_tokens": 2681568.0,
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"entropy": 5.4405768728256225,
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
+
"grad_norm": 3.6614978313446045,
|
| 528 |
+
"learning_rate": 1.4463170300530348e-06,
|
| 529 |
+
"loss": 5.2515,
|
| 530 |
+
"mean_token_accuracy": 0.3218736210465431,
|
| 531 |
+
"num_tokens": 2733587.0,
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"entropy": 5.528175053596496,
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
+
"grad_norm": 1.0849746465682983,
|
| 538 |
+
"learning_rate": 1.434531526222746e-06,
|
| 539 |
+
"loss": 5.3378,
|
| 540 |
+
"mean_token_accuracy": 0.31061659604310987,
|
| 541 |
+
"num_tokens": 2787003.0,
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"entropy": 5.46110897064209,
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
+
"grad_norm": 1.8315683603286743,
|
| 548 |
+
"learning_rate": 1.4227460223924573e-06,
|
| 549 |
+
"loss": 5.2782,
|
| 550 |
+
"mean_token_accuracy": 0.31781029611825945,
|
| 551 |
+
"num_tokens": 2840263.0,
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"entropy": 5.455560960769653,
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
+
"grad_norm": 1.1859091520309448,
|
| 558 |
+
"learning_rate": 1.4109605185621684e-06,
|
| 559 |
+
"loss": 5.2735,
|
| 560 |
+
"mean_token_accuracy": 0.3194814011454582,
|
| 561 |
+
"num_tokens": 2894186.0,
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"entropy": 5.430496115684509,
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
+
"grad_norm": 2.3500001430511475,
|
| 568 |
+
"learning_rate": 1.3991750147318797e-06,
|
| 569 |
+
"loss": 5.2464,
|
| 570 |
+
"mean_token_accuracy": 0.32140792965888976,
|
| 571 |
+
"num_tokens": 2948171.0,
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"entropy": 5.588023023605347,
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
+
"grad_norm": 1.727825403213501,
|
| 578 |
+
"learning_rate": 1.3873895109015909e-06,
|
| 579 |
+
"loss": 5.4028,
|
| 580 |
+
"mean_token_accuracy": 0.3039530631899834,
|
| 581 |
+
"num_tokens": 3002678.0,
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"entropy": 5.410525422096253,
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
+
"grad_norm": 1.3401474952697754,
|
| 588 |
+
"learning_rate": 1.3756040070713022e-06,
|
| 589 |
+
"loss": 5.2298,
|
| 590 |
+
"mean_token_accuracy": 0.324065263569355,
|
| 591 |
+
"num_tokens": 3055844.0,
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"entropy": 5.36959942817688,
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
+
"grad_norm": 1.1892589330673218,
|
| 598 |
+
"learning_rate": 1.3638185032410133e-06,
|
| 599 |
+
"loss": 5.1956,
|
| 600 |
+
"mean_token_accuracy": 0.32639502108097074,
|
| 601 |
+
"num_tokens": 3108636.0,
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"entropy": 5.53826907157898,
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
+
"grad_norm": 1.2652360200881958,
|
| 608 |
+
"learning_rate": 1.3520329994107247e-06,
|
| 609 |
+
"loss": 5.3583,
|
| 610 |
+
"mean_token_accuracy": 0.3074926760792732,
|
| 611 |
+
"num_tokens": 3162627.0,
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"entropy": 5.417449145317078,
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
+
"grad_norm": 1.584312915802002,
|
| 618 |
+
"learning_rate": 1.340247495580436e-06,
|
| 619 |
+
"loss": 5.2388,
|
| 620 |
+
"mean_token_accuracy": 0.32019727885723115,
|
| 621 |
+
"num_tokens": 3216409.0,
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"entropy": 5.241390740871429,
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
+
"grad_norm": 1.5219439268112183,
|
| 628 |
+
"learning_rate": 1.3284619917501471e-06,
|
| 629 |
+
"loss": 5.0645,
|
| 630 |
+
"mean_token_accuracy": 0.3445430138707161,
|
| 631 |
+
"num_tokens": 3266967.0,
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"entropy": 5.405424036979675,
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
+
"grad_norm": 2.1165153980255127,
|
| 638 |
+
"learning_rate": 1.3166764879198585e-06,
|
| 639 |
+
"loss": 5.232,
|
| 640 |
+
"mean_token_accuracy": 0.32085000157356264,
|
| 641 |
+
"num_tokens": 3319877.0,
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"entropy": 5.123006024360657,
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
+
"grad_norm": 1.2189785242080688,
|
| 648 |
+
"learning_rate": 1.3048909840895698e-06,
|
| 649 |
+
"loss": 4.9582,
|
| 650 |
+
"mean_token_accuracy": 0.356108532845974,
|
| 651 |
+
"num_tokens": 3368569.0,
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"entropy": 5.417610831260681,
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
+
"grad_norm": 1.5157604217529297,
|
| 658 |
+
"learning_rate": 1.2931054802592812e-06,
|
| 659 |
+
"loss": 5.2454,
|
| 660 |
+
"mean_token_accuracy": 0.31976755023002623,
|
| 661 |
+
"num_tokens": 3422449.0,
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"entropy": 5.409690895080566,
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
+
"grad_norm": 1.3088161945343018,
|
| 668 |
+
"learning_rate": 1.2813199764289923e-06,
|
| 669 |
+
"loss": 5.2348,
|
| 670 |
+
"mean_token_accuracy": 0.32325415283441544,
|
| 671 |
+
"num_tokens": 3474399.0,
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"entropy": 5.44662567615509,
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
+
"grad_norm": 2.178372621536255,
|
| 678 |
+
"learning_rate": 1.2695344725987036e-06,
|
| 679 |
+
"loss": 5.2661,
|
| 680 |
+
"mean_token_accuracy": 0.3182847076654434,
|
| 681 |
+
"num_tokens": 3527726.0,
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"entropy": 5.512614865303039,
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
+
"grad_norm": 1.3050425052642822,
|
| 688 |
+
"learning_rate": 1.2577489687684147e-06,
|
| 689 |
+
"loss": 5.3416,
|
| 690 |
+
"mean_token_accuracy": 0.3084403133392334,
|
| 691 |
+
"num_tokens": 3581980.0,
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"entropy": 5.379772834777832,
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
+
"grad_norm": 1.4584404230117798,
|
| 698 |
+
"learning_rate": 1.245963464938126e-06,
|
| 699 |
+
"loss": 5.2087,
|
| 700 |
+
"mean_token_accuracy": 0.32388432770967485,
|
| 701 |
+
"num_tokens": 3635393.0,
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"entropy": 5.483665924072266,
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
+
"grad_norm": 1.2157734632492065,
|
| 708 |
+
"learning_rate": 1.2341779611078374e-06,
|
| 709 |
+
"loss": 5.3101,
|
| 710 |
+
"mean_token_accuracy": 0.3121953472495079,
|
| 711 |
+
"num_tokens": 3689894.0,
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
+
"eval_entropy": 5.711394641805904,
|
| 717 |
+
"eval_loss": 5.55628776550293,
|
| 718 |
+
"eval_mean_token_accuracy": 0.2764948787777105,
|
| 719 |
+
"eval_model_preparation_time": 0.0045,
|
| 720 |
+
"eval_num_tokens": 3712724.0,
|
| 721 |
+
"eval_runtime": 50.187,
|
| 722 |
+
"eval_samples_per_second": 8.648,
|
| 723 |
+
"eval_steps_per_second": 4.324,
|
| 724 |
"step": 3474
|
| 725 |
}
|
| 726 |
],
|
| 727 |
"logging_steps": 50,
|
| 728 |
+
"max_steps": 8685,
|
| 729 |
"num_input_tokens_seen": 0,
|
| 730 |
+
"num_train_epochs": 5,
|
| 731 |
"save_steps": 500,
|
| 732 |
"stateful_callbacks": {
|
| 733 |
"TrainerControl": {
|
|
|
|
| 741 |
"attributes": {}
|
| 742 |
}
|
| 743 |
},
|
| 744 |
+
"total_flos": 5.088598592372736e+16,
|
| 745 |
"train_batch_size": 2,
|
| 746 |
"trial_name": null,
|
| 747 |
"trial_params": null
|
checkpoint-3474/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|
checkpoint-5211/adapter_config.json
CHANGED
|
@@ -29,10 +29,10 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
-
"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
checkpoint-5211/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26182176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e4025b4926f5afc8ae1b9483e03aa404961dda3a1814cb46ca7aeea065c3fe0b
|
| 3 |
size 26182176
|
checkpoint-5211/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 52486155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1c6ba0e1a0fa5232307b4180a95f626f91ecd955d6b67f2be41b26598f1360f
|
| 3 |
size 52486155
|
checkpoint-5211/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39b9f29ddfbd9f77ba8789dff06c3a159107fa04d3db99c9007a614a55b3b852
|
| 3 |
size 14645
|
checkpoint-5211/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e174b3c503a209f4286065ce76cb198c717027077a7d229160809df776f0167
|
| 3 |
size 1465
|
checkpoint-5211/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 5211,
|
| 3 |
-
"best_metric": 5.
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-5211",
|
| 5 |
"epoch": 3.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,1086 +10,1086 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"entropy": 3.
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
-
"grad_norm": 3.
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
-
"loss": 13.
|
| 18 |
-
"mean_token_accuracy": 0.
|
| 19 |
-
"num_tokens":
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"entropy": 3.
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
-
"grad_norm": 3.
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
-
"loss": 14.
|
| 28 |
-
"mean_token_accuracy": 0.
|
| 29 |
-
"num_tokens":
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"entropy": 3.
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
-
"grad_norm": 3.
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
-
"loss":
|
| 38 |
-
"mean_token_accuracy": 0.
|
| 39 |
-
"num_tokens":
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"entropy": 3.
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
-
"grad_norm": 4.
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
-
"loss": 13.
|
| 48 |
-
"mean_token_accuracy": 0.
|
| 49 |
-
"num_tokens":
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"entropy": 3.
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
-
"grad_norm": 4.
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
-
"mean_token_accuracy": 0.
|
| 59 |
-
"num_tokens":
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"entropy":
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
-
"grad_norm": 4.
|
| 66 |
-
"learning_rate": 1.
|
| 67 |
-
"loss": 12.
|
| 68 |
-
"mean_token_accuracy": 0.
|
| 69 |
-
"num_tokens":
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
-
"entropy": 4.
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate": 1.
|
| 77 |
-
"loss": 12.
|
| 78 |
-
"mean_token_accuracy": 0.
|
| 79 |
-
"num_tokens":
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"entropy": 4.
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 1.
|
| 87 |
-
"loss": 11.
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"entropy":
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate": 1.
|
| 97 |
-
"loss":
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"entropy": 5.
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 1.
|
| 107 |
-
"loss": 9.
|
| 108 |
-
"mean_token_accuracy": 0.
|
| 109 |
-
"num_tokens":
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"entropy": 6.
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate": 1.
|
| 117 |
-
"loss":
|
| 118 |
-
"mean_token_accuracy": 0.
|
| 119 |
-
"num_tokens":
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"entropy": 6.
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 1.
|
| 127 |
-
"loss": 7.
|
| 128 |
-
"mean_token_accuracy": 0.
|
| 129 |
-
"num_tokens":
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"entropy": 6.
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 1.
|
| 137 |
-
"loss": 6.
|
| 138 |
-
"mean_token_accuracy": 0.
|
| 139 |
-
"num_tokens":
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"entropy": 6.
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
-
"grad_norm": 4.
|
| 146 |
-
"learning_rate": 1.
|
| 147 |
-
"loss": 6.
|
| 148 |
-
"mean_token_accuracy": 0.
|
| 149 |
-
"num_tokens":
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"entropy": 6.
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 6.
|
| 158 |
-
"mean_token_accuracy": 0.
|
| 159 |
-
"num_tokens":
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"entropy": 6.
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
-
"grad_norm":
|
| 166 |
-
"learning_rate": 1.
|
| 167 |
-
"loss": 6.
|
| 168 |
-
"mean_token_accuracy": 0.
|
| 169 |
-
"num_tokens":
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"entropy": 6.
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
-
"grad_norm": 3.
|
| 176 |
-
"learning_rate": 1.
|
| 177 |
-
"loss": 6.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"entropy": 6.
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 1.
|
| 187 |
-
"loss": 6.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"entropy": 6.
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
-
"grad_norm": 2.
|
| 196 |
-
"learning_rate": 1.
|
| 197 |
-
"loss": 6.
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"entropy": 6.
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
-
"grad_norm": 2.
|
| 206 |
-
"learning_rate": 1.
|
| 207 |
-
"loss": 6.
|
| 208 |
-
"mean_token_accuracy": 0.
|
| 209 |
-
"num_tokens":
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"entropy": 6.
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
-
"grad_norm": 2.
|
| 216 |
-
"learning_rate": 1.
|
| 217 |
-
"loss": 6.
|
| 218 |
-
"mean_token_accuracy": 0.
|
| 219 |
-
"num_tokens":
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"entropy": 6.
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 1.
|
| 227 |
-
"loss": 5.
|
| 228 |
-
"mean_token_accuracy": 0.
|
| 229 |
-
"num_tokens":
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"entropy": 5.
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate": 1.
|
| 237 |
-
"loss": 5.
|
| 238 |
-
"mean_token_accuracy": 0.
|
| 239 |
-
"num_tokens":
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"entropy": 6.
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 1.
|
| 247 |
-
"loss": 5.
|
| 248 |
-
"mean_token_accuracy": 0.
|
| 249 |
-
"num_tokens":
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"entropy": 6.
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
-
"grad_norm": 3.
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 5.
|
| 258 |
-
"mean_token_accuracy": 0.
|
| 259 |
-
"num_tokens":
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"entropy": 6.
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
-
"grad_norm": 2.
|
| 266 |
-
"learning_rate": 1.
|
| 267 |
-
"loss": 5.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"entropy": 5.
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
-
"grad_norm": 2.
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 5.
|
| 278 |
-
"mean_token_accuracy": 0.
|
| 279 |
-
"num_tokens":
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"entropy": 6.
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
-
"grad_norm": 3.
|
| 286 |
-
"learning_rate": 1.
|
| 287 |
-
"loss": 5.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"entropy":
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 1.
|
| 297 |
-
"loss": 5.
|
| 298 |
-
"mean_token_accuracy": 0.
|
| 299 |
-
"num_tokens":
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"entropy":
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 5.
|
| 308 |
-
"mean_token_accuracy": 0.
|
| 309 |
-
"num_tokens":
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"entropy":
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 1.
|
| 317 |
-
"loss": 5.
|
| 318 |
-
"mean_token_accuracy": 0.
|
| 319 |
-
"num_tokens":
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"entropy": 5.
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
-
"grad_norm": 2.
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 5.
|
| 328 |
-
"mean_token_accuracy": 0.
|
| 329 |
-
"num_tokens":
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"entropy":
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
-
"grad_norm": 2.
|
| 336 |
-
"learning_rate": 1.
|
| 337 |
-
"loss": 5.
|
| 338 |
-
"mean_token_accuracy": 0.
|
| 339 |
-
"num_tokens":
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"entropy": 5.
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 1.
|
| 347 |
-
"loss": 5.
|
| 348 |
-
"mean_token_accuracy": 0.
|
| 349 |
-
"num_tokens":
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
-
"eval_entropy":
|
| 355 |
-
"eval_loss": 5.
|
| 356 |
-
"eval_mean_token_accuracy": 0.
|
| 357 |
-
"eval_model_preparation_time": 0.
|
| 358 |
-
"eval_num_tokens":
|
| 359 |
-
"eval_runtime":
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second":
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"entropy": 5.
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
-
"grad_norm": 2.
|
| 368 |
-
"learning_rate": 1.
|
| 369 |
-
"loss": 5.
|
| 370 |
-
"mean_token_accuracy": 0.
|
| 371 |
-
"num_tokens":
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"entropy":
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 1.
|
| 379 |
-
"loss": 5.
|
| 380 |
-
"mean_token_accuracy": 0.
|
| 381 |
-
"num_tokens":
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"entropy": 5.
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 1.
|
| 389 |
-
"loss": 5.
|
| 390 |
-
"mean_token_accuracy": 0.
|
| 391 |
-
"num_tokens":
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"entropy": 5.
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
-
"grad_norm":
|
| 398 |
-
"learning_rate": 1.
|
| 399 |
-
"loss": 5.
|
| 400 |
-
"mean_token_accuracy": 0.
|
| 401 |
-
"num_tokens":
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"entropy": 5.
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 1.
|
| 409 |
-
"loss": 5.
|
| 410 |
-
"mean_token_accuracy": 0.
|
| 411 |
-
"num_tokens":
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"entropy": 5.
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
-
"grad_norm":
|
| 418 |
-
"learning_rate": 1.
|
| 419 |
-
"loss": 5.
|
| 420 |
-
"mean_token_accuracy": 0.
|
| 421 |
-
"num_tokens":
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"entropy": 5.
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 1.
|
| 429 |
-
"loss": 5.
|
| 430 |
-
"mean_token_accuracy": 0.
|
| 431 |
-
"num_tokens":
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"entropy": 5.
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss": 5.
|
| 440 |
-
"mean_token_accuracy": 0.
|
| 441 |
-
"num_tokens":
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
-
"entropy": 5.
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
-
"grad_norm": 2.
|
| 448 |
-
"learning_rate": 1.
|
| 449 |
-
"loss": 5.
|
| 450 |
-
"mean_token_accuracy": 0.
|
| 451 |
-
"num_tokens":
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"entropy": 5.
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
-
"grad_norm":
|
| 458 |
-
"learning_rate": 1.
|
| 459 |
-
"loss": 5.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
-
"entropy": 5.
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
-
"grad_norm":
|
| 468 |
-
"learning_rate": 1.
|
| 469 |
-
"loss": 5.
|
| 470 |
-
"mean_token_accuracy": 0.
|
| 471 |
-
"num_tokens":
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"entropy": 5.
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 1.
|
| 479 |
-
"loss": 5.
|
| 480 |
-
"mean_token_accuracy": 0.
|
| 481 |
-
"num_tokens":
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"entropy": 5.
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
-
"grad_norm":
|
| 488 |
-
"learning_rate": 1.
|
| 489 |
-
"loss": 5.
|
| 490 |
-
"mean_token_accuracy": 0.
|
| 491 |
-
"num_tokens":
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"entropy": 5.
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 1.
|
| 499 |
-
"loss": 5.
|
| 500 |
-
"mean_token_accuracy": 0.
|
| 501 |
-
"num_tokens":
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"entropy": 5.
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate": 1.
|
| 509 |
-
"loss": 5.
|
| 510 |
-
"mean_token_accuracy": 0.
|
| 511 |
-
"num_tokens":
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"entropy": 5.
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
-
"grad_norm": 1.
|
| 518 |
-
"learning_rate": 1.
|
| 519 |
-
"loss": 5.
|
| 520 |
-
"mean_token_accuracy": 0.
|
| 521 |
-
"num_tokens":
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"entropy": 5.
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
-
"grad_norm": 3.
|
| 528 |
-
"learning_rate": 1.
|
| 529 |
-
"loss": 5.
|
| 530 |
-
"mean_token_accuracy": 0.
|
| 531 |
-
"num_tokens":
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"entropy": 5.
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
-
"grad_norm": 1.
|
| 538 |
-
"learning_rate": 1.
|
| 539 |
-
"loss": 5.
|
| 540 |
-
"mean_token_accuracy": 0.
|
| 541 |
-
"num_tokens":
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"entropy": 5.
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
-
"grad_norm":
|
| 548 |
-
"learning_rate": 1.
|
| 549 |
-
"loss": 5.
|
| 550 |
-
"mean_token_accuracy": 0.
|
| 551 |
-
"num_tokens":
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"entropy": 5.
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
-
"grad_norm":
|
| 558 |
-
"learning_rate": 1.
|
| 559 |
-
"loss": 5.
|
| 560 |
-
"mean_token_accuracy": 0.
|
| 561 |
-
"num_tokens":
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"entropy": 5.
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
-
"grad_norm": 2.
|
| 568 |
-
"learning_rate": 1.
|
| 569 |
-
"loss": 5.
|
| 570 |
-
"mean_token_accuracy": 0.
|
| 571 |
-
"num_tokens":
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"entropy": 5.
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate": 1.
|
| 579 |
-
"loss": 5.
|
| 580 |
-
"mean_token_accuracy": 0.
|
| 581 |
-
"num_tokens":
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"entropy": 5.
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
-
"grad_norm": 1.
|
| 588 |
-
"learning_rate": 1.
|
| 589 |
-
"loss": 5.
|
| 590 |
-
"mean_token_accuracy": 0.
|
| 591 |
-
"num_tokens":
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
-
"entropy": 5.
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
-
"grad_norm": 1.
|
| 598 |
-
"learning_rate": 1.
|
| 599 |
-
"loss": 5.
|
| 600 |
-
"mean_token_accuracy": 0.
|
| 601 |
-
"num_tokens":
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"entropy": 5.
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
-
"grad_norm": 1.
|
| 608 |
-
"learning_rate": 1.
|
| 609 |
-
"loss": 5.
|
| 610 |
-
"mean_token_accuracy": 0.
|
| 611 |
-
"num_tokens":
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"entropy": 5.
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
-
"grad_norm":
|
| 618 |
-
"learning_rate": 1.
|
| 619 |
-
"loss": 5.
|
| 620 |
-
"mean_token_accuracy": 0.
|
| 621 |
-
"num_tokens":
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
-
"entropy": 5.
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
-
"grad_norm": 1.
|
| 628 |
-
"learning_rate": 1.
|
| 629 |
-
"loss": 5.
|
| 630 |
-
"mean_token_accuracy": 0.
|
| 631 |
-
"num_tokens":
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"entropy": 5.
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate": 1.
|
| 639 |
-
"loss": 5.
|
| 640 |
-
"mean_token_accuracy": 0.
|
| 641 |
-
"num_tokens":
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"entropy": 5.
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate": 1.
|
| 649 |
-
"loss":
|
| 650 |
-
"mean_token_accuracy": 0.
|
| 651 |
-
"num_tokens":
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
-
"entropy": 5.
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
-
"grad_norm": 1.
|
| 658 |
-
"learning_rate": 1.
|
| 659 |
-
"loss": 5.
|
| 660 |
-
"mean_token_accuracy": 0.
|
| 661 |
-
"num_tokens":
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"entropy": 5.
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
-
"grad_norm": 1.
|
| 668 |
-
"learning_rate": 1.
|
| 669 |
-
"loss": 5.
|
| 670 |
-
"mean_token_accuracy": 0.
|
| 671 |
-
"num_tokens":
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"entropy": 5.
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
-
"grad_norm":
|
| 678 |
-
"learning_rate": 1.
|
| 679 |
-
"loss": 5.
|
| 680 |
-
"mean_token_accuracy": 0.
|
| 681 |
-
"num_tokens":
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"entropy": 5.
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
-
"grad_norm":
|
| 688 |
-
"learning_rate": 1.
|
| 689 |
-
"loss": 5.
|
| 690 |
-
"mean_token_accuracy": 0.
|
| 691 |
-
"num_tokens":
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"entropy": 5.
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
-
"grad_norm": 1.
|
| 698 |
-
"learning_rate": 1.
|
| 699 |
-
"loss": 5.
|
| 700 |
-
"mean_token_accuracy": 0.
|
| 701 |
-
"num_tokens":
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"entropy": 5.
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
-
"grad_norm": 1.
|
| 708 |
-
"learning_rate": 1.
|
| 709 |
-
"loss": 5.
|
| 710 |
-
"mean_token_accuracy": 0.
|
| 711 |
-
"num_tokens":
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
-
"eval_entropy": 5.
|
| 717 |
-
"eval_loss": 5.
|
| 718 |
-
"eval_mean_token_accuracy": 0.
|
| 719 |
-
"eval_model_preparation_time": 0.
|
| 720 |
-
"eval_num_tokens":
|
| 721 |
-
"eval_runtime":
|
| 722 |
-
"eval_samples_per_second":
|
| 723 |
-
"eval_steps_per_second":
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"entropy": 5.
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
-
"grad_norm":
|
| 730 |
-
"learning_rate": 1.
|
| 731 |
-
"loss": 5.
|
| 732 |
-
"mean_token_accuracy": 0.
|
| 733 |
-
"num_tokens":
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"entropy": 5.
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
-
"grad_norm": 1.
|
| 740 |
-
"learning_rate": 1.
|
| 741 |
-
"loss": 5.
|
| 742 |
-
"mean_token_accuracy": 0.
|
| 743 |
-
"num_tokens":
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"entropy": 5.
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss": 5.
|
| 752 |
-
"mean_token_accuracy": 0.
|
| 753 |
-
"num_tokens":
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"entropy": 5.
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate":
|
| 761 |
-
"loss": 5.
|
| 762 |
-
"mean_token_accuracy": 0.
|
| 763 |
-
"num_tokens":
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"entropy": 5.
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 5.
|
| 772 |
-
"mean_token_accuracy": 0.
|
| 773 |
-
"num_tokens":
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"entropy": 5.
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
-
"grad_norm": 2.
|
| 780 |
-
"learning_rate":
|
| 781 |
-
"loss": 5.
|
| 782 |
-
"mean_token_accuracy": 0.
|
| 783 |
-
"num_tokens":
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
-
"entropy": 5.
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
-
"grad_norm": 1.
|
| 790 |
-
"learning_rate":
|
| 791 |
-
"loss": 5.
|
| 792 |
-
"mean_token_accuracy": 0.
|
| 793 |
-
"num_tokens":
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"entropy": 5.
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
-
"grad_norm":
|
| 800 |
-
"learning_rate":
|
| 801 |
-
"loss": 5.
|
| 802 |
-
"mean_token_accuracy": 0.
|
| 803 |
-
"num_tokens":
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
-
"entropy": 5.
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
-
"grad_norm":
|
| 810 |
-
"learning_rate":
|
| 811 |
-
"loss": 5.
|
| 812 |
-
"mean_token_accuracy": 0.
|
| 813 |
-
"num_tokens":
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"entropy": 5.
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 5.
|
| 822 |
-
"mean_token_accuracy": 0.
|
| 823 |
-
"num_tokens":
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
-
"entropy": 5.
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
-
"grad_norm": 2.
|
| 830 |
-
"learning_rate":
|
| 831 |
-
"loss": 5.
|
| 832 |
-
"mean_token_accuracy": 0.
|
| 833 |
-
"num_tokens":
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
-
"entropy": 5.
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
-
"grad_norm": 1.
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 5.
|
| 842 |
-
"mean_token_accuracy": 0.
|
| 843 |
-
"num_tokens":
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
-
"entropy": 5.
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
-
"grad_norm": 1.
|
| 850 |
-
"learning_rate":
|
| 851 |
-
"loss": 5.
|
| 852 |
-
"mean_token_accuracy": 0.
|
| 853 |
-
"num_tokens":
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
-
"entropy": 5.
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
-
"grad_norm": 1.
|
| 860 |
-
"learning_rate":
|
| 861 |
-
"loss": 5.
|
| 862 |
-
"mean_token_accuracy": 0.
|
| 863 |
-
"num_tokens":
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"entropy": 5.
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
-
"grad_norm": 1.
|
| 870 |
-
"learning_rate":
|
| 871 |
-
"loss": 5.
|
| 872 |
-
"mean_token_accuracy": 0.
|
| 873 |
-
"num_tokens":
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
-
"entropy": 5.
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
-
"grad_norm": 1.
|
| 880 |
-
"learning_rate":
|
| 881 |
-
"loss": 5.
|
| 882 |
-
"mean_token_accuracy": 0.
|
| 883 |
-
"num_tokens":
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"entropy": 5.
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
-
"grad_norm": 1.
|
| 890 |
-
"learning_rate":
|
| 891 |
-
"loss": 5.
|
| 892 |
-
"mean_token_accuracy": 0.
|
| 893 |
-
"num_tokens":
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
-
"entropy": 5.
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
-
"grad_norm":
|
| 900 |
-
"learning_rate":
|
| 901 |
-
"loss": 5.
|
| 902 |
-
"mean_token_accuracy": 0.
|
| 903 |
-
"num_tokens":
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
-
"entropy": 5.
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 5.
|
| 912 |
-
"mean_token_accuracy": 0.
|
| 913 |
-
"num_tokens":
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
-
"entropy": 5.
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
-
"grad_norm": 1.
|
| 920 |
-
"learning_rate":
|
| 921 |
-
"loss": 5.
|
| 922 |
-
"mean_token_accuracy": 0.
|
| 923 |
-
"num_tokens":
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"entropy": 5.
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
-
"grad_norm": 1.
|
| 930 |
-
"learning_rate":
|
| 931 |
-
"loss": 5.
|
| 932 |
-
"mean_token_accuracy": 0.
|
| 933 |
-
"num_tokens":
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
-
"entropy": 5.
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
-
"grad_norm": 1.
|
| 940 |
-
"learning_rate":
|
| 941 |
-
"loss": 5.
|
| 942 |
-
"mean_token_accuracy": 0.
|
| 943 |
-
"num_tokens":
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
-
"entropy": 5.
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
-
"grad_norm": 1.
|
| 950 |
-
"learning_rate":
|
| 951 |
-
"loss": 5.
|
| 952 |
-
"mean_token_accuracy": 0.
|
| 953 |
-
"num_tokens":
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"entropy": 5.
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
-
"grad_norm":
|
| 960 |
-
"learning_rate":
|
| 961 |
-
"loss": 5.
|
| 962 |
-
"mean_token_accuracy": 0.
|
| 963 |
-
"num_tokens":
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
-
"entropy": 5.
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
-
"grad_norm":
|
| 970 |
-
"learning_rate":
|
| 971 |
-
"loss": 5.
|
| 972 |
-
"mean_token_accuracy": 0.
|
| 973 |
-
"num_tokens":
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"entropy": 5.
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
-
"grad_norm":
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 5.
|
| 982 |
-
"mean_token_accuracy": 0.
|
| 983 |
-
"num_tokens":
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
-
"entropy": 5.
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
-
"grad_norm":
|
| 990 |
-
"learning_rate":
|
| 991 |
-
"loss": 5.
|
| 992 |
-
"mean_token_accuracy": 0.
|
| 993 |
-
"num_tokens":
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
-
"entropy": 5.
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
-
"grad_norm": 1.
|
| 1000 |
-
"learning_rate":
|
| 1001 |
-
"loss": 5.
|
| 1002 |
-
"mean_token_accuracy": 0.
|
| 1003 |
-
"num_tokens":
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"entropy": 5.
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
-
"grad_norm": 1.
|
| 1010 |
-
"learning_rate":
|
| 1011 |
-
"loss": 5.
|
| 1012 |
-
"mean_token_accuracy": 0.
|
| 1013 |
-
"num_tokens":
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
-
"entropy": 5.
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
-
"grad_norm":
|
| 1020 |
-
"learning_rate":
|
| 1021 |
-
"loss": 5.
|
| 1022 |
-
"mean_token_accuracy": 0.
|
| 1023 |
-
"num_tokens":
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"entropy": 5.
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
-
"grad_norm": 1.
|
| 1030 |
-
"learning_rate":
|
| 1031 |
-
"loss": 5.
|
| 1032 |
-
"mean_token_accuracy": 0.
|
| 1033 |
-
"num_tokens":
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
-
"entropy": 5.
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
-
"grad_norm": 1.
|
| 1040 |
-
"learning_rate":
|
| 1041 |
-
"loss": 5.
|
| 1042 |
-
"mean_token_accuracy": 0.
|
| 1043 |
-
"num_tokens":
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"entropy": 5.
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
-
"grad_norm": 1.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 5.
|
| 1052 |
-
"mean_token_accuracy": 0.
|
| 1053 |
-
"num_tokens":
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
-
"entropy": 5.
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
-
"grad_norm": 1.
|
| 1060 |
-
"learning_rate":
|
| 1061 |
-
"loss": 5.
|
| 1062 |
-
"mean_token_accuracy": 0.
|
| 1063 |
-
"num_tokens":
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
-
"entropy": 5.
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
-
"grad_norm": 1.
|
| 1070 |
-
"learning_rate":
|
| 1071 |
-
"loss": 5.
|
| 1072 |
-
"mean_token_accuracy": 0.
|
| 1073 |
-
"num_tokens":
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
-
"eval_entropy": 5.
|
| 1079 |
-
"eval_loss": 5.
|
| 1080 |
-
"eval_mean_token_accuracy": 0.
|
| 1081 |
-
"eval_model_preparation_time": 0.
|
| 1082 |
-
"eval_num_tokens":
|
| 1083 |
-
"eval_runtime":
|
| 1084 |
-
"eval_samples_per_second":
|
| 1085 |
-
"eval_steps_per_second":
|
| 1086 |
"step": 5211
|
| 1087 |
}
|
| 1088 |
],
|
| 1089 |
"logging_steps": 50,
|
| 1090 |
-
"max_steps":
|
| 1091 |
"num_input_tokens_seen": 0,
|
| 1092 |
-
"num_train_epochs":
|
| 1093 |
"save_steps": 500,
|
| 1094 |
"stateful_callbacks": {
|
| 1095 |
"TrainerControl": {
|
|
@@ -1103,7 +1103,7 @@
|
|
| 1103 |
"attributes": {}
|
| 1104 |
}
|
| 1105 |
},
|
| 1106 |
-
"total_flos": 7.
|
| 1107 |
"train_batch_size": 2,
|
| 1108 |
"trial_name": null,
|
| 1109 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 5211,
|
| 3 |
+
"best_metric": 5.53223991394043,
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-5211",
|
| 5 |
"epoch": 3.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"entropy": 3.606692385673523,
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
+
"grad_norm": 3.2999913692474365,
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
+
"loss": 13.6598,
|
| 18 |
+
"mean_token_accuracy": 0.16028020828962325,
|
| 19 |
+
"num_tokens": 53993.0,
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"entropy": 3.618675880432129,
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
+
"grad_norm": 3.101252555847168,
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
+
"loss": 14.0188,
|
| 28 |
+
"mean_token_accuracy": 0.1508466500043869,
|
| 29 |
+
"num_tokens": 110134.0,
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"entropy": 3.5215235900878907,
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
+
"grad_norm": 3.513662815093994,
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
+
"loss": 12.8555,
|
| 38 |
+
"mean_token_accuracy": 0.18527640983462335,
|
| 39 |
+
"num_tokens": 160191.0,
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"entropy": 3.667909698486328,
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
+
"grad_norm": 4.327610492706299,
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
+
"loss": 13.5394,
|
| 48 |
+
"mean_token_accuracy": 0.157139780074358,
|
| 49 |
+
"num_tokens": 214993.0,
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"entropy": 3.768263258934021,
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
+
"grad_norm": 4.290107250213623,
|
| 56 |
+
"learning_rate": 1.988450206246317e-06,
|
| 57 |
+
"loss": 12.8912,
|
| 58 |
+
"mean_token_accuracy": 0.17374794125556947,
|
| 59 |
+
"num_tokens": 268184.0,
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"entropy": 3.990619196891785,
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
+
"grad_norm": 4.444278717041016,
|
| 66 |
+
"learning_rate": 1.976664702416028e-06,
|
| 67 |
+
"loss": 12.455,
|
| 68 |
+
"mean_token_accuracy": 0.17780130118131637,
|
| 69 |
+
"num_tokens": 319458.0,
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"entropy": 4.162646284103394,
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
+
"grad_norm": 5.615262508392334,
|
| 76 |
+
"learning_rate": 1.9648791985857395e-06,
|
| 77 |
+
"loss": 12.0893,
|
| 78 |
+
"mean_token_accuracy": 0.18191319867968558,
|
| 79 |
+
"num_tokens": 373337.0,
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"entropy": 4.532100868225098,
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
+
"grad_norm": 10.074016571044922,
|
| 86 |
+
"learning_rate": 1.9530936947554507e-06,
|
| 87 |
+
"loss": 11.9261,
|
| 88 |
+
"mean_token_accuracy": 0.169477596282959,
|
| 89 |
+
"num_tokens": 427526.0,
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"entropy": 4.923871030807495,
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
+
"grad_norm": 16.220163345336914,
|
| 96 |
+
"learning_rate": 1.9413081909251622e-06,
|
| 97 |
+
"loss": 11.0048,
|
| 98 |
+
"mean_token_accuracy": 0.1704501649737358,
|
| 99 |
+
"num_tokens": 480528.0,
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"entropy": 5.521005854606629,
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
+
"grad_norm": 29.904008865356445,
|
| 106 |
+
"learning_rate": 1.9295226870948733e-06,
|
| 107 |
+
"loss": 9.6524,
|
| 108 |
+
"mean_token_accuracy": 0.16450899541378022,
|
| 109 |
+
"num_tokens": 535314.0,
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"entropy": 6.092623329162597,
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
+
"grad_norm": 17.821575164794922,
|
| 116 |
+
"learning_rate": 1.9177371832645845e-06,
|
| 117 |
+
"loss": 8.1054,
|
| 118 |
+
"mean_token_accuracy": 0.17205011785030366,
|
| 119 |
+
"num_tokens": 588410.0,
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"entropy": 6.385262680053711,
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
+
"grad_norm": 5.502202987670898,
|
| 126 |
+
"learning_rate": 1.9059516794342958e-06,
|
| 127 |
+
"loss": 7.4313,
|
| 128 |
+
"mean_token_accuracy": 0.1734227080643177,
|
| 129 |
+
"num_tokens": 641736.0,
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"entropy": 6.278562617301941,
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
+
"grad_norm": 5.4657697677612305,
|
| 136 |
+
"learning_rate": 1.8941661756040071e-06,
|
| 137 |
+
"loss": 6.9266,
|
| 138 |
+
"mean_token_accuracy": 0.18680249139666558,
|
| 139 |
+
"num_tokens": 692200.0,
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"entropy": 6.553266277313233,
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
+
"grad_norm": 4.955812931060791,
|
| 146 |
+
"learning_rate": 1.8823806717737183e-06,
|
| 147 |
+
"loss": 6.9847,
|
| 148 |
+
"mean_token_accuracy": 0.16679802387952805,
|
| 149 |
+
"num_tokens": 745830.0,
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"entropy": 6.470935583114624,
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
+
"grad_norm": 4.198381423950195,
|
| 156 |
+
"learning_rate": 1.8705951679434296e-06,
|
| 157 |
+
"loss": 6.7277,
|
| 158 |
+
"mean_token_accuracy": 0.17847734570503235,
|
| 159 |
+
"num_tokens": 798872.0,
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"entropy": 6.5620588779449465,
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
+
"grad_norm": 3.1793746948242188,
|
| 166 |
+
"learning_rate": 1.8588096641131407e-06,
|
| 167 |
+
"loss": 6.7032,
|
| 168 |
+
"mean_token_accuracy": 0.17336134731769562,
|
| 169 |
+
"num_tokens": 853045.0,
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"entropy": 6.532204885482788,
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
+
"grad_norm": 3.824537515640259,
|
| 176 |
+
"learning_rate": 1.847024160282852e-06,
|
| 177 |
+
"loss": 6.5762,
|
| 178 |
+
"mean_token_accuracy": 0.1805124071240425,
|
| 179 |
+
"num_tokens": 907679.0,
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"entropy": 6.535988225936889,
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
+
"grad_norm": 4.350001811981201,
|
| 186 |
+
"learning_rate": 1.8352386564525632e-06,
|
| 187 |
+
"loss": 6.505,
|
| 188 |
+
"mean_token_accuracy": 0.1842605724930763,
|
| 189 |
+
"num_tokens": 964170.0,
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"entropy": 6.204533562660218,
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
+
"grad_norm": 2.193660020828247,
|
| 196 |
+
"learning_rate": 1.8234531526222745e-06,
|
| 197 |
+
"loss": 6.1211,
|
| 198 |
+
"mean_token_accuracy": 0.21968430042266845,
|
| 199 |
+
"num_tokens": 1015909.0,
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"entropy": 6.308737449645996,
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
+
"grad_norm": 2.325622320175171,
|
| 206 |
+
"learning_rate": 1.8116676487919857e-06,
|
| 207 |
+
"loss": 6.1653,
|
| 208 |
+
"mean_token_accuracy": 0.21636426240205764,
|
| 209 |
+
"num_tokens": 1068859.0,
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"entropy": 6.332560749053955,
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
+
"grad_norm": 2.0439090728759766,
|
| 216 |
+
"learning_rate": 1.799882144961697e-06,
|
| 217 |
+
"loss": 6.1559,
|
| 218 |
+
"mean_token_accuracy": 0.21859725564718246,
|
| 219 |
+
"num_tokens": 1123202.0,
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"entropy": 6.042124252319336,
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
+
"grad_norm": 3.621903657913208,
|
| 226 |
+
"learning_rate": 1.7880966411314081e-06,
|
| 227 |
+
"loss": 5.8441,
|
| 228 |
+
"mean_token_accuracy": 0.24906315237283708,
|
| 229 |
+
"num_tokens": 1173403.0,
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"entropy": 5.921343173980713,
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
+
"grad_norm": 5.658033847808838,
|
| 236 |
+
"learning_rate": 1.7763111373011195e-06,
|
| 237 |
+
"loss": 5.7104,
|
| 238 |
+
"mean_token_accuracy": 0.2625067520141602,
|
| 239 |
+
"num_tokens": 1225026.0,
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"entropy": 6.093586492538452,
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
+
"grad_norm": 2.4292995929718018,
|
| 246 |
+
"learning_rate": 1.7645256334708308e-06,
|
| 247 |
+
"loss": 5.8658,
|
| 248 |
+
"mean_token_accuracy": 0.24842385441064835,
|
| 249 |
+
"num_tokens": 1279013.0,
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"entropy": 6.119112596511841,
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
+
"grad_norm": 3.369384288787842,
|
| 256 |
+
"learning_rate": 1.752740129640542e-06,
|
| 257 |
+
"loss": 5.8784,
|
| 258 |
+
"mean_token_accuracy": 0.24857850253582,
|
| 259 |
+
"num_tokens": 1332547.0,
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"entropy": 6.025163550376892,
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
+
"grad_norm": 2.5110116004943848,
|
| 266 |
+
"learning_rate": 1.7409546258102533e-06,
|
| 267 |
+
"loss": 5.7769,
|
| 268 |
+
"mean_token_accuracy": 0.25835376888513567,
|
| 269 |
+
"num_tokens": 1385192.0,
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"entropy": 5.877259612083435,
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
+
"grad_norm": 2.4179303646087646,
|
| 276 |
+
"learning_rate": 1.7291691219799646e-06,
|
| 277 |
+
"loss": 5.6284,
|
| 278 |
+
"mean_token_accuracy": 0.2756252554059029,
|
| 279 |
+
"num_tokens": 1437071.0,
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"entropy": 6.002246947288513,
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
+
"grad_norm": 3.494359016418457,
|
| 286 |
+
"learning_rate": 1.717383618149676e-06,
|
| 287 |
+
"loss": 5.747,
|
| 288 |
+
"mean_token_accuracy": 0.26462210685014725,
|
| 289 |
+
"num_tokens": 1490818.0,
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"entropy": 5.991955623626709,
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
+
"grad_norm": 2.340975761413574,
|
| 296 |
+
"learning_rate": 1.705598114319387e-06,
|
| 297 |
+
"loss": 5.7379,
|
| 298 |
+
"mean_token_accuracy": 0.26444981098175047,
|
| 299 |
+
"num_tokens": 1544997.0,
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"entropy": 5.91768889427185,
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
+
"grad_norm": 2.2394514083862305,
|
| 306 |
+
"learning_rate": 1.6938126104890984e-06,
|
| 307 |
+
"loss": 5.6564,
|
| 308 |
+
"mean_token_accuracy": 0.2730415526032448,
|
| 309 |
+
"num_tokens": 1598302.0,
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"entropy": 5.982716989517212,
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
+
"grad_norm": 1.876839518547058,
|
| 316 |
+
"learning_rate": 1.6820271066588098e-06,
|
| 317 |
+
"loss": 5.7215,
|
| 318 |
+
"mean_token_accuracy": 0.26642445534467696,
|
| 319 |
+
"num_tokens": 1655267.0,
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"entropy": 5.820467872619629,
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
+
"grad_norm": 2.219966173171997,
|
| 326 |
+
"learning_rate": 1.6702416028285209e-06,
|
| 327 |
+
"loss": 5.5555,
|
| 328 |
+
"mean_token_accuracy": 0.2856418335437775,
|
| 329 |
+
"num_tokens": 1709199.0,
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"entropy": 5.996349005699158,
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
+
"grad_norm": 2.247213840484619,
|
| 336 |
+
"learning_rate": 1.6584560989982322e-06,
|
| 337 |
+
"loss": 5.7283,
|
| 338 |
+
"mean_token_accuracy": 0.2696125540137291,
|
| 339 |
+
"num_tokens": 1765443.0,
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"entropy": 5.696683068275451,
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
+
"grad_norm": 2.8499979972839355,
|
| 346 |
+
"learning_rate": 1.6466705951679433e-06,
|
| 347 |
+
"loss": 5.4335,
|
| 348 |
+
"mean_token_accuracy": 0.29918427973985673,
|
| 349 |
+
"num_tokens": 1817494.0,
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
+
"eval_entropy": 5.993559589034401,
|
| 355 |
+
"eval_loss": 5.737204551696777,
|
| 356 |
+
"eval_mean_token_accuracy": 0.2618687468739699,
|
| 357 |
+
"eval_model_preparation_time": 0.0045,
|
| 358 |
+
"eval_num_tokens": 1856362.0,
|
| 359 |
+
"eval_runtime": 50.5332,
|
| 360 |
+
"eval_samples_per_second": 8.588,
|
| 361 |
+
"eval_steps_per_second": 4.294,
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"entropy": 5.746842083930969,
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
+
"grad_norm": 2.33052921295166,
|
| 368 |
+
"learning_rate": 1.6348850913376547e-06,
|
| 369 |
+
"loss": 5.4796,
|
| 370 |
+
"mean_token_accuracy": 0.2966849410533905,
|
| 371 |
+
"num_tokens": 1870353.0,
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"entropy": 5.859029049873352,
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
+
"grad_norm": 1.6248886585235596,
|
| 378 |
+
"learning_rate": 1.6230995875073658e-06,
|
| 379 |
+
"loss": 5.5975,
|
| 380 |
+
"mean_token_accuracy": 0.2838129925727844,
|
| 381 |
+
"num_tokens": 1926205.0,
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"entropy": 5.731445336341858,
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
+
"grad_norm": 1.6941566467285156,
|
| 388 |
+
"learning_rate": 1.6113140836770771e-06,
|
| 389 |
+
"loss": 5.476,
|
| 390 |
+
"mean_token_accuracy": 0.2992346465587616,
|
| 391 |
+
"num_tokens": 1979821.0,
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"entropy": 5.6993954515457155,
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
+
"grad_norm": 1.1746597290039062,
|
| 398 |
+
"learning_rate": 1.5995285798467883e-06,
|
| 399 |
+
"loss": 5.4608,
|
| 400 |
+
"mean_token_accuracy": 0.3000726142525673,
|
| 401 |
+
"num_tokens": 2034373.0,
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"entropy": 5.668873124122619,
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
+
"grad_norm": 1.728211760520935,
|
| 408 |
+
"learning_rate": 1.5877430760164996e-06,
|
| 409 |
+
"loss": 5.4347,
|
| 410 |
+
"mean_token_accuracy": 0.3033922725915909,
|
| 411 |
+
"num_tokens": 2087339.0,
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
+
"entropy": 5.624621086120605,
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
+
"grad_norm": 1.4078539609909058,
|
| 418 |
+
"learning_rate": 1.5759575721862107e-06,
|
| 419 |
+
"loss": 5.3954,
|
| 420 |
+
"mean_token_accuracy": 0.30784171640872954,
|
| 421 |
+
"num_tokens": 2139520.0,
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"entropy": 5.7141213130950925,
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
+
"grad_norm": 2.186459541320801,
|
| 428 |
+
"learning_rate": 1.564172068355922e-06,
|
| 429 |
+
"loss": 5.4847,
|
| 430 |
+
"mean_token_accuracy": 0.29594049394130706,
|
| 431 |
+
"num_tokens": 2193987.0,
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"entropy": 5.632415266036987,
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
+
"grad_norm": 1.3601349592208862,
|
| 438 |
+
"learning_rate": 1.5523865645256334e-06,
|
| 439 |
+
"loss": 5.4135,
|
| 440 |
+
"mean_token_accuracy": 0.30366597563028336,
|
| 441 |
+
"num_tokens": 2249616.0,
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
+
"entropy": 5.510904269218445,
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
+
"grad_norm": 2.065760612487793,
|
| 448 |
+
"learning_rate": 1.5406010606953445e-06,
|
| 449 |
+
"loss": 5.2904,
|
| 450 |
+
"mean_token_accuracy": 0.3211754837632179,
|
| 451 |
+
"num_tokens": 2300863.0,
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"entropy": 5.703383626937867,
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
+
"grad_norm": 1.1172698736190796,
|
| 458 |
+
"learning_rate": 1.5288155568650559e-06,
|
| 459 |
+
"loss": 5.4802,
|
| 460 |
+
"mean_token_accuracy": 0.29713701367378237,
|
| 461 |
+
"num_tokens": 2356029.0,
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
+
"entropy": 5.565930342674255,
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
+
"grad_norm": 1.7528513669967651,
|
| 468 |
+
"learning_rate": 1.5170300530347672e-06,
|
| 469 |
+
"loss": 5.3518,
|
| 470 |
+
"mean_token_accuracy": 0.31301232606172563,
|
| 471 |
+
"num_tokens": 2408957.0,
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"entropy": 5.496430187225342,
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
+
"grad_norm": 1.892640233039856,
|
| 478 |
+
"learning_rate": 1.5052445492044786e-06,
|
| 479 |
+
"loss": 5.2967,
|
| 480 |
+
"mean_token_accuracy": 0.3181899458169937,
|
| 481 |
+
"num_tokens": 2462569.0,
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"entropy": 5.725150098800659,
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
+
"grad_norm": 1.774940848350525,
|
| 488 |
+
"learning_rate": 1.4934590453741897e-06,
|
| 489 |
+
"loss": 5.5215,
|
| 490 |
+
"mean_token_accuracy": 0.29055028676986694,
|
| 491 |
+
"num_tokens": 2518544.0,
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"entropy": 5.4884827613830565,
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
+
"grad_norm": 2.2167599201202393,
|
| 498 |
+
"learning_rate": 1.481673541543901e-06,
|
| 499 |
+
"loss": 5.2917,
|
| 500 |
+
"mean_token_accuracy": 0.31803421139717103,
|
| 501 |
+
"num_tokens": 2570863.0,
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"entropy": 5.697079472541809,
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
+
"grad_norm": 1.6489030122756958,
|
| 508 |
+
"learning_rate": 1.4698880377136124e-06,
|
| 509 |
+
"loss": 5.4982,
|
| 510 |
+
"mean_token_accuracy": 0.2925163987278938,
|
| 511 |
+
"num_tokens": 2626998.0,
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"entropy": 5.46209939956665,
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
+
"grad_norm": 1.153914451599121,
|
| 518 |
+
"learning_rate": 1.4581025338833235e-06,
|
| 519 |
+
"loss": 5.2736,
|
| 520 |
+
"mean_token_accuracy": 0.3182168474793434,
|
| 521 |
+
"num_tokens": 2681568.0,
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"entropy": 5.4405768728256225,
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
+
"grad_norm": 3.6614978313446045,
|
| 528 |
+
"learning_rate": 1.4463170300530348e-06,
|
| 529 |
+
"loss": 5.2515,
|
| 530 |
+
"mean_token_accuracy": 0.3218736210465431,
|
| 531 |
+
"num_tokens": 2733587.0,
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"entropy": 5.528175053596496,
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
+
"grad_norm": 1.0849746465682983,
|
| 538 |
+
"learning_rate": 1.434531526222746e-06,
|
| 539 |
+
"loss": 5.3378,
|
| 540 |
+
"mean_token_accuracy": 0.31061659604310987,
|
| 541 |
+
"num_tokens": 2787003.0,
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"entropy": 5.46110897064209,
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
+
"grad_norm": 1.8315683603286743,
|
| 548 |
+
"learning_rate": 1.4227460223924573e-06,
|
| 549 |
+
"loss": 5.2782,
|
| 550 |
+
"mean_token_accuracy": 0.31781029611825945,
|
| 551 |
+
"num_tokens": 2840263.0,
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"entropy": 5.455560960769653,
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
+
"grad_norm": 1.1859091520309448,
|
| 558 |
+
"learning_rate": 1.4109605185621684e-06,
|
| 559 |
+
"loss": 5.2735,
|
| 560 |
+
"mean_token_accuracy": 0.3194814011454582,
|
| 561 |
+
"num_tokens": 2894186.0,
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"entropy": 5.430496115684509,
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
+
"grad_norm": 2.3500001430511475,
|
| 568 |
+
"learning_rate": 1.3991750147318797e-06,
|
| 569 |
+
"loss": 5.2464,
|
| 570 |
+
"mean_token_accuracy": 0.32140792965888976,
|
| 571 |
+
"num_tokens": 2948171.0,
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"entropy": 5.588023023605347,
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
+
"grad_norm": 1.727825403213501,
|
| 578 |
+
"learning_rate": 1.3873895109015909e-06,
|
| 579 |
+
"loss": 5.4028,
|
| 580 |
+
"mean_token_accuracy": 0.3039530631899834,
|
| 581 |
+
"num_tokens": 3002678.0,
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"entropy": 5.410525422096253,
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
+
"grad_norm": 1.3401474952697754,
|
| 588 |
+
"learning_rate": 1.3756040070713022e-06,
|
| 589 |
+
"loss": 5.2298,
|
| 590 |
+
"mean_token_accuracy": 0.324065263569355,
|
| 591 |
+
"num_tokens": 3055844.0,
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"entropy": 5.36959942817688,
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
+
"grad_norm": 1.1892589330673218,
|
| 598 |
+
"learning_rate": 1.3638185032410133e-06,
|
| 599 |
+
"loss": 5.1956,
|
| 600 |
+
"mean_token_accuracy": 0.32639502108097074,
|
| 601 |
+
"num_tokens": 3108636.0,
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"entropy": 5.53826907157898,
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
+
"grad_norm": 1.2652360200881958,
|
| 608 |
+
"learning_rate": 1.3520329994107247e-06,
|
| 609 |
+
"loss": 5.3583,
|
| 610 |
+
"mean_token_accuracy": 0.3074926760792732,
|
| 611 |
+
"num_tokens": 3162627.0,
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"entropy": 5.417449145317078,
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
+
"grad_norm": 1.584312915802002,
|
| 618 |
+
"learning_rate": 1.340247495580436e-06,
|
| 619 |
+
"loss": 5.2388,
|
| 620 |
+
"mean_token_accuracy": 0.32019727885723115,
|
| 621 |
+
"num_tokens": 3216409.0,
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"entropy": 5.241390740871429,
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
+
"grad_norm": 1.5219439268112183,
|
| 628 |
+
"learning_rate": 1.3284619917501471e-06,
|
| 629 |
+
"loss": 5.0645,
|
| 630 |
+
"mean_token_accuracy": 0.3445430138707161,
|
| 631 |
+
"num_tokens": 3266967.0,
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"entropy": 5.405424036979675,
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
+
"grad_norm": 2.1165153980255127,
|
| 638 |
+
"learning_rate": 1.3166764879198585e-06,
|
| 639 |
+
"loss": 5.232,
|
| 640 |
+
"mean_token_accuracy": 0.32085000157356264,
|
| 641 |
+
"num_tokens": 3319877.0,
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"entropy": 5.123006024360657,
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
+
"grad_norm": 1.2189785242080688,
|
| 648 |
+
"learning_rate": 1.3048909840895698e-06,
|
| 649 |
+
"loss": 4.9582,
|
| 650 |
+
"mean_token_accuracy": 0.356108532845974,
|
| 651 |
+
"num_tokens": 3368569.0,
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"entropy": 5.417610831260681,
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
+
"grad_norm": 1.5157604217529297,
|
| 658 |
+
"learning_rate": 1.2931054802592812e-06,
|
| 659 |
+
"loss": 5.2454,
|
| 660 |
+
"mean_token_accuracy": 0.31976755023002623,
|
| 661 |
+
"num_tokens": 3422449.0,
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"entropy": 5.409690895080566,
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
+
"grad_norm": 1.3088161945343018,
|
| 668 |
+
"learning_rate": 1.2813199764289923e-06,
|
| 669 |
+
"loss": 5.2348,
|
| 670 |
+
"mean_token_accuracy": 0.32325415283441544,
|
| 671 |
+
"num_tokens": 3474399.0,
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"entropy": 5.44662567615509,
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
+
"grad_norm": 2.178372621536255,
|
| 678 |
+
"learning_rate": 1.2695344725987036e-06,
|
| 679 |
+
"loss": 5.2661,
|
| 680 |
+
"mean_token_accuracy": 0.3182847076654434,
|
| 681 |
+
"num_tokens": 3527726.0,
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"entropy": 5.512614865303039,
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
+
"grad_norm": 1.3050425052642822,
|
| 688 |
+
"learning_rate": 1.2577489687684147e-06,
|
| 689 |
+
"loss": 5.3416,
|
| 690 |
+
"mean_token_accuracy": 0.3084403133392334,
|
| 691 |
+
"num_tokens": 3581980.0,
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"entropy": 5.379772834777832,
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
+
"grad_norm": 1.4584404230117798,
|
| 698 |
+
"learning_rate": 1.245963464938126e-06,
|
| 699 |
+
"loss": 5.2087,
|
| 700 |
+
"mean_token_accuracy": 0.32388432770967485,
|
| 701 |
+
"num_tokens": 3635393.0,
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"entropy": 5.483665924072266,
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
+
"grad_norm": 1.2157734632492065,
|
| 708 |
+
"learning_rate": 1.2341779611078374e-06,
|
| 709 |
+
"loss": 5.3101,
|
| 710 |
+
"mean_token_accuracy": 0.3121953472495079,
|
| 711 |
+
"num_tokens": 3689894.0,
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
+
"eval_entropy": 5.711394641805904,
|
| 717 |
+
"eval_loss": 5.55628776550293,
|
| 718 |
+
"eval_mean_token_accuracy": 0.2764948787777105,
|
| 719 |
+
"eval_model_preparation_time": 0.0045,
|
| 720 |
+
"eval_num_tokens": 3712724.0,
|
| 721 |
+
"eval_runtime": 50.187,
|
| 722 |
+
"eval_samples_per_second": 8.648,
|
| 723 |
+
"eval_steps_per_second": 4.324,
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"entropy": 5.349283556938172,
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
+
"grad_norm": 1.1696771383285522,
|
| 730 |
+
"learning_rate": 1.2223924572775486e-06,
|
| 731 |
+
"loss": 5.1782,
|
| 732 |
+
"mean_token_accuracy": 0.33028870791196824,
|
| 733 |
+
"num_tokens": 3740861.0,
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"entropy": 5.4721107006073,
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
+
"grad_norm": 1.8449370861053467,
|
| 740 |
+
"learning_rate": 1.2106069534472599e-06,
|
| 741 |
+
"loss": 5.2978,
|
| 742 |
+
"mean_token_accuracy": 0.31511022299528124,
|
| 743 |
+
"num_tokens": 3794869.0,
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"entropy": 5.404226851463318,
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
+
"grad_norm": 3.789496660232544,
|
| 750 |
+
"learning_rate": 1.198821449616971e-06,
|
| 751 |
+
"loss": 5.2371,
|
| 752 |
+
"mean_token_accuracy": 0.32092176616191864,
|
| 753 |
+
"num_tokens": 3848573.0,
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
+
"entropy": 5.435445628166199,
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
+
"grad_norm": 2.2847959995269775,
|
| 760 |
+
"learning_rate": 1.1870359457866824e-06,
|
| 761 |
+
"loss": 5.2662,
|
| 762 |
+
"mean_token_accuracy": 0.3186633634567261,
|
| 763 |
+
"num_tokens": 3901204.0,
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
+
"entropy": 5.4066293334960935,
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
+
"grad_norm": 1.0950902700424194,
|
| 770 |
+
"learning_rate": 1.1752504419563935e-06,
|
| 771 |
+
"loss": 5.2345,
|
| 772 |
+
"mean_token_accuracy": 0.32156052827835085,
|
| 773 |
+
"num_tokens": 3953753.0,
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
+
"entropy": 5.272332944869995,
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
+
"grad_norm": 2.1477339267730713,
|
| 780 |
+
"learning_rate": 1.1634649381261048e-06,
|
| 781 |
+
"loss": 5.1091,
|
| 782 |
+
"mean_token_accuracy": 0.3380983591079712,
|
| 783 |
+
"num_tokens": 4005481.0,
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"entropy": 5.4118804311752315,
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
+
"grad_norm": 1.4509484767913818,
|
| 790 |
+
"learning_rate": 1.151679434295816e-06,
|
| 791 |
+
"loss": 5.2448,
|
| 792 |
+
"mean_token_accuracy": 0.3208243528008461,
|
| 793 |
+
"num_tokens": 4058829.0,
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"entropy": 5.4763900089263915,
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
+
"grad_norm": 1.0856804847717285,
|
| 800 |
+
"learning_rate": 1.1398939304655273e-06,
|
| 801 |
+
"loss": 5.3042,
|
| 802 |
+
"mean_token_accuracy": 0.31338351368904116,
|
| 803 |
+
"num_tokens": 4113326.0,
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
+
"entropy": 5.328452725410461,
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
+
"grad_norm": 3.2843880653381348,
|
| 810 |
+
"learning_rate": 1.1281084266352386e-06,
|
| 811 |
+
"loss": 5.1624,
|
| 812 |
+
"mean_token_accuracy": 0.3305218696594238,
|
| 813 |
+
"num_tokens": 4165454.0,
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"entropy": 5.383157343864441,
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
+
"grad_norm": 2.207082748413086,
|
| 820 |
+
"learning_rate": 1.1163229228049497e-06,
|
| 821 |
+
"loss": 5.2163,
|
| 822 |
+
"mean_token_accuracy": 0.32331310987472534,
|
| 823 |
+
"num_tokens": 4219250.0,
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
+
"entropy": 5.585261764526368,
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
+
"grad_norm": 2.7102835178375244,
|
| 830 |
+
"learning_rate": 1.104537418974661e-06,
|
| 831 |
+
"loss": 5.4137,
|
| 832 |
+
"mean_token_accuracy": 0.29959124475717547,
|
| 833 |
+
"num_tokens": 4274711.0,
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
+
"entropy": 5.434073266983032,
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
+
"grad_norm": 1.3775779008865356,
|
| 840 |
+
"learning_rate": 1.0927519151443724e-06,
|
| 841 |
+
"loss": 5.2644,
|
| 842 |
+
"mean_token_accuracy": 0.3175011593103409,
|
| 843 |
+
"num_tokens": 4328616.0,
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
+
"entropy": 5.462391858100891,
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
+
"grad_norm": 1.4101024866104126,
|
| 850 |
+
"learning_rate": 1.0809664113140838e-06,
|
| 851 |
+
"loss": 5.2924,
|
| 852 |
+
"mean_token_accuracy": 0.3137941011786461,
|
| 853 |
+
"num_tokens": 4382416.0,
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
+
"entropy": 5.529892563819885,
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
+
"grad_norm": 1.2311837673187256,
|
| 860 |
+
"learning_rate": 1.0691809074837949e-06,
|
| 861 |
+
"loss": 5.364,
|
| 862 |
+
"mean_token_accuracy": 0.3046491605043411,
|
| 863 |
+
"num_tokens": 4437848.0,
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
+
"entropy": 5.4370484542846675,
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
+
"grad_norm": 1.0929864645004272,
|
| 870 |
+
"learning_rate": 1.0573954036535062e-06,
|
| 871 |
+
"loss": 5.2734,
|
| 872 |
+
"mean_token_accuracy": 0.3169013774394989,
|
| 873 |
+
"num_tokens": 4491185.0,
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
+
"entropy": 5.395377616882325,
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
+
"grad_norm": 1.5457273721694946,
|
| 880 |
+
"learning_rate": 1.0456098998232174e-06,
|
| 881 |
+
"loss": 5.2276,
|
| 882 |
+
"mean_token_accuracy": 0.32221508473157884,
|
| 883 |
+
"num_tokens": 4544086.0,
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"entropy": 5.443737335205078,
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
+
"grad_norm": 1.4844346046447754,
|
| 890 |
+
"learning_rate": 1.0338243959929287e-06,
|
| 891 |
+
"loss": 5.2786,
|
| 892 |
+
"mean_token_accuracy": 0.3157751387357712,
|
| 893 |
+
"num_tokens": 4597677.0,
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
+
"entropy": 5.419876251220703,
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
+
"grad_norm": 1.2481963634490967,
|
| 900 |
+
"learning_rate": 1.02203889216264e-06,
|
| 901 |
+
"loss": 5.2564,
|
| 902 |
+
"mean_token_accuracy": 0.31889803290367125,
|
| 903 |
+
"num_tokens": 4651343.0,
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
+
"entropy": 5.578677978515625,
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
+
"grad_norm": 2.0005414485931396,
|
| 910 |
+
"learning_rate": 1.0102533883323512e-06,
|
| 911 |
+
"loss": 5.4145,
|
| 912 |
+
"mean_token_accuracy": 0.30037090003490446,
|
| 913 |
+
"num_tokens": 4705985.0,
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
+
"entropy": 5.279946126937866,
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
+
"grad_norm": 1.080521821975708,
|
| 920 |
+
"learning_rate": 9.984678845020625e-07,
|
| 921 |
+
"loss": 5.1226,
|
| 922 |
+
"mean_token_accuracy": 0.3341303279995918,
|
| 923 |
+
"num_tokens": 4757741.0,
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"entropy": 5.551463279724121,
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
+
"grad_norm": 1.28898024559021,
|
| 930 |
+
"learning_rate": 9.866823806717736e-07,
|
| 931 |
+
"loss": 5.3832,
|
| 932 |
+
"mean_token_accuracy": 0.3028248634934425,
|
| 933 |
+
"num_tokens": 4812808.0,
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
+
"entropy": 5.3787487554550175,
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
+
"grad_norm": 1.5697983503341675,
|
| 940 |
+
"learning_rate": 9.74896876841485e-07,
|
| 941 |
+
"loss": 5.2141,
|
| 942 |
+
"mean_token_accuracy": 0.3227942296862602,
|
| 943 |
+
"num_tokens": 4866572.0,
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
+
"entropy": 5.460358958244324,
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
+
"grad_norm": 1.3180441856384277,
|
| 950 |
+
"learning_rate": 9.63111373011196e-07,
|
| 951 |
+
"loss": 5.2954,
|
| 952 |
+
"mean_token_accuracy": 0.31269474506378175,
|
| 953 |
+
"num_tokens": 4921312.0,
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"entropy": 5.434084935188293,
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
+
"grad_norm": 1.2409590482711792,
|
| 960 |
+
"learning_rate": 9.513258691809074e-07,
|
| 961 |
+
"loss": 5.271,
|
| 962 |
+
"mean_token_accuracy": 0.3172155100107193,
|
| 963 |
+
"num_tokens": 4974289.0,
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
+
"entropy": 5.406955418586731,
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
+
"grad_norm": 1.4782609939575195,
|
| 970 |
+
"learning_rate": 9.395403653506187e-07,
|
| 971 |
+
"loss": 5.2473,
|
| 972 |
+
"mean_token_accuracy": 0.32031788885593415,
|
| 973 |
+
"num_tokens": 5028149.0,
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
+
"entropy": 5.206603040695191,
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
+
"grad_norm": 2.351633071899414,
|
| 980 |
+
"learning_rate": 9.2775486152033e-07,
|
| 981 |
+
"loss": 5.0478,
|
| 982 |
+
"mean_token_accuracy": 0.3428420132398605,
|
| 983 |
+
"num_tokens": 5079349.0,
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
+
"entropy": 5.388812799453735,
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
+
"grad_norm": 7.564618110656738,
|
| 990 |
+
"learning_rate": 9.159693576900412e-07,
|
| 991 |
+
"loss": 5.2281,
|
| 992 |
+
"mean_token_accuracy": 0.3222071170806885,
|
| 993 |
+
"num_tokens": 5132564.0,
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"entropy": 5.374106278419495,
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
+
"grad_norm": 1.4734679460525513,
|
| 1000 |
+
"learning_rate": 9.041838538597525e-07,
|
| 1001 |
+
"loss": 5.2161,
|
| 1002 |
+
"mean_token_accuracy": 0.3219477406144142,
|
| 1003 |
+
"num_tokens": 5185921.0,
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
+
"entropy": 5.232998585700988,
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
+
"grad_norm": 1.4175471067428589,
|
| 1010 |
+
"learning_rate": 8.923983500294637e-07,
|
| 1011 |
+
"loss": 5.0769,
|
| 1012 |
+
"mean_token_accuracy": 0.3403926733136177,
|
| 1013 |
+
"num_tokens": 5237521.0,
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
+
"entropy": 5.394891719818116,
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
+
"grad_norm": 4.951873779296875,
|
| 1020 |
+
"learning_rate": 8.806128461991749e-07,
|
| 1021 |
+
"loss": 5.2344,
|
| 1022 |
+
"mean_token_accuracy": 0.3213117456436157,
|
| 1023 |
+
"num_tokens": 5291104.0,
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"entropy": 5.413805012702942,
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
+
"grad_norm": 1.679518461227417,
|
| 1030 |
+
"learning_rate": 8.688273423688863e-07,
|
| 1031 |
+
"loss": 5.2597,
|
| 1032 |
+
"mean_token_accuracy": 0.3165634173154831,
|
| 1033 |
+
"num_tokens": 5345058.0,
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
+
"entropy": 5.256177935600281,
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
+
"grad_norm": 1.8892916440963745,
|
| 1040 |
+
"learning_rate": 8.570418385385975e-07,
|
| 1041 |
+
"loss": 5.1004,
|
| 1042 |
+
"mean_token_accuracy": 0.3369427987933159,
|
| 1043 |
+
"num_tokens": 5395918.0,
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
+
"entropy": 5.259814453125,
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
+
"grad_norm": 1.3802675008773804,
|
| 1050 |
+
"learning_rate": 8.452563347083087e-07,
|
| 1051 |
+
"loss": 5.1057,
|
| 1052 |
+
"mean_token_accuracy": 0.3362414276599884,
|
| 1053 |
+
"num_tokens": 5448086.0,
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
+
"entropy": 5.416206178665161,
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
+
"grad_norm": 1.7677236795425415,
|
| 1060 |
+
"learning_rate": 8.3347083087802e-07,
|
| 1061 |
+
"loss": 5.2562,
|
| 1062 |
+
"mean_token_accuracy": 0.31725785195827483,
|
| 1063 |
+
"num_tokens": 5501959.0,
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
+
"entropy": 5.507337794303894,
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
+
"grad_norm": 1.021727442741394,
|
| 1070 |
+
"learning_rate": 8.216853270477313e-07,
|
| 1071 |
+
"loss": 5.344,
|
| 1072 |
+
"mean_token_accuracy": 0.30679062128067014,
|
| 1073 |
+
"num_tokens": 5557908.0,
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
+
"eval_entropy": 5.682707933786278,
|
| 1079 |
+
"eval_loss": 5.53223991394043,
|
| 1080 |
+
"eval_mean_token_accuracy": 0.27747743456594404,
|
| 1081 |
+
"eval_model_preparation_time": 0.0045,
|
| 1082 |
+
"eval_num_tokens": 5569086.0,
|
| 1083 |
+
"eval_runtime": 49.9944,
|
| 1084 |
+
"eval_samples_per_second": 8.681,
|
| 1085 |
+
"eval_steps_per_second": 4.34,
|
| 1086 |
"step": 5211
|
| 1087 |
}
|
| 1088 |
],
|
| 1089 |
"logging_steps": 50,
|
| 1090 |
+
"max_steps": 8685,
|
| 1091 |
"num_input_tokens_seen": 0,
|
| 1092 |
+
"num_train_epochs": 5,
|
| 1093 |
"save_steps": 500,
|
| 1094 |
"stateful_callbacks": {
|
| 1095 |
"TrainerControl": {
|
|
|
|
| 1103 |
"attributes": {}
|
| 1104 |
}
|
| 1105 |
},
|
| 1106 |
+
"total_flos": 7.632397197691392e+16,
|
| 1107 |
"train_batch_size": 2,
|
| 1108 |
"trial_name": null,
|
| 1109 |
"trial_params": null
|
checkpoint-5211/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|
checkpoint-6948/adapter_config.json
CHANGED
|
@@ -29,10 +29,10 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
-
"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
checkpoint-6948/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26182176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad17a7c29bcc91941e8e904522c1e1408f363f45d397b0f1a5a0c57829008c18
|
| 3 |
size 26182176
|
checkpoint-6948/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 52486155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad7e1547dc98180a197d1e09ff5462cc7435be478f2eadfbb7a35f3fe318cbac
|
| 3 |
size 52486155
|
checkpoint-6948/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8368c41a01c88b53eacde3119bcf65f0f4d5b3c36a14adcf04f08f24495c404
|
| 3 |
size 14645
|
checkpoint-6948/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6fc55ff863a5856e69cbaf9eb3a96203d5bcf04c7648e579610743cc43b484f9
|
| 3 |
size 1465
|
checkpoint-6948/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 6948,
|
| 3 |
-
"best_metric": 5.
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-6948",
|
| 5 |
"epoch": 4.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,1438 +10,1438 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"entropy": 3.
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
-
"grad_norm": 3.
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
-
"loss": 13.
|
| 18 |
-
"mean_token_accuracy": 0.
|
| 19 |
-
"num_tokens":
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"entropy": 3.
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
-
"grad_norm": 3.
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
-
"loss": 14.
|
| 28 |
-
"mean_token_accuracy": 0.
|
| 29 |
-
"num_tokens":
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"entropy": 3.
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
-
"grad_norm": 3.
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
-
"loss":
|
| 38 |
-
"mean_token_accuracy": 0.
|
| 39 |
-
"num_tokens":
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"entropy": 3.
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
-
"grad_norm": 4.
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
-
"loss": 13.
|
| 48 |
-
"mean_token_accuracy": 0.
|
| 49 |
-
"num_tokens":
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"entropy": 3.
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
-
"grad_norm": 4.
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
-
"mean_token_accuracy": 0.
|
| 59 |
-
"num_tokens":
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"entropy":
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
-
"grad_norm": 4.
|
| 66 |
-
"learning_rate": 1.
|
| 67 |
-
"loss": 12.
|
| 68 |
-
"mean_token_accuracy": 0.
|
| 69 |
-
"num_tokens":
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
-
"entropy": 4.
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate": 1.
|
| 77 |
-
"loss": 12.
|
| 78 |
-
"mean_token_accuracy": 0.
|
| 79 |
-
"num_tokens":
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"entropy": 4.
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 1.
|
| 87 |
-
"loss": 11.
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"entropy":
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate": 1.
|
| 97 |
-
"loss":
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"entropy": 5.
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 1.
|
| 107 |
-
"loss": 9.
|
| 108 |
-
"mean_token_accuracy": 0.
|
| 109 |
-
"num_tokens":
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"entropy": 6.
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate": 1.
|
| 117 |
-
"loss":
|
| 118 |
-
"mean_token_accuracy": 0.
|
| 119 |
-
"num_tokens":
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"entropy": 6.
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 1.
|
| 127 |
-
"loss": 7.
|
| 128 |
-
"mean_token_accuracy": 0.
|
| 129 |
-
"num_tokens":
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"entropy": 6.
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 1.
|
| 137 |
-
"loss": 6.
|
| 138 |
-
"mean_token_accuracy": 0.
|
| 139 |
-
"num_tokens":
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"entropy": 6.
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
-
"grad_norm": 4.
|
| 146 |
-
"learning_rate": 1.
|
| 147 |
-
"loss": 6.
|
| 148 |
-
"mean_token_accuracy": 0.
|
| 149 |
-
"num_tokens":
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"entropy": 6.
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 6.
|
| 158 |
-
"mean_token_accuracy": 0.
|
| 159 |
-
"num_tokens":
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"entropy": 6.
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
-
"grad_norm":
|
| 166 |
-
"learning_rate": 1.
|
| 167 |
-
"loss": 6.
|
| 168 |
-
"mean_token_accuracy": 0.
|
| 169 |
-
"num_tokens":
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"entropy": 6.
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
-
"grad_norm": 3.
|
| 176 |
-
"learning_rate": 1.
|
| 177 |
-
"loss": 6.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"entropy": 6.
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 1.
|
| 187 |
-
"loss": 6.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"entropy": 6.
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
-
"grad_norm": 2.
|
| 196 |
-
"learning_rate": 1.
|
| 197 |
-
"loss": 6.
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"entropy": 6.
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
-
"grad_norm": 2.
|
| 206 |
-
"learning_rate": 1.
|
| 207 |
-
"loss": 6.
|
| 208 |
-
"mean_token_accuracy": 0.
|
| 209 |
-
"num_tokens":
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"entropy": 6.
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
-
"grad_norm": 2.
|
| 216 |
-
"learning_rate": 1.
|
| 217 |
-
"loss": 6.
|
| 218 |
-
"mean_token_accuracy": 0.
|
| 219 |
-
"num_tokens":
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"entropy": 6.
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 1.
|
| 227 |
-
"loss": 5.
|
| 228 |
-
"mean_token_accuracy": 0.
|
| 229 |
-
"num_tokens":
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"entropy": 5.
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate": 1.
|
| 237 |
-
"loss": 5.
|
| 238 |
-
"mean_token_accuracy": 0.
|
| 239 |
-
"num_tokens":
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"entropy": 6.
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 1.
|
| 247 |
-
"loss": 5.
|
| 248 |
-
"mean_token_accuracy": 0.
|
| 249 |
-
"num_tokens":
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"entropy": 6.
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
-
"grad_norm": 3.
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 5.
|
| 258 |
-
"mean_token_accuracy": 0.
|
| 259 |
-
"num_tokens":
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"entropy": 6.
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
-
"grad_norm": 2.
|
| 266 |
-
"learning_rate": 1.
|
| 267 |
-
"loss": 5.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"entropy": 5.
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
-
"grad_norm": 2.
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 5.
|
| 278 |
-
"mean_token_accuracy": 0.
|
| 279 |
-
"num_tokens":
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"entropy": 6.
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
-
"grad_norm": 3.
|
| 286 |
-
"learning_rate": 1.
|
| 287 |
-
"loss": 5.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"entropy":
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 1.
|
| 297 |
-
"loss": 5.
|
| 298 |
-
"mean_token_accuracy": 0.
|
| 299 |
-
"num_tokens":
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"entropy":
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 5.
|
| 308 |
-
"mean_token_accuracy": 0.
|
| 309 |
-
"num_tokens":
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"entropy":
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 1.
|
| 317 |
-
"loss": 5.
|
| 318 |
-
"mean_token_accuracy": 0.
|
| 319 |
-
"num_tokens":
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"entropy": 5.
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
-
"grad_norm": 2.
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 5.
|
| 328 |
-
"mean_token_accuracy": 0.
|
| 329 |
-
"num_tokens":
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"entropy":
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
-
"grad_norm": 2.
|
| 336 |
-
"learning_rate": 1.
|
| 337 |
-
"loss": 5.
|
| 338 |
-
"mean_token_accuracy": 0.
|
| 339 |
-
"num_tokens":
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"entropy": 5.
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 1.
|
| 347 |
-
"loss": 5.
|
| 348 |
-
"mean_token_accuracy": 0.
|
| 349 |
-
"num_tokens":
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
-
"eval_entropy":
|
| 355 |
-
"eval_loss": 5.
|
| 356 |
-
"eval_mean_token_accuracy": 0.
|
| 357 |
-
"eval_model_preparation_time": 0.
|
| 358 |
-
"eval_num_tokens":
|
| 359 |
-
"eval_runtime":
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second":
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"entropy": 5.
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
-
"grad_norm": 2.
|
| 368 |
-
"learning_rate": 1.
|
| 369 |
-
"loss": 5.
|
| 370 |
-
"mean_token_accuracy": 0.
|
| 371 |
-
"num_tokens":
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"entropy":
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 1.
|
| 379 |
-
"loss": 5.
|
| 380 |
-
"mean_token_accuracy": 0.
|
| 381 |
-
"num_tokens":
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"entropy": 5.
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 1.
|
| 389 |
-
"loss": 5.
|
| 390 |
-
"mean_token_accuracy": 0.
|
| 391 |
-
"num_tokens":
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"entropy": 5.
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
-
"grad_norm":
|
| 398 |
-
"learning_rate": 1.
|
| 399 |
-
"loss": 5.
|
| 400 |
-
"mean_token_accuracy": 0.
|
| 401 |
-
"num_tokens":
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"entropy": 5.
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 1.
|
| 409 |
-
"loss": 5.
|
| 410 |
-
"mean_token_accuracy": 0.
|
| 411 |
-
"num_tokens":
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"entropy": 5.
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
-
"grad_norm":
|
| 418 |
-
"learning_rate": 1.
|
| 419 |
-
"loss": 5.
|
| 420 |
-
"mean_token_accuracy": 0.
|
| 421 |
-
"num_tokens":
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"entropy": 5.
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 1.
|
| 429 |
-
"loss": 5.
|
| 430 |
-
"mean_token_accuracy": 0.
|
| 431 |
-
"num_tokens":
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"entropy": 5.
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss": 5.
|
| 440 |
-
"mean_token_accuracy": 0.
|
| 441 |
-
"num_tokens":
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
-
"entropy": 5.
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
-
"grad_norm": 2.
|
| 448 |
-
"learning_rate": 1.
|
| 449 |
-
"loss": 5.
|
| 450 |
-
"mean_token_accuracy": 0.
|
| 451 |
-
"num_tokens":
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"entropy": 5.
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
-
"grad_norm":
|
| 458 |
-
"learning_rate": 1.
|
| 459 |
-
"loss": 5.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
-
"entropy": 5.
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
-
"grad_norm":
|
| 468 |
-
"learning_rate": 1.
|
| 469 |
-
"loss": 5.
|
| 470 |
-
"mean_token_accuracy": 0.
|
| 471 |
-
"num_tokens":
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"entropy": 5.
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 1.
|
| 479 |
-
"loss": 5.
|
| 480 |
-
"mean_token_accuracy": 0.
|
| 481 |
-
"num_tokens":
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"entropy": 5.
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
-
"grad_norm":
|
| 488 |
-
"learning_rate": 1.
|
| 489 |
-
"loss": 5.
|
| 490 |
-
"mean_token_accuracy": 0.
|
| 491 |
-
"num_tokens":
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"entropy": 5.
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 1.
|
| 499 |
-
"loss": 5.
|
| 500 |
-
"mean_token_accuracy": 0.
|
| 501 |
-
"num_tokens":
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"entropy": 5.
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate": 1.
|
| 509 |
-
"loss": 5.
|
| 510 |
-
"mean_token_accuracy": 0.
|
| 511 |
-
"num_tokens":
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"entropy": 5.
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
-
"grad_norm": 1.
|
| 518 |
-
"learning_rate": 1.
|
| 519 |
-
"loss": 5.
|
| 520 |
-
"mean_token_accuracy": 0.
|
| 521 |
-
"num_tokens":
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"entropy": 5.
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
-
"grad_norm": 3.
|
| 528 |
-
"learning_rate": 1.
|
| 529 |
-
"loss": 5.
|
| 530 |
-
"mean_token_accuracy": 0.
|
| 531 |
-
"num_tokens":
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"entropy": 5.
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
-
"grad_norm": 1.
|
| 538 |
-
"learning_rate": 1.
|
| 539 |
-
"loss": 5.
|
| 540 |
-
"mean_token_accuracy": 0.
|
| 541 |
-
"num_tokens":
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"entropy": 5.
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
-
"grad_norm":
|
| 548 |
-
"learning_rate": 1.
|
| 549 |
-
"loss": 5.
|
| 550 |
-
"mean_token_accuracy": 0.
|
| 551 |
-
"num_tokens":
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"entropy": 5.
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
-
"grad_norm":
|
| 558 |
-
"learning_rate": 1.
|
| 559 |
-
"loss": 5.
|
| 560 |
-
"mean_token_accuracy": 0.
|
| 561 |
-
"num_tokens":
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"entropy": 5.
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
-
"grad_norm": 2.
|
| 568 |
-
"learning_rate": 1.
|
| 569 |
-
"loss": 5.
|
| 570 |
-
"mean_token_accuracy": 0.
|
| 571 |
-
"num_tokens":
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"entropy": 5.
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate": 1.
|
| 579 |
-
"loss": 5.
|
| 580 |
-
"mean_token_accuracy": 0.
|
| 581 |
-
"num_tokens":
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"entropy": 5.
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
-
"grad_norm": 1.
|
| 588 |
-
"learning_rate": 1.
|
| 589 |
-
"loss": 5.
|
| 590 |
-
"mean_token_accuracy": 0.
|
| 591 |
-
"num_tokens":
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
-
"entropy": 5.
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
-
"grad_norm": 1.
|
| 598 |
-
"learning_rate": 1.
|
| 599 |
-
"loss": 5.
|
| 600 |
-
"mean_token_accuracy": 0.
|
| 601 |
-
"num_tokens":
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"entropy": 5.
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
-
"grad_norm": 1.
|
| 608 |
-
"learning_rate": 1.
|
| 609 |
-
"loss": 5.
|
| 610 |
-
"mean_token_accuracy": 0.
|
| 611 |
-
"num_tokens":
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"entropy": 5.
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
-
"grad_norm":
|
| 618 |
-
"learning_rate": 1.
|
| 619 |
-
"loss": 5.
|
| 620 |
-
"mean_token_accuracy": 0.
|
| 621 |
-
"num_tokens":
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
-
"entropy": 5.
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
-
"grad_norm": 1.
|
| 628 |
-
"learning_rate": 1.
|
| 629 |
-
"loss": 5.
|
| 630 |
-
"mean_token_accuracy": 0.
|
| 631 |
-
"num_tokens":
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"entropy": 5.
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate": 1.
|
| 639 |
-
"loss": 5.
|
| 640 |
-
"mean_token_accuracy": 0.
|
| 641 |
-
"num_tokens":
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"entropy": 5.
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate": 1.
|
| 649 |
-
"loss":
|
| 650 |
-
"mean_token_accuracy": 0.
|
| 651 |
-
"num_tokens":
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
-
"entropy": 5.
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
-
"grad_norm": 1.
|
| 658 |
-
"learning_rate": 1.
|
| 659 |
-
"loss": 5.
|
| 660 |
-
"mean_token_accuracy": 0.
|
| 661 |
-
"num_tokens":
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"entropy": 5.
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
-
"grad_norm": 1.
|
| 668 |
-
"learning_rate": 1.
|
| 669 |
-
"loss": 5.
|
| 670 |
-
"mean_token_accuracy": 0.
|
| 671 |
-
"num_tokens":
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"entropy": 5.
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
-
"grad_norm":
|
| 678 |
-
"learning_rate": 1.
|
| 679 |
-
"loss": 5.
|
| 680 |
-
"mean_token_accuracy": 0.
|
| 681 |
-
"num_tokens":
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"entropy": 5.
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
-
"grad_norm":
|
| 688 |
-
"learning_rate": 1.
|
| 689 |
-
"loss": 5.
|
| 690 |
-
"mean_token_accuracy": 0.
|
| 691 |
-
"num_tokens":
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"entropy": 5.
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
-
"grad_norm": 1.
|
| 698 |
-
"learning_rate": 1.
|
| 699 |
-
"loss": 5.
|
| 700 |
-
"mean_token_accuracy": 0.
|
| 701 |
-
"num_tokens":
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"entropy": 5.
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
-
"grad_norm": 1.
|
| 708 |
-
"learning_rate": 1.
|
| 709 |
-
"loss": 5.
|
| 710 |
-
"mean_token_accuracy": 0.
|
| 711 |
-
"num_tokens":
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
-
"eval_entropy": 5.
|
| 717 |
-
"eval_loss": 5.
|
| 718 |
-
"eval_mean_token_accuracy": 0.
|
| 719 |
-
"eval_model_preparation_time": 0.
|
| 720 |
-
"eval_num_tokens":
|
| 721 |
-
"eval_runtime":
|
| 722 |
-
"eval_samples_per_second":
|
| 723 |
-
"eval_steps_per_second":
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"entropy": 5.
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
-
"grad_norm":
|
| 730 |
-
"learning_rate": 1.
|
| 731 |
-
"loss": 5.
|
| 732 |
-
"mean_token_accuracy": 0.
|
| 733 |
-
"num_tokens":
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"entropy": 5.
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
-
"grad_norm": 1.
|
| 740 |
-
"learning_rate": 1.
|
| 741 |
-
"loss": 5.
|
| 742 |
-
"mean_token_accuracy": 0.
|
| 743 |
-
"num_tokens":
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"entropy": 5.
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss": 5.
|
| 752 |
-
"mean_token_accuracy": 0.
|
| 753 |
-
"num_tokens":
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"entropy": 5.
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate":
|
| 761 |
-
"loss": 5.
|
| 762 |
-
"mean_token_accuracy": 0.
|
| 763 |
-
"num_tokens":
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"entropy": 5.
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 5.
|
| 772 |
-
"mean_token_accuracy": 0.
|
| 773 |
-
"num_tokens":
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"entropy": 5.
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
-
"grad_norm": 2.
|
| 780 |
-
"learning_rate":
|
| 781 |
-
"loss": 5.
|
| 782 |
-
"mean_token_accuracy": 0.
|
| 783 |
-
"num_tokens":
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
-
"entropy": 5.
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
-
"grad_norm": 1.
|
| 790 |
-
"learning_rate":
|
| 791 |
-
"loss": 5.
|
| 792 |
-
"mean_token_accuracy": 0.
|
| 793 |
-
"num_tokens":
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"entropy": 5.
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
-
"grad_norm":
|
| 800 |
-
"learning_rate":
|
| 801 |
-
"loss": 5.
|
| 802 |
-
"mean_token_accuracy": 0.
|
| 803 |
-
"num_tokens":
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
-
"entropy": 5.
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
-
"grad_norm":
|
| 810 |
-
"learning_rate":
|
| 811 |
-
"loss": 5.
|
| 812 |
-
"mean_token_accuracy": 0.
|
| 813 |
-
"num_tokens":
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"entropy": 5.
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 5.
|
| 822 |
-
"mean_token_accuracy": 0.
|
| 823 |
-
"num_tokens":
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
-
"entropy": 5.
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
-
"grad_norm": 2.
|
| 830 |
-
"learning_rate":
|
| 831 |
-
"loss": 5.
|
| 832 |
-
"mean_token_accuracy": 0.
|
| 833 |
-
"num_tokens":
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
-
"entropy": 5.
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
-
"grad_norm": 1.
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 5.
|
| 842 |
-
"mean_token_accuracy": 0.
|
| 843 |
-
"num_tokens":
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
-
"entropy": 5.
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
-
"grad_norm": 1.
|
| 850 |
-
"learning_rate":
|
| 851 |
-
"loss": 5.
|
| 852 |
-
"mean_token_accuracy": 0.
|
| 853 |
-
"num_tokens":
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
-
"entropy": 5.
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
-
"grad_norm": 1.
|
| 860 |
-
"learning_rate":
|
| 861 |
-
"loss": 5.
|
| 862 |
-
"mean_token_accuracy": 0.
|
| 863 |
-
"num_tokens":
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"entropy": 5.
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
-
"grad_norm": 1.
|
| 870 |
-
"learning_rate":
|
| 871 |
-
"loss": 5.
|
| 872 |
-
"mean_token_accuracy": 0.
|
| 873 |
-
"num_tokens":
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
-
"entropy": 5.
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
-
"grad_norm": 1.
|
| 880 |
-
"learning_rate":
|
| 881 |
-
"loss": 5.
|
| 882 |
-
"mean_token_accuracy": 0.
|
| 883 |
-
"num_tokens":
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"entropy": 5.
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
-
"grad_norm": 1.
|
| 890 |
-
"learning_rate":
|
| 891 |
-
"loss": 5.
|
| 892 |
-
"mean_token_accuracy": 0.
|
| 893 |
-
"num_tokens":
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
-
"entropy": 5.
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
-
"grad_norm":
|
| 900 |
-
"learning_rate":
|
| 901 |
-
"loss": 5.
|
| 902 |
-
"mean_token_accuracy": 0.
|
| 903 |
-
"num_tokens":
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
-
"entropy": 5.
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 5.
|
| 912 |
-
"mean_token_accuracy": 0.
|
| 913 |
-
"num_tokens":
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
-
"entropy": 5.
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
-
"grad_norm": 1.
|
| 920 |
-
"learning_rate":
|
| 921 |
-
"loss": 5.
|
| 922 |
-
"mean_token_accuracy": 0.
|
| 923 |
-
"num_tokens":
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"entropy": 5.
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
-
"grad_norm": 1.
|
| 930 |
-
"learning_rate":
|
| 931 |
-
"loss": 5.
|
| 932 |
-
"mean_token_accuracy": 0.
|
| 933 |
-
"num_tokens":
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
-
"entropy": 5.
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
-
"grad_norm": 1.
|
| 940 |
-
"learning_rate":
|
| 941 |
-
"loss": 5.
|
| 942 |
-
"mean_token_accuracy": 0.
|
| 943 |
-
"num_tokens":
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
-
"entropy": 5.
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
-
"grad_norm": 1.
|
| 950 |
-
"learning_rate":
|
| 951 |
-
"loss": 5.
|
| 952 |
-
"mean_token_accuracy": 0.
|
| 953 |
-
"num_tokens":
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"entropy": 5.
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
-
"grad_norm":
|
| 960 |
-
"learning_rate":
|
| 961 |
-
"loss": 5.
|
| 962 |
-
"mean_token_accuracy": 0.
|
| 963 |
-
"num_tokens":
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
-
"entropy": 5.
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
-
"grad_norm":
|
| 970 |
-
"learning_rate":
|
| 971 |
-
"loss": 5.
|
| 972 |
-
"mean_token_accuracy": 0.
|
| 973 |
-
"num_tokens":
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"entropy": 5.
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
-
"grad_norm":
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 5.
|
| 982 |
-
"mean_token_accuracy": 0.
|
| 983 |
-
"num_tokens":
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
-
"entropy": 5.
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
-
"grad_norm":
|
| 990 |
-
"learning_rate":
|
| 991 |
-
"loss": 5.
|
| 992 |
-
"mean_token_accuracy": 0.
|
| 993 |
-
"num_tokens":
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
-
"entropy": 5.
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
-
"grad_norm": 1.
|
| 1000 |
-
"learning_rate":
|
| 1001 |
-
"loss": 5.
|
| 1002 |
-
"mean_token_accuracy": 0.
|
| 1003 |
-
"num_tokens":
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"entropy": 5.
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
-
"grad_norm": 1.
|
| 1010 |
-
"learning_rate":
|
| 1011 |
-
"loss": 5.
|
| 1012 |
-
"mean_token_accuracy": 0.
|
| 1013 |
-
"num_tokens":
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
-
"entropy": 5.
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
-
"grad_norm":
|
| 1020 |
-
"learning_rate":
|
| 1021 |
-
"loss": 5.
|
| 1022 |
-
"mean_token_accuracy": 0.
|
| 1023 |
-
"num_tokens":
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"entropy": 5.
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
-
"grad_norm": 1.
|
| 1030 |
-
"learning_rate":
|
| 1031 |
-
"loss": 5.
|
| 1032 |
-
"mean_token_accuracy": 0.
|
| 1033 |
-
"num_tokens":
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
-
"entropy": 5.
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
-
"grad_norm": 1.
|
| 1040 |
-
"learning_rate":
|
| 1041 |
-
"loss": 5.
|
| 1042 |
-
"mean_token_accuracy": 0.
|
| 1043 |
-
"num_tokens":
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"entropy": 5.
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
-
"grad_norm": 1.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 5.
|
| 1052 |
-
"mean_token_accuracy": 0.
|
| 1053 |
-
"num_tokens":
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
-
"entropy": 5.
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
-
"grad_norm": 1.
|
| 1060 |
-
"learning_rate":
|
| 1061 |
-
"loss": 5.
|
| 1062 |
-
"mean_token_accuracy": 0.
|
| 1063 |
-
"num_tokens":
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
-
"entropy": 5.
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
-
"grad_norm": 1.
|
| 1070 |
-
"learning_rate":
|
| 1071 |
-
"loss": 5.
|
| 1072 |
-
"mean_token_accuracy": 0.
|
| 1073 |
-
"num_tokens":
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
-
"eval_entropy": 5.
|
| 1079 |
-
"eval_loss": 5.
|
| 1080 |
-
"eval_mean_token_accuracy": 0.
|
| 1081 |
-
"eval_model_preparation_time": 0.
|
| 1082 |
-
"eval_num_tokens":
|
| 1083 |
-
"eval_runtime":
|
| 1084 |
-
"eval_samples_per_second":
|
| 1085 |
-
"eval_steps_per_second":
|
| 1086 |
"step": 5211
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
-
"entropy": 5.
|
| 1090 |
"epoch": 3.0224525043177892,
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 5.
|
| 1094 |
-
"mean_token_accuracy": 0.
|
| 1095 |
-
"num_tokens":
|
| 1096 |
"step": 5250
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
-
"entropy": 5.
|
| 1100 |
"epoch": 3.051237766263673,
|
| 1101 |
-
"grad_norm":
|
| 1102 |
-
"learning_rate":
|
| 1103 |
-
"loss": 5.
|
| 1104 |
-
"mean_token_accuracy": 0.
|
| 1105 |
-
"num_tokens":
|
| 1106 |
"step": 5300
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
-
"entropy": 5.
|
| 1110 |
"epoch": 3.0800230282095566,
|
| 1111 |
-
"grad_norm": 2.
|
| 1112 |
-
"learning_rate":
|
| 1113 |
-
"loss": 5.
|
| 1114 |
-
"mean_token_accuracy": 0.
|
| 1115 |
-
"num_tokens":
|
| 1116 |
"step": 5350
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
-
"entropy": 5.
|
| 1120 |
"epoch": 3.1088082901554404,
|
| 1121 |
-
"grad_norm": 1.
|
| 1122 |
-
"learning_rate":
|
| 1123 |
-
"loss": 5.
|
| 1124 |
-
"mean_token_accuracy": 0.
|
| 1125 |
-
"num_tokens":
|
| 1126 |
"step": 5400
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
-
"entropy": 5.
|
| 1130 |
"epoch": 3.137593552101324,
|
| 1131 |
-
"grad_norm":
|
| 1132 |
-
"learning_rate":
|
| 1133 |
-
"loss": 5.
|
| 1134 |
-
"mean_token_accuracy": 0.
|
| 1135 |
-
"num_tokens":
|
| 1136 |
"step": 5450
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"entropy": 5.
|
| 1140 |
"epoch": 3.166378814047208,
|
| 1141 |
-
"grad_norm": 1.
|
| 1142 |
-
"learning_rate":
|
| 1143 |
-
"loss": 5.
|
| 1144 |
-
"mean_token_accuracy": 0.
|
| 1145 |
-
"num_tokens":
|
| 1146 |
"step": 5500
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
-
"entropy": 5.
|
| 1150 |
"epoch": 3.1951640759930915,
|
| 1151 |
-
"grad_norm": 1.
|
| 1152 |
-
"learning_rate":
|
| 1153 |
-
"loss": 5.
|
| 1154 |
-
"mean_token_accuracy": 0.
|
| 1155 |
-
"num_tokens":
|
| 1156 |
"step": 5550
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
-
"entropy": 5.
|
| 1160 |
"epoch": 3.223949337938975,
|
| 1161 |
-
"grad_norm":
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 5.
|
| 1164 |
-
"mean_token_accuracy": 0.
|
| 1165 |
-
"num_tokens":
|
| 1166 |
"step": 5600
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
-
"entropy": 5.
|
| 1170 |
"epoch": 3.252734599884859,
|
| 1171 |
-
"grad_norm":
|
| 1172 |
-
"learning_rate":
|
| 1173 |
-
"loss": 5.
|
| 1174 |
-
"mean_token_accuracy": 0.
|
| 1175 |
-
"num_tokens":
|
| 1176 |
"step": 5650
|
| 1177 |
},
|
| 1178 |
{
|
| 1179 |
-
"entropy": 5.
|
| 1180 |
"epoch": 3.2815198618307426,
|
| 1181 |
-
"grad_norm":
|
| 1182 |
-
"learning_rate":
|
| 1183 |
-
"loss": 5.
|
| 1184 |
-
"mean_token_accuracy": 0.
|
| 1185 |
-
"num_tokens":
|
| 1186 |
"step": 5700
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
-
"entropy": 5.
|
| 1190 |
"epoch": 3.3103051237766263,
|
| 1191 |
-
"grad_norm": 1.
|
| 1192 |
-
"learning_rate":
|
| 1193 |
-
"loss": 5.
|
| 1194 |
-
"mean_token_accuracy": 0.
|
| 1195 |
-
"num_tokens":
|
| 1196 |
"step": 5750
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
-
"entropy": 5.
|
| 1200 |
"epoch": 3.33909038572251,
|
| 1201 |
-
"grad_norm":
|
| 1202 |
-
"learning_rate":
|
| 1203 |
-
"loss": 5.
|
| 1204 |
-
"mean_token_accuracy": 0.
|
| 1205 |
-
"num_tokens":
|
| 1206 |
"step": 5800
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
-
"entropy": 5.
|
| 1210 |
"epoch": 3.3678756476683938,
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss":
|
| 1214 |
-
"mean_token_accuracy": 0.
|
| 1215 |
-
"num_tokens":
|
| 1216 |
"step": 5850
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
-
"entropy": 5.
|
| 1220 |
"epoch": 3.3966609096142775,
|
| 1221 |
-
"grad_norm": 1.
|
| 1222 |
-
"learning_rate":
|
| 1223 |
-
"loss": 5.
|
| 1224 |
-
"mean_token_accuracy": 0.
|
| 1225 |
-
"num_tokens":
|
| 1226 |
"step": 5900
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
-
"entropy": 5.
|
| 1230 |
"epoch": 3.425446171560161,
|
| 1231 |
-
"grad_norm":
|
| 1232 |
-
"learning_rate":
|
| 1233 |
-
"loss": 5.
|
| 1234 |
-
"mean_token_accuracy": 0.
|
| 1235 |
-
"num_tokens":
|
| 1236 |
"step": 5950
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
-
"entropy": 5.
|
| 1240 |
"epoch": 3.454231433506045,
|
| 1241 |
-
"grad_norm": 1.
|
| 1242 |
-
"learning_rate":
|
| 1243 |
-
"loss": 5.
|
| 1244 |
-
"mean_token_accuracy": 0.
|
| 1245 |
-
"num_tokens":
|
| 1246 |
"step": 6000
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
-
"entropy": 5.
|
| 1250 |
"epoch": 3.4830166954519286,
|
| 1251 |
-
"grad_norm": 1.
|
| 1252 |
-
"learning_rate":
|
| 1253 |
-
"loss": 5.
|
| 1254 |
-
"mean_token_accuracy": 0.
|
| 1255 |
-
"num_tokens":
|
| 1256 |
"step": 6050
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
-
"entropy": 5.
|
| 1260 |
"epoch": 3.5118019573978123,
|
| 1261 |
-
"grad_norm":
|
| 1262 |
-
"learning_rate":
|
| 1263 |
-
"loss": 5.
|
| 1264 |
-
"mean_token_accuracy": 0.
|
| 1265 |
-
"num_tokens":
|
| 1266 |
"step": 6100
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
-
"entropy": 5.
|
| 1270 |
"epoch": 3.540587219343696,
|
| 1271 |
-
"grad_norm": 1.
|
| 1272 |
-
"learning_rate":
|
| 1273 |
-
"loss": 5.
|
| 1274 |
-
"mean_token_accuracy": 0.
|
| 1275 |
-
"num_tokens":
|
| 1276 |
"step": 6150
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
-
"entropy": 5.
|
| 1280 |
"epoch": 3.5693724812895797,
|
| 1281 |
-
"grad_norm":
|
| 1282 |
-
"learning_rate":
|
| 1283 |
-
"loss": 5.
|
| 1284 |
-
"mean_token_accuracy": 0.
|
| 1285 |
-
"num_tokens":
|
| 1286 |
"step": 6200
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
-
"entropy": 5.
|
| 1290 |
"epoch": 3.5981577432354634,
|
| 1291 |
-
"grad_norm":
|
| 1292 |
-
"learning_rate":
|
| 1293 |
-
"loss": 5.
|
| 1294 |
-
"mean_token_accuracy": 0.
|
| 1295 |
-
"num_tokens":
|
| 1296 |
"step": 6250
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
-
"entropy": 5.
|
| 1300 |
"epoch": 3.626943005181347,
|
| 1301 |
-
"grad_norm":
|
| 1302 |
-
"learning_rate":
|
| 1303 |
-
"loss": 5.
|
| 1304 |
-
"mean_token_accuracy": 0.
|
| 1305 |
-
"num_tokens":
|
| 1306 |
"step": 6300
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
-
"entropy": 5.
|
| 1310 |
"epoch": 3.655728267127231,
|
| 1311 |
-
"grad_norm": 1.
|
| 1312 |
-
"learning_rate":
|
| 1313 |
-
"loss": 5.
|
| 1314 |
-
"mean_token_accuracy": 0.
|
| 1315 |
-
"num_tokens":
|
| 1316 |
"step": 6350
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
-
"entropy": 5.
|
| 1320 |
"epoch": 3.6845135290731146,
|
| 1321 |
-
"grad_norm":
|
| 1322 |
-
"learning_rate":
|
| 1323 |
-
"loss": 5.
|
| 1324 |
-
"mean_token_accuracy": 0.
|
| 1325 |
-
"num_tokens":
|
| 1326 |
"step": 6400
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
-
"entropy": 5.
|
| 1330 |
"epoch": 3.7132987910189983,
|
| 1331 |
-
"grad_norm": 1.
|
| 1332 |
-
"learning_rate":
|
| 1333 |
-
"loss": 5.
|
| 1334 |
-
"mean_token_accuracy": 0.
|
| 1335 |
-
"num_tokens":
|
| 1336 |
"step": 6450
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
-
"entropy": 5.
|
| 1340 |
"epoch": 3.742084052964882,
|
| 1341 |
-
"grad_norm": 1.
|
| 1342 |
-
"learning_rate":
|
| 1343 |
-
"loss": 5.
|
| 1344 |
-
"mean_token_accuracy": 0.
|
| 1345 |
-
"num_tokens":
|
| 1346 |
"step": 6500
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
-
"entropy": 5.
|
| 1350 |
"epoch": 3.7708693149107657,
|
| 1351 |
-
"grad_norm": 2.
|
| 1352 |
-
"learning_rate":
|
| 1353 |
-
"loss": 5.
|
| 1354 |
-
"mean_token_accuracy": 0.
|
| 1355 |
-
"num_tokens":
|
| 1356 |
"step": 6550
|
| 1357 |
},
|
| 1358 |
{
|
| 1359 |
-
"entropy": 5.
|
| 1360 |
"epoch": 3.7996545768566494,
|
| 1361 |
-
"grad_norm": 1.
|
| 1362 |
-
"learning_rate":
|
| 1363 |
-
"loss": 5.
|
| 1364 |
-
"mean_token_accuracy": 0.
|
| 1365 |
-
"num_tokens":
|
| 1366 |
"step": 6600
|
| 1367 |
},
|
| 1368 |
{
|
| 1369 |
-
"entropy": 5.
|
| 1370 |
"epoch": 3.828439838802533,
|
| 1371 |
-
"grad_norm":
|
| 1372 |
-
"learning_rate":
|
| 1373 |
-
"loss": 5.
|
| 1374 |
-
"mean_token_accuracy": 0.
|
| 1375 |
-
"num_tokens":
|
| 1376 |
"step": 6650
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
-
"entropy": 5.
|
| 1380 |
"epoch": 3.857225100748417,
|
| 1381 |
-
"grad_norm": 1.
|
| 1382 |
-
"learning_rate":
|
| 1383 |
-
"loss": 5.
|
| 1384 |
-
"mean_token_accuracy": 0.
|
| 1385 |
-
"num_tokens":
|
| 1386 |
"step": 6700
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
-
"entropy": 5.
|
| 1390 |
"epoch": 3.8860103626943006,
|
| 1391 |
-
"grad_norm":
|
| 1392 |
-
"learning_rate":
|
| 1393 |
-
"loss": 5.
|
| 1394 |
-
"mean_token_accuracy": 0.
|
| 1395 |
-
"num_tokens":
|
| 1396 |
"step": 6750
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
-
"entropy": 5.
|
| 1400 |
"epoch": 3.9147956246401843,
|
| 1401 |
-
"grad_norm": 1.
|
| 1402 |
-
"learning_rate": 4.
|
| 1403 |
-
"loss": 5.
|
| 1404 |
-
"mean_token_accuracy": 0.
|
| 1405 |
-
"num_tokens":
|
| 1406 |
"step": 6800
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
-
"entropy": 5.
|
| 1410 |
"epoch": 3.943580886586068,
|
| 1411 |
-
"grad_norm": 0.
|
| 1412 |
-
"learning_rate":
|
| 1413 |
-
"loss": 5.
|
| 1414 |
-
"mean_token_accuracy": 0.
|
| 1415 |
-
"num_tokens":
|
| 1416 |
"step": 6850
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
-
"entropy": 5.
|
| 1420 |
"epoch": 3.9723661485319517,
|
| 1421 |
-
"grad_norm":
|
| 1422 |
-
"learning_rate":
|
| 1423 |
-
"loss":
|
| 1424 |
-
"mean_token_accuracy": 0.
|
| 1425 |
-
"num_tokens":
|
| 1426 |
"step": 6900
|
| 1427 |
},
|
| 1428 |
{
|
| 1429 |
"epoch": 4.0,
|
| 1430 |
-
"eval_entropy": 5.
|
| 1431 |
-
"eval_loss": 5.
|
| 1432 |
-
"eval_mean_token_accuracy": 0.
|
| 1433 |
-
"eval_model_preparation_time": 0.
|
| 1434 |
-
"eval_num_tokens":
|
| 1435 |
-
"eval_runtime":
|
| 1436 |
-
"eval_samples_per_second":
|
| 1437 |
-
"eval_steps_per_second":
|
| 1438 |
"step": 6948
|
| 1439 |
}
|
| 1440 |
],
|
| 1441 |
"logging_steps": 50,
|
| 1442 |
-
"max_steps":
|
| 1443 |
"num_input_tokens_seen": 0,
|
| 1444 |
-
"num_train_epochs":
|
| 1445 |
"save_steps": 500,
|
| 1446 |
"stateful_callbacks": {
|
| 1447 |
"TrainerControl": {
|
|
@@ -1450,12 +1450,12 @@
|
|
| 1450 |
"should_evaluate": false,
|
| 1451 |
"should_log": false,
|
| 1452 |
"should_save": true,
|
| 1453 |
-
"should_training_stop":
|
| 1454 |
},
|
| 1455 |
"attributes": {}
|
| 1456 |
}
|
| 1457 |
},
|
| 1458 |
-
"total_flos": 1.
|
| 1459 |
"train_batch_size": 2,
|
| 1460 |
"trial_name": null,
|
| 1461 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 6948,
|
| 3 |
+
"best_metric": 5.525067329406738,
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-6948",
|
| 5 |
"epoch": 4.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"entropy": 3.606692385673523,
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
+
"grad_norm": 3.2999913692474365,
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
+
"loss": 13.6598,
|
| 18 |
+
"mean_token_accuracy": 0.16028020828962325,
|
| 19 |
+
"num_tokens": 53993.0,
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"entropy": 3.618675880432129,
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
+
"grad_norm": 3.101252555847168,
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
+
"loss": 14.0188,
|
| 28 |
+
"mean_token_accuracy": 0.1508466500043869,
|
| 29 |
+
"num_tokens": 110134.0,
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"entropy": 3.5215235900878907,
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
+
"grad_norm": 3.513662815093994,
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
+
"loss": 12.8555,
|
| 38 |
+
"mean_token_accuracy": 0.18527640983462335,
|
| 39 |
+
"num_tokens": 160191.0,
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"entropy": 3.667909698486328,
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
+
"grad_norm": 4.327610492706299,
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
+
"loss": 13.5394,
|
| 48 |
+
"mean_token_accuracy": 0.157139780074358,
|
| 49 |
+
"num_tokens": 214993.0,
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"entropy": 3.768263258934021,
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
+
"grad_norm": 4.290107250213623,
|
| 56 |
+
"learning_rate": 1.988450206246317e-06,
|
| 57 |
+
"loss": 12.8912,
|
| 58 |
+
"mean_token_accuracy": 0.17374794125556947,
|
| 59 |
+
"num_tokens": 268184.0,
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"entropy": 3.990619196891785,
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
+
"grad_norm": 4.444278717041016,
|
| 66 |
+
"learning_rate": 1.976664702416028e-06,
|
| 67 |
+
"loss": 12.455,
|
| 68 |
+
"mean_token_accuracy": 0.17780130118131637,
|
| 69 |
+
"num_tokens": 319458.0,
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"entropy": 4.162646284103394,
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
+
"grad_norm": 5.615262508392334,
|
| 76 |
+
"learning_rate": 1.9648791985857395e-06,
|
| 77 |
+
"loss": 12.0893,
|
| 78 |
+
"mean_token_accuracy": 0.18191319867968558,
|
| 79 |
+
"num_tokens": 373337.0,
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"entropy": 4.532100868225098,
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
+
"grad_norm": 10.074016571044922,
|
| 86 |
+
"learning_rate": 1.9530936947554507e-06,
|
| 87 |
+
"loss": 11.9261,
|
| 88 |
+
"mean_token_accuracy": 0.169477596282959,
|
| 89 |
+
"num_tokens": 427526.0,
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"entropy": 4.923871030807495,
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
+
"grad_norm": 16.220163345336914,
|
| 96 |
+
"learning_rate": 1.9413081909251622e-06,
|
| 97 |
+
"loss": 11.0048,
|
| 98 |
+
"mean_token_accuracy": 0.1704501649737358,
|
| 99 |
+
"num_tokens": 480528.0,
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"entropy": 5.521005854606629,
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
+
"grad_norm": 29.904008865356445,
|
| 106 |
+
"learning_rate": 1.9295226870948733e-06,
|
| 107 |
+
"loss": 9.6524,
|
| 108 |
+
"mean_token_accuracy": 0.16450899541378022,
|
| 109 |
+
"num_tokens": 535314.0,
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"entropy": 6.092623329162597,
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
+
"grad_norm": 17.821575164794922,
|
| 116 |
+
"learning_rate": 1.9177371832645845e-06,
|
| 117 |
+
"loss": 8.1054,
|
| 118 |
+
"mean_token_accuracy": 0.17205011785030366,
|
| 119 |
+
"num_tokens": 588410.0,
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"entropy": 6.385262680053711,
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
+
"grad_norm": 5.502202987670898,
|
| 126 |
+
"learning_rate": 1.9059516794342958e-06,
|
| 127 |
+
"loss": 7.4313,
|
| 128 |
+
"mean_token_accuracy": 0.1734227080643177,
|
| 129 |
+
"num_tokens": 641736.0,
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"entropy": 6.278562617301941,
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
+
"grad_norm": 5.4657697677612305,
|
| 136 |
+
"learning_rate": 1.8941661756040071e-06,
|
| 137 |
+
"loss": 6.9266,
|
| 138 |
+
"mean_token_accuracy": 0.18680249139666558,
|
| 139 |
+
"num_tokens": 692200.0,
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"entropy": 6.553266277313233,
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
+
"grad_norm": 4.955812931060791,
|
| 146 |
+
"learning_rate": 1.8823806717737183e-06,
|
| 147 |
+
"loss": 6.9847,
|
| 148 |
+
"mean_token_accuracy": 0.16679802387952805,
|
| 149 |
+
"num_tokens": 745830.0,
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"entropy": 6.470935583114624,
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
+
"grad_norm": 4.198381423950195,
|
| 156 |
+
"learning_rate": 1.8705951679434296e-06,
|
| 157 |
+
"loss": 6.7277,
|
| 158 |
+
"mean_token_accuracy": 0.17847734570503235,
|
| 159 |
+
"num_tokens": 798872.0,
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"entropy": 6.5620588779449465,
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
+
"grad_norm": 3.1793746948242188,
|
| 166 |
+
"learning_rate": 1.8588096641131407e-06,
|
| 167 |
+
"loss": 6.7032,
|
| 168 |
+
"mean_token_accuracy": 0.17336134731769562,
|
| 169 |
+
"num_tokens": 853045.0,
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"entropy": 6.532204885482788,
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
+
"grad_norm": 3.824537515640259,
|
| 176 |
+
"learning_rate": 1.847024160282852e-06,
|
| 177 |
+
"loss": 6.5762,
|
| 178 |
+
"mean_token_accuracy": 0.1805124071240425,
|
| 179 |
+
"num_tokens": 907679.0,
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"entropy": 6.535988225936889,
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
+
"grad_norm": 4.350001811981201,
|
| 186 |
+
"learning_rate": 1.8352386564525632e-06,
|
| 187 |
+
"loss": 6.505,
|
| 188 |
+
"mean_token_accuracy": 0.1842605724930763,
|
| 189 |
+
"num_tokens": 964170.0,
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"entropy": 6.204533562660218,
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
+
"grad_norm": 2.193660020828247,
|
| 196 |
+
"learning_rate": 1.8234531526222745e-06,
|
| 197 |
+
"loss": 6.1211,
|
| 198 |
+
"mean_token_accuracy": 0.21968430042266845,
|
| 199 |
+
"num_tokens": 1015909.0,
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"entropy": 6.308737449645996,
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
+
"grad_norm": 2.325622320175171,
|
| 206 |
+
"learning_rate": 1.8116676487919857e-06,
|
| 207 |
+
"loss": 6.1653,
|
| 208 |
+
"mean_token_accuracy": 0.21636426240205764,
|
| 209 |
+
"num_tokens": 1068859.0,
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"entropy": 6.332560749053955,
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
+
"grad_norm": 2.0439090728759766,
|
| 216 |
+
"learning_rate": 1.799882144961697e-06,
|
| 217 |
+
"loss": 6.1559,
|
| 218 |
+
"mean_token_accuracy": 0.21859725564718246,
|
| 219 |
+
"num_tokens": 1123202.0,
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"entropy": 6.042124252319336,
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
+
"grad_norm": 3.621903657913208,
|
| 226 |
+
"learning_rate": 1.7880966411314081e-06,
|
| 227 |
+
"loss": 5.8441,
|
| 228 |
+
"mean_token_accuracy": 0.24906315237283708,
|
| 229 |
+
"num_tokens": 1173403.0,
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"entropy": 5.921343173980713,
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
+
"grad_norm": 5.658033847808838,
|
| 236 |
+
"learning_rate": 1.7763111373011195e-06,
|
| 237 |
+
"loss": 5.7104,
|
| 238 |
+
"mean_token_accuracy": 0.2625067520141602,
|
| 239 |
+
"num_tokens": 1225026.0,
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"entropy": 6.093586492538452,
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
+
"grad_norm": 2.4292995929718018,
|
| 246 |
+
"learning_rate": 1.7645256334708308e-06,
|
| 247 |
+
"loss": 5.8658,
|
| 248 |
+
"mean_token_accuracy": 0.24842385441064835,
|
| 249 |
+
"num_tokens": 1279013.0,
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"entropy": 6.119112596511841,
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
+
"grad_norm": 3.369384288787842,
|
| 256 |
+
"learning_rate": 1.752740129640542e-06,
|
| 257 |
+
"loss": 5.8784,
|
| 258 |
+
"mean_token_accuracy": 0.24857850253582,
|
| 259 |
+
"num_tokens": 1332547.0,
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"entropy": 6.025163550376892,
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
+
"grad_norm": 2.5110116004943848,
|
| 266 |
+
"learning_rate": 1.7409546258102533e-06,
|
| 267 |
+
"loss": 5.7769,
|
| 268 |
+
"mean_token_accuracy": 0.25835376888513567,
|
| 269 |
+
"num_tokens": 1385192.0,
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"entropy": 5.877259612083435,
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
+
"grad_norm": 2.4179303646087646,
|
| 276 |
+
"learning_rate": 1.7291691219799646e-06,
|
| 277 |
+
"loss": 5.6284,
|
| 278 |
+
"mean_token_accuracy": 0.2756252554059029,
|
| 279 |
+
"num_tokens": 1437071.0,
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"entropy": 6.002246947288513,
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
+
"grad_norm": 3.494359016418457,
|
| 286 |
+
"learning_rate": 1.717383618149676e-06,
|
| 287 |
+
"loss": 5.747,
|
| 288 |
+
"mean_token_accuracy": 0.26462210685014725,
|
| 289 |
+
"num_tokens": 1490818.0,
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"entropy": 5.991955623626709,
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
+
"grad_norm": 2.340975761413574,
|
| 296 |
+
"learning_rate": 1.705598114319387e-06,
|
| 297 |
+
"loss": 5.7379,
|
| 298 |
+
"mean_token_accuracy": 0.26444981098175047,
|
| 299 |
+
"num_tokens": 1544997.0,
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"entropy": 5.91768889427185,
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
+
"grad_norm": 2.2394514083862305,
|
| 306 |
+
"learning_rate": 1.6938126104890984e-06,
|
| 307 |
+
"loss": 5.6564,
|
| 308 |
+
"mean_token_accuracy": 0.2730415526032448,
|
| 309 |
+
"num_tokens": 1598302.0,
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"entropy": 5.982716989517212,
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
+
"grad_norm": 1.876839518547058,
|
| 316 |
+
"learning_rate": 1.6820271066588098e-06,
|
| 317 |
+
"loss": 5.7215,
|
| 318 |
+
"mean_token_accuracy": 0.26642445534467696,
|
| 319 |
+
"num_tokens": 1655267.0,
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"entropy": 5.820467872619629,
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
+
"grad_norm": 2.219966173171997,
|
| 326 |
+
"learning_rate": 1.6702416028285209e-06,
|
| 327 |
+
"loss": 5.5555,
|
| 328 |
+
"mean_token_accuracy": 0.2856418335437775,
|
| 329 |
+
"num_tokens": 1709199.0,
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"entropy": 5.996349005699158,
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
+
"grad_norm": 2.247213840484619,
|
| 336 |
+
"learning_rate": 1.6584560989982322e-06,
|
| 337 |
+
"loss": 5.7283,
|
| 338 |
+
"mean_token_accuracy": 0.2696125540137291,
|
| 339 |
+
"num_tokens": 1765443.0,
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"entropy": 5.696683068275451,
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
+
"grad_norm": 2.8499979972839355,
|
| 346 |
+
"learning_rate": 1.6466705951679433e-06,
|
| 347 |
+
"loss": 5.4335,
|
| 348 |
+
"mean_token_accuracy": 0.29918427973985673,
|
| 349 |
+
"num_tokens": 1817494.0,
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
+
"eval_entropy": 5.993559589034401,
|
| 355 |
+
"eval_loss": 5.737204551696777,
|
| 356 |
+
"eval_mean_token_accuracy": 0.2618687468739699,
|
| 357 |
+
"eval_model_preparation_time": 0.0045,
|
| 358 |
+
"eval_num_tokens": 1856362.0,
|
| 359 |
+
"eval_runtime": 50.5332,
|
| 360 |
+
"eval_samples_per_second": 8.588,
|
| 361 |
+
"eval_steps_per_second": 4.294,
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"entropy": 5.746842083930969,
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
+
"grad_norm": 2.33052921295166,
|
| 368 |
+
"learning_rate": 1.6348850913376547e-06,
|
| 369 |
+
"loss": 5.4796,
|
| 370 |
+
"mean_token_accuracy": 0.2966849410533905,
|
| 371 |
+
"num_tokens": 1870353.0,
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"entropy": 5.859029049873352,
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
+
"grad_norm": 1.6248886585235596,
|
| 378 |
+
"learning_rate": 1.6230995875073658e-06,
|
| 379 |
+
"loss": 5.5975,
|
| 380 |
+
"mean_token_accuracy": 0.2838129925727844,
|
| 381 |
+
"num_tokens": 1926205.0,
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"entropy": 5.731445336341858,
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
+
"grad_norm": 1.6941566467285156,
|
| 388 |
+
"learning_rate": 1.6113140836770771e-06,
|
| 389 |
+
"loss": 5.476,
|
| 390 |
+
"mean_token_accuracy": 0.2992346465587616,
|
| 391 |
+
"num_tokens": 1979821.0,
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"entropy": 5.6993954515457155,
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
+
"grad_norm": 1.1746597290039062,
|
| 398 |
+
"learning_rate": 1.5995285798467883e-06,
|
| 399 |
+
"loss": 5.4608,
|
| 400 |
+
"mean_token_accuracy": 0.3000726142525673,
|
| 401 |
+
"num_tokens": 2034373.0,
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"entropy": 5.668873124122619,
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
+
"grad_norm": 1.728211760520935,
|
| 408 |
+
"learning_rate": 1.5877430760164996e-06,
|
| 409 |
+
"loss": 5.4347,
|
| 410 |
+
"mean_token_accuracy": 0.3033922725915909,
|
| 411 |
+
"num_tokens": 2087339.0,
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
+
"entropy": 5.624621086120605,
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
+
"grad_norm": 1.4078539609909058,
|
| 418 |
+
"learning_rate": 1.5759575721862107e-06,
|
| 419 |
+
"loss": 5.3954,
|
| 420 |
+
"mean_token_accuracy": 0.30784171640872954,
|
| 421 |
+
"num_tokens": 2139520.0,
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"entropy": 5.7141213130950925,
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
+
"grad_norm": 2.186459541320801,
|
| 428 |
+
"learning_rate": 1.564172068355922e-06,
|
| 429 |
+
"loss": 5.4847,
|
| 430 |
+
"mean_token_accuracy": 0.29594049394130706,
|
| 431 |
+
"num_tokens": 2193987.0,
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"entropy": 5.632415266036987,
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
+
"grad_norm": 1.3601349592208862,
|
| 438 |
+
"learning_rate": 1.5523865645256334e-06,
|
| 439 |
+
"loss": 5.4135,
|
| 440 |
+
"mean_token_accuracy": 0.30366597563028336,
|
| 441 |
+
"num_tokens": 2249616.0,
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
+
"entropy": 5.510904269218445,
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
+
"grad_norm": 2.065760612487793,
|
| 448 |
+
"learning_rate": 1.5406010606953445e-06,
|
| 449 |
+
"loss": 5.2904,
|
| 450 |
+
"mean_token_accuracy": 0.3211754837632179,
|
| 451 |
+
"num_tokens": 2300863.0,
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"entropy": 5.703383626937867,
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
+
"grad_norm": 1.1172698736190796,
|
| 458 |
+
"learning_rate": 1.5288155568650559e-06,
|
| 459 |
+
"loss": 5.4802,
|
| 460 |
+
"mean_token_accuracy": 0.29713701367378237,
|
| 461 |
+
"num_tokens": 2356029.0,
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
+
"entropy": 5.565930342674255,
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
+
"grad_norm": 1.7528513669967651,
|
| 468 |
+
"learning_rate": 1.5170300530347672e-06,
|
| 469 |
+
"loss": 5.3518,
|
| 470 |
+
"mean_token_accuracy": 0.31301232606172563,
|
| 471 |
+
"num_tokens": 2408957.0,
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"entropy": 5.496430187225342,
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
+
"grad_norm": 1.892640233039856,
|
| 478 |
+
"learning_rate": 1.5052445492044786e-06,
|
| 479 |
+
"loss": 5.2967,
|
| 480 |
+
"mean_token_accuracy": 0.3181899458169937,
|
| 481 |
+
"num_tokens": 2462569.0,
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"entropy": 5.725150098800659,
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
+
"grad_norm": 1.774940848350525,
|
| 488 |
+
"learning_rate": 1.4934590453741897e-06,
|
| 489 |
+
"loss": 5.5215,
|
| 490 |
+
"mean_token_accuracy": 0.29055028676986694,
|
| 491 |
+
"num_tokens": 2518544.0,
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"entropy": 5.4884827613830565,
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
+
"grad_norm": 2.2167599201202393,
|
| 498 |
+
"learning_rate": 1.481673541543901e-06,
|
| 499 |
+
"loss": 5.2917,
|
| 500 |
+
"mean_token_accuracy": 0.31803421139717103,
|
| 501 |
+
"num_tokens": 2570863.0,
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"entropy": 5.697079472541809,
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
+
"grad_norm": 1.6489030122756958,
|
| 508 |
+
"learning_rate": 1.4698880377136124e-06,
|
| 509 |
+
"loss": 5.4982,
|
| 510 |
+
"mean_token_accuracy": 0.2925163987278938,
|
| 511 |
+
"num_tokens": 2626998.0,
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"entropy": 5.46209939956665,
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
+
"grad_norm": 1.153914451599121,
|
| 518 |
+
"learning_rate": 1.4581025338833235e-06,
|
| 519 |
+
"loss": 5.2736,
|
| 520 |
+
"mean_token_accuracy": 0.3182168474793434,
|
| 521 |
+
"num_tokens": 2681568.0,
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"entropy": 5.4405768728256225,
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
+
"grad_norm": 3.6614978313446045,
|
| 528 |
+
"learning_rate": 1.4463170300530348e-06,
|
| 529 |
+
"loss": 5.2515,
|
| 530 |
+
"mean_token_accuracy": 0.3218736210465431,
|
| 531 |
+
"num_tokens": 2733587.0,
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"entropy": 5.528175053596496,
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
+
"grad_norm": 1.0849746465682983,
|
| 538 |
+
"learning_rate": 1.434531526222746e-06,
|
| 539 |
+
"loss": 5.3378,
|
| 540 |
+
"mean_token_accuracy": 0.31061659604310987,
|
| 541 |
+
"num_tokens": 2787003.0,
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"entropy": 5.46110897064209,
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
+
"grad_norm": 1.8315683603286743,
|
| 548 |
+
"learning_rate": 1.4227460223924573e-06,
|
| 549 |
+
"loss": 5.2782,
|
| 550 |
+
"mean_token_accuracy": 0.31781029611825945,
|
| 551 |
+
"num_tokens": 2840263.0,
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"entropy": 5.455560960769653,
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
+
"grad_norm": 1.1859091520309448,
|
| 558 |
+
"learning_rate": 1.4109605185621684e-06,
|
| 559 |
+
"loss": 5.2735,
|
| 560 |
+
"mean_token_accuracy": 0.3194814011454582,
|
| 561 |
+
"num_tokens": 2894186.0,
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"entropy": 5.430496115684509,
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
+
"grad_norm": 2.3500001430511475,
|
| 568 |
+
"learning_rate": 1.3991750147318797e-06,
|
| 569 |
+
"loss": 5.2464,
|
| 570 |
+
"mean_token_accuracy": 0.32140792965888976,
|
| 571 |
+
"num_tokens": 2948171.0,
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"entropy": 5.588023023605347,
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
+
"grad_norm": 1.727825403213501,
|
| 578 |
+
"learning_rate": 1.3873895109015909e-06,
|
| 579 |
+
"loss": 5.4028,
|
| 580 |
+
"mean_token_accuracy": 0.3039530631899834,
|
| 581 |
+
"num_tokens": 3002678.0,
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"entropy": 5.410525422096253,
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
+
"grad_norm": 1.3401474952697754,
|
| 588 |
+
"learning_rate": 1.3756040070713022e-06,
|
| 589 |
+
"loss": 5.2298,
|
| 590 |
+
"mean_token_accuracy": 0.324065263569355,
|
| 591 |
+
"num_tokens": 3055844.0,
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"entropy": 5.36959942817688,
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
+
"grad_norm": 1.1892589330673218,
|
| 598 |
+
"learning_rate": 1.3638185032410133e-06,
|
| 599 |
+
"loss": 5.1956,
|
| 600 |
+
"mean_token_accuracy": 0.32639502108097074,
|
| 601 |
+
"num_tokens": 3108636.0,
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"entropy": 5.53826907157898,
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
+
"grad_norm": 1.2652360200881958,
|
| 608 |
+
"learning_rate": 1.3520329994107247e-06,
|
| 609 |
+
"loss": 5.3583,
|
| 610 |
+
"mean_token_accuracy": 0.3074926760792732,
|
| 611 |
+
"num_tokens": 3162627.0,
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"entropy": 5.417449145317078,
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
+
"grad_norm": 1.584312915802002,
|
| 618 |
+
"learning_rate": 1.340247495580436e-06,
|
| 619 |
+
"loss": 5.2388,
|
| 620 |
+
"mean_token_accuracy": 0.32019727885723115,
|
| 621 |
+
"num_tokens": 3216409.0,
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"entropy": 5.241390740871429,
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
+
"grad_norm": 1.5219439268112183,
|
| 628 |
+
"learning_rate": 1.3284619917501471e-06,
|
| 629 |
+
"loss": 5.0645,
|
| 630 |
+
"mean_token_accuracy": 0.3445430138707161,
|
| 631 |
+
"num_tokens": 3266967.0,
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"entropy": 5.405424036979675,
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
+
"grad_norm": 2.1165153980255127,
|
| 638 |
+
"learning_rate": 1.3166764879198585e-06,
|
| 639 |
+
"loss": 5.232,
|
| 640 |
+
"mean_token_accuracy": 0.32085000157356264,
|
| 641 |
+
"num_tokens": 3319877.0,
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"entropy": 5.123006024360657,
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
+
"grad_norm": 1.2189785242080688,
|
| 648 |
+
"learning_rate": 1.3048909840895698e-06,
|
| 649 |
+
"loss": 4.9582,
|
| 650 |
+
"mean_token_accuracy": 0.356108532845974,
|
| 651 |
+
"num_tokens": 3368569.0,
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"entropy": 5.417610831260681,
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
+
"grad_norm": 1.5157604217529297,
|
| 658 |
+
"learning_rate": 1.2931054802592812e-06,
|
| 659 |
+
"loss": 5.2454,
|
| 660 |
+
"mean_token_accuracy": 0.31976755023002623,
|
| 661 |
+
"num_tokens": 3422449.0,
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"entropy": 5.409690895080566,
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
+
"grad_norm": 1.3088161945343018,
|
| 668 |
+
"learning_rate": 1.2813199764289923e-06,
|
| 669 |
+
"loss": 5.2348,
|
| 670 |
+
"mean_token_accuracy": 0.32325415283441544,
|
| 671 |
+
"num_tokens": 3474399.0,
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"entropy": 5.44662567615509,
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
+
"grad_norm": 2.178372621536255,
|
| 678 |
+
"learning_rate": 1.2695344725987036e-06,
|
| 679 |
+
"loss": 5.2661,
|
| 680 |
+
"mean_token_accuracy": 0.3182847076654434,
|
| 681 |
+
"num_tokens": 3527726.0,
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"entropy": 5.512614865303039,
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
+
"grad_norm": 1.3050425052642822,
|
| 688 |
+
"learning_rate": 1.2577489687684147e-06,
|
| 689 |
+
"loss": 5.3416,
|
| 690 |
+
"mean_token_accuracy": 0.3084403133392334,
|
| 691 |
+
"num_tokens": 3581980.0,
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"entropy": 5.379772834777832,
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
+
"grad_norm": 1.4584404230117798,
|
| 698 |
+
"learning_rate": 1.245963464938126e-06,
|
| 699 |
+
"loss": 5.2087,
|
| 700 |
+
"mean_token_accuracy": 0.32388432770967485,
|
| 701 |
+
"num_tokens": 3635393.0,
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"entropy": 5.483665924072266,
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
+
"grad_norm": 1.2157734632492065,
|
| 708 |
+
"learning_rate": 1.2341779611078374e-06,
|
| 709 |
+
"loss": 5.3101,
|
| 710 |
+
"mean_token_accuracy": 0.3121953472495079,
|
| 711 |
+
"num_tokens": 3689894.0,
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
+
"eval_entropy": 5.711394641805904,
|
| 717 |
+
"eval_loss": 5.55628776550293,
|
| 718 |
+
"eval_mean_token_accuracy": 0.2764948787777105,
|
| 719 |
+
"eval_model_preparation_time": 0.0045,
|
| 720 |
+
"eval_num_tokens": 3712724.0,
|
| 721 |
+
"eval_runtime": 50.187,
|
| 722 |
+
"eval_samples_per_second": 8.648,
|
| 723 |
+
"eval_steps_per_second": 4.324,
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"entropy": 5.349283556938172,
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
+
"grad_norm": 1.1696771383285522,
|
| 730 |
+
"learning_rate": 1.2223924572775486e-06,
|
| 731 |
+
"loss": 5.1782,
|
| 732 |
+
"mean_token_accuracy": 0.33028870791196824,
|
| 733 |
+
"num_tokens": 3740861.0,
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"entropy": 5.4721107006073,
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
+
"grad_norm": 1.8449370861053467,
|
| 740 |
+
"learning_rate": 1.2106069534472599e-06,
|
| 741 |
+
"loss": 5.2978,
|
| 742 |
+
"mean_token_accuracy": 0.31511022299528124,
|
| 743 |
+
"num_tokens": 3794869.0,
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"entropy": 5.404226851463318,
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
+
"grad_norm": 3.789496660232544,
|
| 750 |
+
"learning_rate": 1.198821449616971e-06,
|
| 751 |
+
"loss": 5.2371,
|
| 752 |
+
"mean_token_accuracy": 0.32092176616191864,
|
| 753 |
+
"num_tokens": 3848573.0,
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
+
"entropy": 5.435445628166199,
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
+
"grad_norm": 2.2847959995269775,
|
| 760 |
+
"learning_rate": 1.1870359457866824e-06,
|
| 761 |
+
"loss": 5.2662,
|
| 762 |
+
"mean_token_accuracy": 0.3186633634567261,
|
| 763 |
+
"num_tokens": 3901204.0,
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
+
"entropy": 5.4066293334960935,
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
+
"grad_norm": 1.0950902700424194,
|
| 770 |
+
"learning_rate": 1.1752504419563935e-06,
|
| 771 |
+
"loss": 5.2345,
|
| 772 |
+
"mean_token_accuracy": 0.32156052827835085,
|
| 773 |
+
"num_tokens": 3953753.0,
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
+
"entropy": 5.272332944869995,
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
+
"grad_norm": 2.1477339267730713,
|
| 780 |
+
"learning_rate": 1.1634649381261048e-06,
|
| 781 |
+
"loss": 5.1091,
|
| 782 |
+
"mean_token_accuracy": 0.3380983591079712,
|
| 783 |
+
"num_tokens": 4005481.0,
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"entropy": 5.4118804311752315,
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
+
"grad_norm": 1.4509484767913818,
|
| 790 |
+
"learning_rate": 1.151679434295816e-06,
|
| 791 |
+
"loss": 5.2448,
|
| 792 |
+
"mean_token_accuracy": 0.3208243528008461,
|
| 793 |
+
"num_tokens": 4058829.0,
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"entropy": 5.4763900089263915,
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
+
"grad_norm": 1.0856804847717285,
|
| 800 |
+
"learning_rate": 1.1398939304655273e-06,
|
| 801 |
+
"loss": 5.3042,
|
| 802 |
+
"mean_token_accuracy": 0.31338351368904116,
|
| 803 |
+
"num_tokens": 4113326.0,
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
+
"entropy": 5.328452725410461,
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
+
"grad_norm": 3.2843880653381348,
|
| 810 |
+
"learning_rate": 1.1281084266352386e-06,
|
| 811 |
+
"loss": 5.1624,
|
| 812 |
+
"mean_token_accuracy": 0.3305218696594238,
|
| 813 |
+
"num_tokens": 4165454.0,
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"entropy": 5.383157343864441,
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
+
"grad_norm": 2.207082748413086,
|
| 820 |
+
"learning_rate": 1.1163229228049497e-06,
|
| 821 |
+
"loss": 5.2163,
|
| 822 |
+
"mean_token_accuracy": 0.32331310987472534,
|
| 823 |
+
"num_tokens": 4219250.0,
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
+
"entropy": 5.585261764526368,
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
+
"grad_norm": 2.7102835178375244,
|
| 830 |
+
"learning_rate": 1.104537418974661e-06,
|
| 831 |
+
"loss": 5.4137,
|
| 832 |
+
"mean_token_accuracy": 0.29959124475717547,
|
| 833 |
+
"num_tokens": 4274711.0,
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
+
"entropy": 5.434073266983032,
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
+
"grad_norm": 1.3775779008865356,
|
| 840 |
+
"learning_rate": 1.0927519151443724e-06,
|
| 841 |
+
"loss": 5.2644,
|
| 842 |
+
"mean_token_accuracy": 0.3175011593103409,
|
| 843 |
+
"num_tokens": 4328616.0,
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
+
"entropy": 5.462391858100891,
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
+
"grad_norm": 1.4101024866104126,
|
| 850 |
+
"learning_rate": 1.0809664113140838e-06,
|
| 851 |
+
"loss": 5.2924,
|
| 852 |
+
"mean_token_accuracy": 0.3137941011786461,
|
| 853 |
+
"num_tokens": 4382416.0,
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
+
"entropy": 5.529892563819885,
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
+
"grad_norm": 1.2311837673187256,
|
| 860 |
+
"learning_rate": 1.0691809074837949e-06,
|
| 861 |
+
"loss": 5.364,
|
| 862 |
+
"mean_token_accuracy": 0.3046491605043411,
|
| 863 |
+
"num_tokens": 4437848.0,
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
+
"entropy": 5.4370484542846675,
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
+
"grad_norm": 1.0929864645004272,
|
| 870 |
+
"learning_rate": 1.0573954036535062e-06,
|
| 871 |
+
"loss": 5.2734,
|
| 872 |
+
"mean_token_accuracy": 0.3169013774394989,
|
| 873 |
+
"num_tokens": 4491185.0,
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
+
"entropy": 5.395377616882325,
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
+
"grad_norm": 1.5457273721694946,
|
| 880 |
+
"learning_rate": 1.0456098998232174e-06,
|
| 881 |
+
"loss": 5.2276,
|
| 882 |
+
"mean_token_accuracy": 0.32221508473157884,
|
| 883 |
+
"num_tokens": 4544086.0,
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"entropy": 5.443737335205078,
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
+
"grad_norm": 1.4844346046447754,
|
| 890 |
+
"learning_rate": 1.0338243959929287e-06,
|
| 891 |
+
"loss": 5.2786,
|
| 892 |
+
"mean_token_accuracy": 0.3157751387357712,
|
| 893 |
+
"num_tokens": 4597677.0,
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
+
"entropy": 5.419876251220703,
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
+
"grad_norm": 1.2481963634490967,
|
| 900 |
+
"learning_rate": 1.02203889216264e-06,
|
| 901 |
+
"loss": 5.2564,
|
| 902 |
+
"mean_token_accuracy": 0.31889803290367125,
|
| 903 |
+
"num_tokens": 4651343.0,
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
+
"entropy": 5.578677978515625,
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
+
"grad_norm": 2.0005414485931396,
|
| 910 |
+
"learning_rate": 1.0102533883323512e-06,
|
| 911 |
+
"loss": 5.4145,
|
| 912 |
+
"mean_token_accuracy": 0.30037090003490446,
|
| 913 |
+
"num_tokens": 4705985.0,
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
+
"entropy": 5.279946126937866,
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
+
"grad_norm": 1.080521821975708,
|
| 920 |
+
"learning_rate": 9.984678845020625e-07,
|
| 921 |
+
"loss": 5.1226,
|
| 922 |
+
"mean_token_accuracy": 0.3341303279995918,
|
| 923 |
+
"num_tokens": 4757741.0,
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"entropy": 5.551463279724121,
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
+
"grad_norm": 1.28898024559021,
|
| 930 |
+
"learning_rate": 9.866823806717736e-07,
|
| 931 |
+
"loss": 5.3832,
|
| 932 |
+
"mean_token_accuracy": 0.3028248634934425,
|
| 933 |
+
"num_tokens": 4812808.0,
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
+
"entropy": 5.3787487554550175,
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
+
"grad_norm": 1.5697983503341675,
|
| 940 |
+
"learning_rate": 9.74896876841485e-07,
|
| 941 |
+
"loss": 5.2141,
|
| 942 |
+
"mean_token_accuracy": 0.3227942296862602,
|
| 943 |
+
"num_tokens": 4866572.0,
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
+
"entropy": 5.460358958244324,
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
+
"grad_norm": 1.3180441856384277,
|
| 950 |
+
"learning_rate": 9.63111373011196e-07,
|
| 951 |
+
"loss": 5.2954,
|
| 952 |
+
"mean_token_accuracy": 0.31269474506378175,
|
| 953 |
+
"num_tokens": 4921312.0,
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"entropy": 5.434084935188293,
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
+
"grad_norm": 1.2409590482711792,
|
| 960 |
+
"learning_rate": 9.513258691809074e-07,
|
| 961 |
+
"loss": 5.271,
|
| 962 |
+
"mean_token_accuracy": 0.3172155100107193,
|
| 963 |
+
"num_tokens": 4974289.0,
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
+
"entropy": 5.406955418586731,
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
+
"grad_norm": 1.4782609939575195,
|
| 970 |
+
"learning_rate": 9.395403653506187e-07,
|
| 971 |
+
"loss": 5.2473,
|
| 972 |
+
"mean_token_accuracy": 0.32031788885593415,
|
| 973 |
+
"num_tokens": 5028149.0,
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
+
"entropy": 5.206603040695191,
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
+
"grad_norm": 2.351633071899414,
|
| 980 |
+
"learning_rate": 9.2775486152033e-07,
|
| 981 |
+
"loss": 5.0478,
|
| 982 |
+
"mean_token_accuracy": 0.3428420132398605,
|
| 983 |
+
"num_tokens": 5079349.0,
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
+
"entropy": 5.388812799453735,
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
+
"grad_norm": 7.564618110656738,
|
| 990 |
+
"learning_rate": 9.159693576900412e-07,
|
| 991 |
+
"loss": 5.2281,
|
| 992 |
+
"mean_token_accuracy": 0.3222071170806885,
|
| 993 |
+
"num_tokens": 5132564.0,
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"entropy": 5.374106278419495,
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
+
"grad_norm": 1.4734679460525513,
|
| 1000 |
+
"learning_rate": 9.041838538597525e-07,
|
| 1001 |
+
"loss": 5.2161,
|
| 1002 |
+
"mean_token_accuracy": 0.3219477406144142,
|
| 1003 |
+
"num_tokens": 5185921.0,
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
+
"entropy": 5.232998585700988,
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
+
"grad_norm": 1.4175471067428589,
|
| 1010 |
+
"learning_rate": 8.923983500294637e-07,
|
| 1011 |
+
"loss": 5.0769,
|
| 1012 |
+
"mean_token_accuracy": 0.3403926733136177,
|
| 1013 |
+
"num_tokens": 5237521.0,
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
+
"entropy": 5.394891719818116,
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
+
"grad_norm": 4.951873779296875,
|
| 1020 |
+
"learning_rate": 8.806128461991749e-07,
|
| 1021 |
+
"loss": 5.2344,
|
| 1022 |
+
"mean_token_accuracy": 0.3213117456436157,
|
| 1023 |
+
"num_tokens": 5291104.0,
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"entropy": 5.413805012702942,
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
+
"grad_norm": 1.679518461227417,
|
| 1030 |
+
"learning_rate": 8.688273423688863e-07,
|
| 1031 |
+
"loss": 5.2597,
|
| 1032 |
+
"mean_token_accuracy": 0.3165634173154831,
|
| 1033 |
+
"num_tokens": 5345058.0,
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
+
"entropy": 5.256177935600281,
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
+
"grad_norm": 1.8892916440963745,
|
| 1040 |
+
"learning_rate": 8.570418385385975e-07,
|
| 1041 |
+
"loss": 5.1004,
|
| 1042 |
+
"mean_token_accuracy": 0.3369427987933159,
|
| 1043 |
+
"num_tokens": 5395918.0,
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
+
"entropy": 5.259814453125,
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
+
"grad_norm": 1.3802675008773804,
|
| 1050 |
+
"learning_rate": 8.452563347083087e-07,
|
| 1051 |
+
"loss": 5.1057,
|
| 1052 |
+
"mean_token_accuracy": 0.3362414276599884,
|
| 1053 |
+
"num_tokens": 5448086.0,
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
+
"entropy": 5.416206178665161,
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
+
"grad_norm": 1.7677236795425415,
|
| 1060 |
+
"learning_rate": 8.3347083087802e-07,
|
| 1061 |
+
"loss": 5.2562,
|
| 1062 |
+
"mean_token_accuracy": 0.31725785195827483,
|
| 1063 |
+
"num_tokens": 5501959.0,
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
+
"entropy": 5.507337794303894,
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
+
"grad_norm": 1.021727442741394,
|
| 1070 |
+
"learning_rate": 8.216853270477313e-07,
|
| 1071 |
+
"loss": 5.344,
|
| 1072 |
+
"mean_token_accuracy": 0.30679062128067014,
|
| 1073 |
+
"num_tokens": 5557908.0,
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
+
"eval_entropy": 5.682707933786278,
|
| 1079 |
+
"eval_loss": 5.53223991394043,
|
| 1080 |
+
"eval_mean_token_accuracy": 0.27747743456594404,
|
| 1081 |
+
"eval_model_preparation_time": 0.0045,
|
| 1082 |
+
"eval_num_tokens": 5569086.0,
|
| 1083 |
+
"eval_runtime": 49.9944,
|
| 1084 |
+
"eval_samples_per_second": 8.681,
|
| 1085 |
+
"eval_steps_per_second": 4.34,
|
| 1086 |
"step": 5211
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
+
"entropy": 5.209756035804748,
|
| 1090 |
"epoch": 3.0224525043177892,
|
| 1091 |
+
"grad_norm": 1.725786566734314,
|
| 1092 |
+
"learning_rate": 8.098998232174425e-07,
|
| 1093 |
+
"loss": 5.0541,
|
| 1094 |
+
"mean_token_accuracy": 0.34166110813617706,
|
| 1095 |
+
"num_tokens": 5608917.0,
|
| 1096 |
"step": 5250
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
+
"entropy": 5.396296281814575,
|
| 1100 |
"epoch": 3.051237766263673,
|
| 1101 |
+
"grad_norm": 0.7720207571983337,
|
| 1102 |
+
"learning_rate": 7.981143193871538e-07,
|
| 1103 |
+
"loss": 5.2337,
|
| 1104 |
+
"mean_token_accuracy": 0.32116260558366777,
|
| 1105 |
+
"num_tokens": 5662712.0,
|
| 1106 |
"step": 5300
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
+
"entropy": 5.341518473625183,
|
| 1110 |
"epoch": 3.0800230282095566,
|
| 1111 |
+
"grad_norm": 2.2686808109283447,
|
| 1112 |
+
"learning_rate": 7.86328815556865e-07,
|
| 1113 |
+
"loss": 5.1824,
|
| 1114 |
+
"mean_token_accuracy": 0.32726580530405047,
|
| 1115 |
+
"num_tokens": 5715921.0,
|
| 1116 |
"step": 5350
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
+
"entropy": 5.376176896095276,
|
| 1120 |
"epoch": 3.1088082901554404,
|
| 1121 |
+
"grad_norm": 1.2420796155929565,
|
| 1122 |
+
"learning_rate": 7.745433117265762e-07,
|
| 1123 |
+
"loss": 5.2162,
|
| 1124 |
+
"mean_token_accuracy": 0.32142678707838057,
|
| 1125 |
+
"num_tokens": 5769436.0,
|
| 1126 |
"step": 5400
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
+
"entropy": 5.4553061914443965,
|
| 1130 |
"epoch": 3.137593552101324,
|
| 1131 |
+
"grad_norm": 1.2402859926223755,
|
| 1132 |
+
"learning_rate": 7.627578078962876e-07,
|
| 1133 |
+
"loss": 5.2971,
|
| 1134 |
+
"mean_token_accuracy": 0.31396267503499986,
|
| 1135 |
+
"num_tokens": 5823126.0,
|
| 1136 |
"step": 5450
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"entropy": 5.385247969627381,
|
| 1140 |
"epoch": 3.166378814047208,
|
| 1141 |
+
"grad_norm": 1.112062931060791,
|
| 1142 |
+
"learning_rate": 7.509723040659988e-07,
|
| 1143 |
+
"loss": 5.2324,
|
| 1144 |
+
"mean_token_accuracy": 0.3207343602180481,
|
| 1145 |
+
"num_tokens": 5875751.0,
|
| 1146 |
"step": 5500
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
+
"entropy": 5.55422221660614,
|
| 1150 |
"epoch": 3.1951640759930915,
|
| 1151 |
+
"grad_norm": 1.5440446138381958,
|
| 1152 |
+
"learning_rate": 7.3918680023571e-07,
|
| 1153 |
+
"loss": 5.3902,
|
| 1154 |
+
"mean_token_accuracy": 0.3006985321640968,
|
| 1155 |
+
"num_tokens": 5932163.0,
|
| 1156 |
"step": 5550
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
+
"entropy": 5.403217372894287,
|
| 1160 |
"epoch": 3.223949337938975,
|
| 1161 |
+
"grad_norm": 0.8481096625328064,
|
| 1162 |
+
"learning_rate": 7.274012964054213e-07,
|
| 1163 |
+
"loss": 5.2417,
|
| 1164 |
+
"mean_token_accuracy": 0.3210747820138931,
|
| 1165 |
+
"num_tokens": 5985889.0,
|
| 1166 |
"step": 5600
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
+
"entropy": 5.388293180465698,
|
| 1170 |
"epoch": 3.252734599884859,
|
| 1171 |
+
"grad_norm": 0.9305989146232605,
|
| 1172 |
+
"learning_rate": 7.156157925751326e-07,
|
| 1173 |
+
"loss": 5.2319,
|
| 1174 |
+
"mean_token_accuracy": 0.3206030324101448,
|
| 1175 |
+
"num_tokens": 6040052.0,
|
| 1176 |
"step": 5650
|
| 1177 |
},
|
| 1178 |
{
|
| 1179 |
+
"entropy": 5.401709322929382,
|
| 1180 |
"epoch": 3.2815198618307426,
|
| 1181 |
+
"grad_norm": 0.8080459237098694,
|
| 1182 |
+
"learning_rate": 7.038302887448438e-07,
|
| 1183 |
+
"loss": 5.2438,
|
| 1184 |
+
"mean_token_accuracy": 0.3199671137332916,
|
| 1185 |
+
"num_tokens": 6092350.0,
|
| 1186 |
"step": 5700
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
+
"entropy": 5.4320423412323,
|
| 1190 |
"epoch": 3.3103051237766263,
|
| 1191 |
+
"grad_norm": 1.9186089038848877,
|
| 1192 |
+
"learning_rate": 6.920447849145551e-07,
|
| 1193 |
+
"loss": 5.2696,
|
| 1194 |
+
"mean_token_accuracy": 0.31657984614372253,
|
| 1195 |
+
"num_tokens": 6146112.0,
|
| 1196 |
"step": 5750
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
+
"entropy": 5.276471285820008,
|
| 1200 |
"epoch": 3.33909038572251,
|
| 1201 |
+
"grad_norm": 1.032879114151001,
|
| 1202 |
+
"learning_rate": 6.802592810842663e-07,
|
| 1203 |
+
"loss": 5.1224,
|
| 1204 |
+
"mean_token_accuracy": 0.3347566506266594,
|
| 1205 |
+
"num_tokens": 6197916.0,
|
| 1206 |
"step": 5800
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
+
"entropy": 5.122317051887512,
|
| 1210 |
"epoch": 3.3678756476683938,
|
| 1211 |
+
"grad_norm": 3.156858444213867,
|
| 1212 |
+
"learning_rate": 6.684737772539775e-07,
|
| 1213 |
+
"loss": 4.9706,
|
| 1214 |
+
"mean_token_accuracy": 0.35455317378044127,
|
| 1215 |
+
"num_tokens": 6247565.0,
|
| 1216 |
"step": 5850
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
+
"entropy": 5.346597375869751,
|
| 1220 |
"epoch": 3.3966609096142775,
|
| 1221 |
+
"grad_norm": 1.2619549036026,
|
| 1222 |
+
"learning_rate": 6.566882734236889e-07,
|
| 1223 |
+
"loss": 5.1902,
|
| 1224 |
+
"mean_token_accuracy": 0.3258721518516541,
|
| 1225 |
+
"num_tokens": 6300481.0,
|
| 1226 |
"step": 5900
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
+
"entropy": 5.413151068687439,
|
| 1230 |
"epoch": 3.425446171560161,
|
| 1231 |
+
"grad_norm": 1.801740050315857,
|
| 1232 |
+
"learning_rate": 6.449027695934001e-07,
|
| 1233 |
+
"loss": 5.2513,
|
| 1234 |
+
"mean_token_accuracy": 0.3187857499718666,
|
| 1235 |
+
"num_tokens": 6353098.0,
|
| 1236 |
"step": 5950
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
+
"entropy": 5.464186942577362,
|
| 1240 |
"epoch": 3.454231433506045,
|
| 1241 |
+
"grad_norm": 1.6306997537612915,
|
| 1242 |
+
"learning_rate": 6.331172657631113e-07,
|
| 1243 |
+
"loss": 5.3043,
|
| 1244 |
+
"mean_token_accuracy": 0.31154109388589857,
|
| 1245 |
+
"num_tokens": 6407984.0,
|
| 1246 |
"step": 6000
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
+
"entropy": 5.401795778274536,
|
| 1250 |
"epoch": 3.4830166954519286,
|
| 1251 |
+
"grad_norm": 1.1694583892822266,
|
| 1252 |
+
"learning_rate": 6.213317619328226e-07,
|
| 1253 |
+
"loss": 5.2427,
|
| 1254 |
+
"mean_token_accuracy": 0.31954523265361784,
|
| 1255 |
+
"num_tokens": 6461854.0,
|
| 1256 |
"step": 6050
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
+
"entropy": 5.317689285278321,
|
| 1260 |
"epoch": 3.5118019573978123,
|
| 1261 |
+
"grad_norm": 0.9361855387687683,
|
| 1262 |
+
"learning_rate": 6.095462581025339e-07,
|
| 1263 |
+
"loss": 5.1588,
|
| 1264 |
+
"mean_token_accuracy": 0.330586878657341,
|
| 1265 |
+
"num_tokens": 6514882.0,
|
| 1266 |
"step": 6100
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
+
"entropy": 5.478708257675171,
|
| 1270 |
"epoch": 3.540587219343696,
|
| 1271 |
+
"grad_norm": 1.05711030960083,
|
| 1272 |
+
"learning_rate": 5.977607542722451e-07,
|
| 1273 |
+
"loss": 5.321,
|
| 1274 |
+
"mean_token_accuracy": 0.3104448106884956,
|
| 1275 |
+
"num_tokens": 6569455.0,
|
| 1276 |
"step": 6150
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
+
"entropy": 5.309361801147461,
|
| 1280 |
"epoch": 3.5693724812895797,
|
| 1281 |
+
"grad_norm": 1.3499550819396973,
|
| 1282 |
+
"learning_rate": 5.859752504419564e-07,
|
| 1283 |
+
"loss": 5.153,
|
| 1284 |
+
"mean_token_accuracy": 0.331512533724308,
|
| 1285 |
+
"num_tokens": 6621734.0,
|
| 1286 |
"step": 6200
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
+
"entropy": 5.296572666168213,
|
| 1290 |
"epoch": 3.5981577432354634,
|
| 1291 |
+
"grad_norm": 1.940708875656128,
|
| 1292 |
+
"learning_rate": 5.741897466116676e-07,
|
| 1293 |
+
"loss": 5.14,
|
| 1294 |
+
"mean_token_accuracy": 0.3299832499027252,
|
| 1295 |
+
"num_tokens": 6674994.0,
|
| 1296 |
"step": 6250
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
+
"entropy": 5.544284400939941,
|
| 1300 |
"epoch": 3.626943005181347,
|
| 1301 |
+
"grad_norm": 1.8903827667236328,
|
| 1302 |
+
"learning_rate": 5.624042427813788e-07,
|
| 1303 |
+
"loss": 5.3885,
|
| 1304 |
+
"mean_token_accuracy": 0.3016947290301323,
|
| 1305 |
+
"num_tokens": 6730674.0,
|
| 1306 |
"step": 6300
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
+
"entropy": 5.333053431510925,
|
| 1310 |
"epoch": 3.655728267127231,
|
| 1311 |
+
"grad_norm": 1.1618578433990479,
|
| 1312 |
+
"learning_rate": 5.506187389510902e-07,
|
| 1313 |
+
"loss": 5.1781,
|
| 1314 |
+
"mean_token_accuracy": 0.3275001719594002,
|
| 1315 |
+
"num_tokens": 6784235.0,
|
| 1316 |
"step": 6350
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
+
"entropy": 5.4938449716568,
|
| 1320 |
"epoch": 3.6845135290731146,
|
| 1321 |
+
"grad_norm": 1.384329080581665,
|
| 1322 |
+
"learning_rate": 5.388332351208014e-07,
|
| 1323 |
+
"loss": 5.3399,
|
| 1324 |
+
"mean_token_accuracy": 0.3068840709328651,
|
| 1325 |
+
"num_tokens": 6839590.0,
|
| 1326 |
"step": 6400
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
+
"entropy": 5.277545223236084,
|
| 1330 |
"epoch": 3.7132987910189983,
|
| 1331 |
+
"grad_norm": 1.8918265104293823,
|
| 1332 |
+
"learning_rate": 5.270477312905126e-07,
|
| 1333 |
+
"loss": 5.1221,
|
| 1334 |
+
"mean_token_accuracy": 0.33364981949329375,
|
| 1335 |
+
"num_tokens": 6891301.0,
|
| 1336 |
"step": 6450
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
+
"entropy": 5.40100293636322,
|
| 1340 |
"epoch": 3.742084052964882,
|
| 1341 |
+
"grad_norm": 1.6968809366226196,
|
| 1342 |
+
"learning_rate": 5.152622274602239e-07,
|
| 1343 |
+
"loss": 5.2471,
|
| 1344 |
+
"mean_token_accuracy": 0.31912936180830004,
|
| 1345 |
+
"num_tokens": 6945510.0,
|
| 1346 |
"step": 6500
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
+
"entropy": 5.561220169067383,
|
| 1350 |
"epoch": 3.7708693149107657,
|
| 1351 |
+
"grad_norm": 2.066960573196411,
|
| 1352 |
+
"learning_rate": 5.034767236299352e-07,
|
| 1353 |
+
"loss": 5.4026,
|
| 1354 |
+
"mean_token_accuracy": 0.2984810543060303,
|
| 1355 |
+
"num_tokens": 7001870.0,
|
| 1356 |
"step": 6550
|
| 1357 |
},
|
| 1358 |
{
|
| 1359 |
+
"entropy": 5.3108087682724,
|
| 1360 |
"epoch": 3.7996545768566494,
|
| 1361 |
+
"grad_norm": 1.6065007448196411,
|
| 1362 |
+
"learning_rate": 4.916912197996464e-07,
|
| 1363 |
+
"loss": 5.155,
|
| 1364 |
+
"mean_token_accuracy": 0.3304683968424797,
|
| 1365 |
+
"num_tokens": 7053974.0,
|
| 1366 |
"step": 6600
|
| 1367 |
},
|
| 1368 |
{
|
| 1369 |
+
"entropy": 5.323807754516602,
|
| 1370 |
"epoch": 3.828439838802533,
|
| 1371 |
+
"grad_norm": 2.6806318759918213,
|
| 1372 |
+
"learning_rate": 4.799057159693577e-07,
|
| 1373 |
+
"loss": 5.1653,
|
| 1374 |
+
"mean_token_accuracy": 0.3294159671664238,
|
| 1375 |
+
"num_tokens": 7107061.0,
|
| 1376 |
"step": 6650
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
+
"entropy": 5.4716163873672485,
|
| 1380 |
"epoch": 3.857225100748417,
|
| 1381 |
+
"grad_norm": 1.8264856338500977,
|
| 1382 |
+
"learning_rate": 4.6812021213906895e-07,
|
| 1383 |
+
"loss": 5.3124,
|
| 1384 |
+
"mean_token_accuracy": 0.3109353107213974,
|
| 1385 |
+
"num_tokens": 7161697.0,
|
| 1386 |
"step": 6700
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
+
"entropy": 5.382365622520447,
|
| 1390 |
"epoch": 3.8860103626943006,
|
| 1391 |
+
"grad_norm": 0.9954923987388611,
|
| 1392 |
+
"learning_rate": 4.563347083087802e-07,
|
| 1393 |
+
"loss": 5.2237,
|
| 1394 |
+
"mean_token_accuracy": 0.32161149621009827,
|
| 1395 |
+
"num_tokens": 7215524.0,
|
| 1396 |
"step": 6750
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
+
"entropy": 5.277496585845947,
|
| 1400 |
"epoch": 3.9147956246401843,
|
| 1401 |
+
"grad_norm": 1.267786979675293,
|
| 1402 |
+
"learning_rate": 4.445492044784914e-07,
|
| 1403 |
+
"loss": 5.1265,
|
| 1404 |
+
"mean_token_accuracy": 0.3319795566797257,
|
| 1405 |
+
"num_tokens": 7267329.0,
|
| 1406 |
"step": 6800
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
+
"entropy": 5.550942025184631,
|
| 1410 |
"epoch": 3.943580886586068,
|
| 1411 |
+
"grad_norm": 0.9425063133239746,
|
| 1412 |
+
"learning_rate": 4.3276370064820265e-07,
|
| 1413 |
+
"loss": 5.3898,
|
| 1414 |
+
"mean_token_accuracy": 0.30050904959440233,
|
| 1415 |
+
"num_tokens": 7324070.0,
|
| 1416 |
"step": 6850
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
+
"entropy": 5.125799627304077,
|
| 1420 |
"epoch": 3.9723661485319517,
|
| 1421 |
+
"grad_norm": 5.447021007537842,
|
| 1422 |
+
"learning_rate": 4.20978196817914e-07,
|
| 1423 |
+
"loss": 4.9781,
|
| 1424 |
+
"mean_token_accuracy": 0.3520450854301453,
|
| 1425 |
+
"num_tokens": 7375083.0,
|
| 1426 |
"step": 6900
|
| 1427 |
},
|
| 1428 |
{
|
| 1429 |
"epoch": 4.0,
|
| 1430 |
+
"eval_entropy": 5.6681923492712905,
|
| 1431 |
+
"eval_loss": 5.525067329406738,
|
| 1432 |
+
"eval_mean_token_accuracy": 0.2779707208893816,
|
| 1433 |
+
"eval_model_preparation_time": 0.0045,
|
| 1434 |
+
"eval_num_tokens": 7425448.0,
|
| 1435 |
+
"eval_runtime": 49.7944,
|
| 1436 |
+
"eval_samples_per_second": 8.716,
|
| 1437 |
+
"eval_steps_per_second": 4.358,
|
| 1438 |
"step": 6948
|
| 1439 |
}
|
| 1440 |
],
|
| 1441 |
"logging_steps": 50,
|
| 1442 |
+
"max_steps": 8685,
|
| 1443 |
"num_input_tokens_seen": 0,
|
| 1444 |
+
"num_train_epochs": 5,
|
| 1445 |
"save_steps": 500,
|
| 1446 |
"stateful_callbacks": {
|
| 1447 |
"TrainerControl": {
|
|
|
|
| 1450 |
"should_evaluate": false,
|
| 1451 |
"should_log": false,
|
| 1452 |
"should_save": true,
|
| 1453 |
+
"should_training_stop": false
|
| 1454 |
},
|
| 1455 |
"attributes": {}
|
| 1456 |
}
|
| 1457 |
},
|
| 1458 |
+
"total_flos": 1.016969752533504e+17,
|
| 1459 |
"train_batch_size": 2,
|
| 1460 |
"trial_name": null,
|
| 1461 |
"trial_params": null
|
checkpoint-6948/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|
checkpoint-8685/adapter_config.json
CHANGED
|
@@ -25,14 +25,14 @@
|
|
| 25 |
"peft_type": "LORA",
|
| 26 |
"peft_version": "0.18.0",
|
| 27 |
"qalora_group_size": 16,
|
| 28 |
-
"r":
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
|
|
|
| 32 |
"k_proj",
|
| 33 |
"v_proj",
|
| 34 |
-
"q_proj"
|
| 35 |
-
"o_proj"
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 25 |
"peft_type": "LORA",
|
| 26 |
"peft_version": "0.18.0",
|
| 27 |
"qalora_group_size": 16,
|
| 28 |
+
"r": 24,
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"o_proj",
|
| 33 |
"k_proj",
|
| 34 |
"v_proj",
|
| 35 |
+
"q_proj"
|
|
|
|
| 36 |
],
|
| 37 |
"target_parameters": null,
|
| 38 |
"task_type": "CAUSAL_LM",
|
checkpoint-8685/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8b4ecb107db701acdc04f96300149f10454a4f22cc800cab0b968eae74c3415
|
| 3 |
+
size 26182176
|
checkpoint-8685/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1299200e098e830dd07921adbf8a60a6476e8cf36d7b7f707c1a922d4319d4d
|
| 3 |
+
size 52486155
|
checkpoint-8685/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:400a0ef098a3a7945e367fce95960239dc97197e53382a77b353016a149f9f93
|
| 3 |
size 14645
|
checkpoint-8685/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22e56af9f486c216b1657390be7da71b3945baf2e0925cfac7cc23d69d3cd231
|
| 3 |
size 1465
|
checkpoint-8685/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 8685,
|
| 3 |
-
"best_metric": 5.
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-8685",
|
| 5 |
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,1800 +10,1800 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"entropy": 3.
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
-
"grad_norm":
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
-
"loss": 13.
|
| 18 |
-
"mean_token_accuracy": 0.
|
| 19 |
-
"num_tokens":
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"entropy": 3.
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
-
"grad_norm":
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
-
"loss": 14.
|
| 28 |
-
"mean_token_accuracy": 0.
|
| 29 |
-
"num_tokens":
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"entropy": 3.
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
-
"grad_norm":
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
-
"loss":
|
| 38 |
-
"mean_token_accuracy": 0.
|
| 39 |
-
"num_tokens":
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"entropy": 3.
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
-
"grad_norm":
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
-
"loss": 13.
|
| 48 |
-
"mean_token_accuracy": 0.
|
| 49 |
-
"num_tokens":
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"entropy": 3.
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
-
"mean_token_accuracy": 0.
|
| 59 |
-
"num_tokens":
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"entropy":
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
-
"grad_norm":
|
| 66 |
-
"learning_rate": 1.
|
| 67 |
-
"loss": 12.
|
| 68 |
-
"mean_token_accuracy": 0.
|
| 69 |
-
"num_tokens":
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
-
"entropy": 4.
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate": 1.
|
| 77 |
-
"loss": 12.
|
| 78 |
-
"mean_token_accuracy": 0.
|
| 79 |
-
"num_tokens":
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"entropy": 4.
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 1.
|
| 87 |
-
"loss":
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"entropy":
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate": 1.
|
| 97 |
-
"loss":
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"entropy": 5.
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 1.
|
| 107 |
-
"loss": 9.
|
| 108 |
-
"mean_token_accuracy": 0.
|
| 109 |
-
"num_tokens":
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"entropy": 6.
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate": 1.
|
| 117 |
-
"loss": 8.
|
| 118 |
-
"mean_token_accuracy": 0.
|
| 119 |
-
"num_tokens":
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"entropy": 6.
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 1.
|
| 127 |
-
"loss": 7.
|
| 128 |
-
"mean_token_accuracy": 0.
|
| 129 |
-
"num_tokens":
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"entropy": 6.
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 1.
|
| 137 |
-
"loss": 6.
|
| 138 |
-
"mean_token_accuracy": 0.
|
| 139 |
-
"num_tokens":
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"entropy": 6.
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
-
"grad_norm":
|
| 146 |
-
"learning_rate": 1.
|
| 147 |
-
"loss": 6.
|
| 148 |
-
"mean_token_accuracy": 0.
|
| 149 |
-
"num_tokens":
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"entropy": 6.
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
-
"grad_norm": 4.
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 6.
|
| 158 |
-
"mean_token_accuracy": 0.
|
| 159 |
-
"num_tokens":
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"entropy": 6.
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
-
"grad_norm": 3.
|
| 166 |
-
"learning_rate": 1.
|
| 167 |
-
"loss": 6.
|
| 168 |
-
"mean_token_accuracy": 0.
|
| 169 |
-
"num_tokens":
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"entropy": 6.
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate": 1.
|
| 177 |
-
"loss": 6.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"entropy": 6.
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 1.
|
| 187 |
-
"loss": 6.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"entropy": 6.
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate": 1.
|
| 197 |
-
"loss":
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"entropy": 6.
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
-
"grad_norm":
|
| 206 |
-
"learning_rate": 1.
|
| 207 |
-
"loss": 6.
|
| 208 |
-
"mean_token_accuracy": 0.
|
| 209 |
-
"num_tokens":
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"entropy": 6.
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
-
"grad_norm":
|
| 216 |
-
"learning_rate": 1.
|
| 217 |
-
"loss": 6.
|
| 218 |
-
"mean_token_accuracy": 0.
|
| 219 |
-
"num_tokens":
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"entropy": 6.
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 1.
|
| 227 |
-
"loss": 5.
|
| 228 |
-
"mean_token_accuracy": 0.
|
| 229 |
-
"num_tokens":
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"entropy": 5.
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate": 1.
|
| 237 |
-
"loss": 5.
|
| 238 |
-
"mean_token_accuracy": 0.
|
| 239 |
-
"num_tokens":
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"entropy": 6.
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 1.
|
| 247 |
-
"loss": 5.
|
| 248 |
-
"mean_token_accuracy": 0.
|
| 249 |
-
"num_tokens":
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"entropy": 6.
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
-
"grad_norm":
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 5.
|
| 258 |
-
"mean_token_accuracy": 0.
|
| 259 |
-
"num_tokens":
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"entropy": 6.
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate": 1.
|
| 267 |
-
"loss": 5.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"entropy": 5.
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
-
"grad_norm":
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 5.
|
| 278 |
-
"mean_token_accuracy": 0.
|
| 279 |
-
"num_tokens":
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"entropy": 6.
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
-
"grad_norm":
|
| 286 |
-
"learning_rate": 1.
|
| 287 |
-
"loss": 5.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"entropy":
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 1.
|
| 297 |
-
"loss": 5.
|
| 298 |
-
"mean_token_accuracy": 0.
|
| 299 |
-
"num_tokens":
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"entropy": 5.
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 5.
|
| 308 |
-
"mean_token_accuracy": 0.
|
| 309 |
-
"num_tokens":
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"entropy":
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 1.
|
| 317 |
-
"loss": 5.
|
| 318 |
-
"mean_token_accuracy": 0.
|
| 319 |
-
"num_tokens":
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"entropy": 5.
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
-
"grad_norm":
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 5.
|
| 328 |
-
"mean_token_accuracy": 0.
|
| 329 |
-
"num_tokens":
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"entropy":
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
-
"grad_norm":
|
| 336 |
-
"learning_rate": 1.
|
| 337 |
-
"loss": 5.
|
| 338 |
-
"mean_token_accuracy": 0.
|
| 339 |
-
"num_tokens":
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"entropy": 5.
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 1.
|
| 347 |
-
"loss": 5.
|
| 348 |
-
"mean_token_accuracy": 0.
|
| 349 |
-
"num_tokens":
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
-
"eval_entropy":
|
| 355 |
-
"eval_loss": 5.
|
| 356 |
-
"eval_mean_token_accuracy": 0.
|
| 357 |
-
"eval_model_preparation_time": 0.
|
| 358 |
-
"eval_num_tokens":
|
| 359 |
-
"eval_runtime":
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second":
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"entropy": 5.
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
-
"grad_norm":
|
| 368 |
-
"learning_rate": 1.
|
| 369 |
-
"loss": 5.
|
| 370 |
-
"mean_token_accuracy": 0.
|
| 371 |
-
"num_tokens":
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"entropy": 5.
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 1.
|
| 379 |
-
"loss": 5.
|
| 380 |
-
"mean_token_accuracy": 0.
|
| 381 |
-
"num_tokens":
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"entropy": 5.
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 1.
|
| 389 |
-
"loss": 5.
|
| 390 |
-
"mean_token_accuracy": 0.
|
| 391 |
-
"num_tokens":
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"entropy": 5.
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
-
"grad_norm":
|
| 398 |
-
"learning_rate": 1.
|
| 399 |
-
"loss": 5.
|
| 400 |
-
"mean_token_accuracy": 0.
|
| 401 |
-
"num_tokens":
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"entropy": 5.
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 1.
|
| 409 |
-
"loss": 5.
|
| 410 |
-
"mean_token_accuracy": 0.
|
| 411 |
-
"num_tokens":
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"entropy": 5.
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
-
"grad_norm":
|
| 418 |
-
"learning_rate": 1.
|
| 419 |
-
"loss": 5.
|
| 420 |
-
"mean_token_accuracy": 0.
|
| 421 |
-
"num_tokens":
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"entropy": 5.
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 1.
|
| 429 |
-
"loss": 5.
|
| 430 |
-
"mean_token_accuracy": 0.
|
| 431 |
-
"num_tokens":
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"entropy": 5.
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss": 5.
|
| 440 |
-
"mean_token_accuracy": 0.
|
| 441 |
-
"num_tokens":
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
-
"entropy": 5.
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
-
"grad_norm":
|
| 448 |
-
"learning_rate": 1.
|
| 449 |
-
"loss": 5.
|
| 450 |
-
"mean_token_accuracy": 0.
|
| 451 |
-
"num_tokens":
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"entropy": 5.
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
-
"grad_norm":
|
| 458 |
-
"learning_rate": 1.
|
| 459 |
-
"loss": 5.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
-
"entropy": 5.
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
-
"grad_norm":
|
| 468 |
-
"learning_rate": 1.
|
| 469 |
-
"loss": 5.
|
| 470 |
-
"mean_token_accuracy": 0.
|
| 471 |
-
"num_tokens":
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"entropy": 5.
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 1.
|
| 479 |
-
"loss": 5.
|
| 480 |
-
"mean_token_accuracy": 0.
|
| 481 |
-
"num_tokens":
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"entropy": 5.
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
-
"grad_norm":
|
| 488 |
-
"learning_rate": 1.
|
| 489 |
-
"loss": 5.
|
| 490 |
-
"mean_token_accuracy": 0.
|
| 491 |
-
"num_tokens":
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"entropy": 5.
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 1.
|
| 499 |
-
"loss": 5.
|
| 500 |
-
"mean_token_accuracy": 0.
|
| 501 |
-
"num_tokens":
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"entropy": 5.
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate": 1.
|
| 509 |
-
"loss": 5.
|
| 510 |
-
"mean_token_accuracy": 0.
|
| 511 |
-
"num_tokens":
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"entropy": 5.
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
-
"grad_norm":
|
| 518 |
-
"learning_rate": 1.
|
| 519 |
-
"loss": 5.
|
| 520 |
-
"mean_token_accuracy": 0.
|
| 521 |
-
"num_tokens":
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"entropy": 5.
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
-
"grad_norm":
|
| 528 |
-
"learning_rate": 1.
|
| 529 |
-
"loss": 5.
|
| 530 |
-
"mean_token_accuracy": 0.
|
| 531 |
-
"num_tokens":
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"entropy": 5.
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
-
"grad_norm":
|
| 538 |
-
"learning_rate": 1.
|
| 539 |
-
"loss": 5.
|
| 540 |
-
"mean_token_accuracy": 0.
|
| 541 |
-
"num_tokens":
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"entropy": 5.
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
-
"grad_norm":
|
| 548 |
-
"learning_rate": 1.
|
| 549 |
-
"loss": 5.
|
| 550 |
-
"mean_token_accuracy": 0.
|
| 551 |
-
"num_tokens":
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"entropy": 5.
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
-
"grad_norm":
|
| 558 |
-
"learning_rate": 1.
|
| 559 |
-
"loss": 5.
|
| 560 |
-
"mean_token_accuracy": 0.
|
| 561 |
-
"num_tokens":
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"entropy": 5.
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
-
"grad_norm":
|
| 568 |
-
"learning_rate": 1.
|
| 569 |
-
"loss": 5.
|
| 570 |
-
"mean_token_accuracy": 0.
|
| 571 |
-
"num_tokens":
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"entropy": 5.
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate": 1.
|
| 579 |
-
"loss": 5.
|
| 580 |
-
"mean_token_accuracy": 0.
|
| 581 |
-
"num_tokens":
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"entropy": 5.
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
-
"grad_norm":
|
| 588 |
-
"learning_rate": 1.
|
| 589 |
-
"loss": 5.
|
| 590 |
-
"mean_token_accuracy": 0.
|
| 591 |
-
"num_tokens":
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
-
"entropy": 5.
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
-
"grad_norm":
|
| 598 |
-
"learning_rate": 1.
|
| 599 |
-
"loss": 5.
|
| 600 |
-
"mean_token_accuracy": 0.
|
| 601 |
-
"num_tokens":
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"entropy": 5.
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
-
"grad_norm":
|
| 608 |
-
"learning_rate": 1.
|
| 609 |
-
"loss": 5.
|
| 610 |
-
"mean_token_accuracy": 0.
|
| 611 |
-
"num_tokens":
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"entropy": 5.
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
-
"grad_norm":
|
| 618 |
-
"learning_rate": 1.
|
| 619 |
-
"loss": 5.
|
| 620 |
-
"mean_token_accuracy": 0.
|
| 621 |
-
"num_tokens":
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
-
"entropy": 5.
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
-
"grad_norm":
|
| 628 |
-
"learning_rate": 1.
|
| 629 |
-
"loss": 5.
|
| 630 |
-
"mean_token_accuracy": 0.
|
| 631 |
-
"num_tokens":
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"entropy": 5.
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
-
"grad_norm": 2.
|
| 638 |
-
"learning_rate": 1.
|
| 639 |
-
"loss": 5.
|
| 640 |
-
"mean_token_accuracy": 0.
|
| 641 |
-
"num_tokens":
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"entropy": 5.
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate": 1.
|
| 649 |
-
"loss":
|
| 650 |
-
"mean_token_accuracy": 0.
|
| 651 |
-
"num_tokens":
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
-
"entropy": 5.
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
-
"grad_norm":
|
| 658 |
-
"learning_rate": 1.
|
| 659 |
-
"loss": 5.
|
| 660 |
-
"mean_token_accuracy": 0.
|
| 661 |
-
"num_tokens":
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"entropy": 5.
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
-
"grad_norm":
|
| 668 |
-
"learning_rate": 1.
|
| 669 |
-
"loss": 5.
|
| 670 |
-
"mean_token_accuracy": 0.
|
| 671 |
-
"num_tokens":
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"entropy": 5.
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
-
"grad_norm": 2.
|
| 678 |
-
"learning_rate": 1.
|
| 679 |
-
"loss": 5.
|
| 680 |
-
"mean_token_accuracy": 0.
|
| 681 |
-
"num_tokens":
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"entropy": 5.
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
-
"grad_norm":
|
| 688 |
-
"learning_rate": 1.
|
| 689 |
-
"loss": 5.
|
| 690 |
-
"mean_token_accuracy": 0.
|
| 691 |
-
"num_tokens":
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"entropy": 5.
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
-
"grad_norm":
|
| 698 |
-
"learning_rate": 1.
|
| 699 |
-
"loss": 5.
|
| 700 |
-
"mean_token_accuracy": 0.
|
| 701 |
-
"num_tokens":
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"entropy": 5.
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate": 1.
|
| 709 |
-
"loss": 5.
|
| 710 |
-
"mean_token_accuracy": 0.
|
| 711 |
-
"num_tokens":
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
-
"eval_entropy": 5.
|
| 717 |
-
"eval_loss": 5.
|
| 718 |
-
"eval_mean_token_accuracy": 0.
|
| 719 |
-
"eval_model_preparation_time": 0.
|
| 720 |
-
"eval_num_tokens":
|
| 721 |
-
"eval_runtime":
|
| 722 |
-
"eval_samples_per_second":
|
| 723 |
-
"eval_steps_per_second":
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"entropy": 5.
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
-
"grad_norm":
|
| 730 |
-
"learning_rate": 1.
|
| 731 |
-
"loss": 5.
|
| 732 |
-
"mean_token_accuracy": 0.
|
| 733 |
-
"num_tokens":
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"entropy": 5.
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
-
"grad_norm":
|
| 740 |
-
"learning_rate": 1.
|
| 741 |
-
"loss": 5.
|
| 742 |
-
"mean_token_accuracy": 0.
|
| 743 |
-
"num_tokens":
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"entropy": 5.
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate": 1.
|
| 751 |
-
"loss": 5.
|
| 752 |
-
"mean_token_accuracy": 0.
|
| 753 |
-
"num_tokens":
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"entropy": 5.
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
-
"grad_norm": 2.
|
| 760 |
-
"learning_rate": 1.
|
| 761 |
-
"loss": 5.
|
| 762 |
-
"mean_token_accuracy": 0.
|
| 763 |
-
"num_tokens":
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"entropy": 5.
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate": 1.
|
| 771 |
-
"loss": 5.
|
| 772 |
-
"mean_token_accuracy": 0.
|
| 773 |
-
"num_tokens":
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"entropy": 5.
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
-
"grad_norm": 2.
|
| 780 |
-
"learning_rate": 1.
|
| 781 |
-
"loss": 5.
|
| 782 |
-
"mean_token_accuracy": 0.
|
| 783 |
-
"num_tokens":
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
-
"entropy": 5.
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
-
"grad_norm":
|
| 790 |
-
"learning_rate": 1.
|
| 791 |
-
"loss": 5.
|
| 792 |
-
"mean_token_accuracy": 0.
|
| 793 |
-
"num_tokens":
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"entropy": 5.
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
-
"grad_norm":
|
| 800 |
-
"learning_rate": 1.
|
| 801 |
-
"loss": 5.
|
| 802 |
-
"mean_token_accuracy": 0.
|
| 803 |
-
"num_tokens":
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
-
"entropy": 5.
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
-
"grad_norm":
|
| 810 |
-
"learning_rate": 1.
|
| 811 |
-
"loss": 5.
|
| 812 |
-
"mean_token_accuracy": 0.
|
| 813 |
-
"num_tokens":
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"entropy": 5.
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate": 1.
|
| 821 |
-
"loss": 5.
|
| 822 |
-
"mean_token_accuracy": 0.
|
| 823 |
-
"num_tokens":
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
-
"entropy": 5.
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
-
"grad_norm":
|
| 830 |
-
"learning_rate": 1.
|
| 831 |
-
"loss": 5.
|
| 832 |
-
"mean_token_accuracy": 0.
|
| 833 |
-
"num_tokens":
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
-
"entropy": 5.
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
-
"grad_norm":
|
| 840 |
-
"learning_rate": 1.
|
| 841 |
-
"loss": 5.
|
| 842 |
-
"mean_token_accuracy": 0.
|
| 843 |
-
"num_tokens":
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
-
"entropy": 5.
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
-
"grad_norm":
|
| 850 |
-
"learning_rate": 1.
|
| 851 |
-
"loss": 5.
|
| 852 |
-
"mean_token_accuracy": 0.
|
| 853 |
-
"num_tokens":
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
-
"entropy": 5.
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
-
"grad_norm":
|
| 860 |
-
"learning_rate": 1.
|
| 861 |
-
"loss": 5.
|
| 862 |
-
"mean_token_accuracy": 0.
|
| 863 |
-
"num_tokens":
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"entropy": 5.
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
-
"grad_norm": 1.
|
| 870 |
-
"learning_rate": 1.
|
| 871 |
-
"loss": 5.
|
| 872 |
-
"mean_token_accuracy": 0.
|
| 873 |
-
"num_tokens":
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
-
"entropy": 5.
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
-
"grad_norm": 1.
|
| 880 |
-
"learning_rate": 1.
|
| 881 |
-
"loss": 5.
|
| 882 |
-
"mean_token_accuracy": 0.
|
| 883 |
-
"num_tokens":
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"entropy": 5.
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
-
"grad_norm":
|
| 890 |
-
"learning_rate": 1.
|
| 891 |
-
"loss": 5.
|
| 892 |
-
"mean_token_accuracy": 0.
|
| 893 |
-
"num_tokens":
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
-
"entropy": 5.
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
-
"grad_norm":
|
| 900 |
-
"learning_rate": 1.
|
| 901 |
-
"loss": 5.
|
| 902 |
-
"mean_token_accuracy": 0.
|
| 903 |
-
"num_tokens":
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
-
"entropy": 5.
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate": 1.
|
| 911 |
-
"loss": 5.
|
| 912 |
-
"mean_token_accuracy": 0.
|
| 913 |
-
"num_tokens":
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
-
"entropy": 5.
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
-
"grad_norm":
|
| 920 |
-
"learning_rate":
|
| 921 |
-
"loss": 5.
|
| 922 |
-
"mean_token_accuracy": 0.
|
| 923 |
-
"num_tokens":
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"entropy": 5.
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
-
"grad_norm":
|
| 930 |
-
"learning_rate":
|
| 931 |
-
"loss": 5.
|
| 932 |
-
"mean_token_accuracy": 0.
|
| 933 |
-
"num_tokens":
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
-
"entropy": 5.
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
-
"grad_norm":
|
| 940 |
-
"learning_rate":
|
| 941 |
-
"loss": 5.
|
| 942 |
-
"mean_token_accuracy": 0.
|
| 943 |
-
"num_tokens":
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
-
"entropy": 5.
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
-
"grad_norm":
|
| 950 |
-
"learning_rate":
|
| 951 |
-
"loss": 5.
|
| 952 |
-
"mean_token_accuracy": 0.
|
| 953 |
-
"num_tokens":
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"entropy": 5.
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
-
"grad_norm":
|
| 960 |
-
"learning_rate":
|
| 961 |
-
"loss": 5.
|
| 962 |
-
"mean_token_accuracy": 0.
|
| 963 |
-
"num_tokens":
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
-
"entropy": 5.
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
-
"grad_norm":
|
| 970 |
-
"learning_rate":
|
| 971 |
-
"loss": 5.
|
| 972 |
-
"mean_token_accuracy": 0.
|
| 973 |
-
"num_tokens":
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"entropy": 5.
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
-
"grad_norm":
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 5.
|
| 982 |
-
"mean_token_accuracy": 0.
|
| 983 |
-
"num_tokens":
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
-
"entropy": 5.
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
-
"grad_norm":
|
| 990 |
-
"learning_rate":
|
| 991 |
-
"loss": 5.
|
| 992 |
-
"mean_token_accuracy": 0.
|
| 993 |
-
"num_tokens":
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
-
"entropy": 5.
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
-
"grad_norm":
|
| 1000 |
-
"learning_rate":
|
| 1001 |
-
"loss": 5.
|
| 1002 |
-
"mean_token_accuracy": 0.
|
| 1003 |
-
"num_tokens":
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"entropy": 5.
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
-
"grad_norm":
|
| 1010 |
-
"learning_rate":
|
| 1011 |
-
"loss": 5.
|
| 1012 |
-
"mean_token_accuracy": 0.
|
| 1013 |
-
"num_tokens":
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
-
"entropy": 5.
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
-
"grad_norm":
|
| 1020 |
-
"learning_rate":
|
| 1021 |
-
"loss": 5.
|
| 1022 |
-
"mean_token_accuracy": 0.
|
| 1023 |
-
"num_tokens":
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"entropy": 5.
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
-
"grad_norm":
|
| 1030 |
-
"learning_rate":
|
| 1031 |
-
"loss": 5.
|
| 1032 |
-
"mean_token_accuracy": 0.
|
| 1033 |
-
"num_tokens":
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
-
"entropy": 5.
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
-
"grad_norm":
|
| 1040 |
-
"learning_rate":
|
| 1041 |
-
"loss": 5.
|
| 1042 |
-
"mean_token_accuracy": 0.
|
| 1043 |
-
"num_tokens":
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"entropy": 5.
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
-
"grad_norm":
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 5.
|
| 1052 |
-
"mean_token_accuracy": 0.
|
| 1053 |
-
"num_tokens":
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
-
"entropy": 5.
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
-
"grad_norm":
|
| 1060 |
-
"learning_rate":
|
| 1061 |
-
"loss": 5.
|
| 1062 |
-
"mean_token_accuracy": 0.
|
| 1063 |
-
"num_tokens":
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
-
"entropy": 5.
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
-
"grad_norm": 1.
|
| 1070 |
-
"learning_rate":
|
| 1071 |
-
"loss": 5.
|
| 1072 |
-
"mean_token_accuracy": 0.
|
| 1073 |
-
"num_tokens":
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
-
"eval_entropy": 5.
|
| 1079 |
-
"eval_loss": 5.
|
| 1080 |
-
"eval_mean_token_accuracy": 0.
|
| 1081 |
-
"eval_model_preparation_time": 0.
|
| 1082 |
-
"eval_num_tokens":
|
| 1083 |
-
"eval_runtime":
|
| 1084 |
-
"eval_samples_per_second":
|
| 1085 |
-
"eval_steps_per_second":
|
| 1086 |
"step": 5211
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
-
"entropy": 5.
|
| 1090 |
"epoch": 3.0224525043177892,
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 5.
|
| 1094 |
-
"mean_token_accuracy": 0.
|
| 1095 |
-
"num_tokens":
|
| 1096 |
"step": 5250
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
-
"entropy": 5.
|
| 1100 |
"epoch": 3.051237766263673,
|
| 1101 |
-
"grad_norm":
|
| 1102 |
-
"learning_rate":
|
| 1103 |
-
"loss": 5.
|
| 1104 |
-
"mean_token_accuracy": 0.
|
| 1105 |
-
"num_tokens":
|
| 1106 |
"step": 5300
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
-
"entropy": 5.
|
| 1110 |
"epoch": 3.0800230282095566,
|
| 1111 |
-
"grad_norm": 2.
|
| 1112 |
-
"learning_rate":
|
| 1113 |
-
"loss": 5.
|
| 1114 |
-
"mean_token_accuracy": 0.
|
| 1115 |
-
"num_tokens":
|
| 1116 |
"step": 5350
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
-
"entropy": 5.
|
| 1120 |
"epoch": 3.1088082901554404,
|
| 1121 |
-
"grad_norm": 1.
|
| 1122 |
-
"learning_rate":
|
| 1123 |
-
"loss": 5.
|
| 1124 |
-
"mean_token_accuracy": 0.
|
| 1125 |
-
"num_tokens":
|
| 1126 |
"step": 5400
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
-
"entropy": 5.
|
| 1130 |
"epoch": 3.137593552101324,
|
| 1131 |
-
"grad_norm": 1.
|
| 1132 |
-
"learning_rate":
|
| 1133 |
-
"loss": 5.
|
| 1134 |
-
"mean_token_accuracy": 0.
|
| 1135 |
-
"num_tokens":
|
| 1136 |
"step": 5450
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"entropy": 5.
|
| 1140 |
"epoch": 3.166378814047208,
|
| 1141 |
-
"grad_norm": 1.
|
| 1142 |
-
"learning_rate":
|
| 1143 |
-
"loss": 5.
|
| 1144 |
-
"mean_token_accuracy": 0.
|
| 1145 |
-
"num_tokens":
|
| 1146 |
"step": 5500
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
-
"entropy": 5.
|
| 1150 |
"epoch": 3.1951640759930915,
|
| 1151 |
-
"grad_norm":
|
| 1152 |
-
"learning_rate":
|
| 1153 |
-
"loss": 5.
|
| 1154 |
-
"mean_token_accuracy": 0.
|
| 1155 |
-
"num_tokens":
|
| 1156 |
"step": 5550
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
-
"entropy": 5.
|
| 1160 |
"epoch": 3.223949337938975,
|
| 1161 |
-
"grad_norm":
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 5.
|
| 1164 |
-
"mean_token_accuracy": 0.
|
| 1165 |
-
"num_tokens":
|
| 1166 |
"step": 5600
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
-
"entropy": 5.
|
| 1170 |
"epoch": 3.252734599884859,
|
| 1171 |
-
"grad_norm":
|
| 1172 |
-
"learning_rate":
|
| 1173 |
-
"loss": 5.
|
| 1174 |
-
"mean_token_accuracy": 0.
|
| 1175 |
-
"num_tokens":
|
| 1176 |
"step": 5650
|
| 1177 |
},
|
| 1178 |
{
|
| 1179 |
-
"entropy": 5.
|
| 1180 |
"epoch": 3.2815198618307426,
|
| 1181 |
-
"grad_norm":
|
| 1182 |
-
"learning_rate":
|
| 1183 |
-
"loss": 5.
|
| 1184 |
-
"mean_token_accuracy": 0.
|
| 1185 |
-
"num_tokens":
|
| 1186 |
"step": 5700
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
-
"entropy": 5.
|
| 1190 |
"epoch": 3.3103051237766263,
|
| 1191 |
-
"grad_norm":
|
| 1192 |
-
"learning_rate":
|
| 1193 |
-
"loss": 5.
|
| 1194 |
-
"mean_token_accuracy": 0.
|
| 1195 |
-
"num_tokens":
|
| 1196 |
"step": 5750
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
-
"entropy": 5.
|
| 1200 |
"epoch": 3.33909038572251,
|
| 1201 |
-
"grad_norm":
|
| 1202 |
-
"learning_rate":
|
| 1203 |
-
"loss": 5.
|
| 1204 |
-
"mean_token_accuracy": 0.
|
| 1205 |
-
"num_tokens":
|
| 1206 |
"step": 5800
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
-
"entropy": 5.
|
| 1210 |
"epoch": 3.3678756476683938,
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss":
|
| 1214 |
-
"mean_token_accuracy": 0.
|
| 1215 |
-
"num_tokens":
|
| 1216 |
"step": 5850
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
-
"entropy": 5.
|
| 1220 |
"epoch": 3.3966609096142775,
|
| 1221 |
-
"grad_norm": 1.
|
| 1222 |
-
"learning_rate":
|
| 1223 |
-
"loss": 5.
|
| 1224 |
-
"mean_token_accuracy": 0.
|
| 1225 |
-
"num_tokens":
|
| 1226 |
"step": 5900
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
-
"entropy": 5.
|
| 1230 |
"epoch": 3.425446171560161,
|
| 1231 |
-
"grad_norm":
|
| 1232 |
-
"learning_rate":
|
| 1233 |
-
"loss": 5.
|
| 1234 |
-
"mean_token_accuracy": 0.
|
| 1235 |
-
"num_tokens":
|
| 1236 |
"step": 5950
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
-
"entropy": 5.
|
| 1240 |
"epoch": 3.454231433506045,
|
| 1241 |
-
"grad_norm":
|
| 1242 |
-
"learning_rate":
|
| 1243 |
-
"loss": 5.
|
| 1244 |
-
"mean_token_accuracy": 0.
|
| 1245 |
-
"num_tokens":
|
| 1246 |
"step": 6000
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
-
"entropy": 5.
|
| 1250 |
"epoch": 3.4830166954519286,
|
| 1251 |
-
"grad_norm":
|
| 1252 |
-
"learning_rate":
|
| 1253 |
-
"loss": 5.
|
| 1254 |
-
"mean_token_accuracy": 0.
|
| 1255 |
-
"num_tokens":
|
| 1256 |
"step": 6050
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
-
"entropy": 5.
|
| 1260 |
"epoch": 3.5118019573978123,
|
| 1261 |
-
"grad_norm":
|
| 1262 |
-
"learning_rate":
|
| 1263 |
-
"loss": 5.
|
| 1264 |
-
"mean_token_accuracy": 0.
|
| 1265 |
-
"num_tokens":
|
| 1266 |
"step": 6100
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
-
"entropy": 5.
|
| 1270 |
"epoch": 3.540587219343696,
|
| 1271 |
-
"grad_norm": 1.
|
| 1272 |
-
"learning_rate":
|
| 1273 |
-
"loss": 5.
|
| 1274 |
-
"mean_token_accuracy": 0.
|
| 1275 |
-
"num_tokens":
|
| 1276 |
"step": 6150
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
-
"entropy": 5.
|
| 1280 |
"epoch": 3.5693724812895797,
|
| 1281 |
-
"grad_norm":
|
| 1282 |
-
"learning_rate":
|
| 1283 |
-
"loss": 5.
|
| 1284 |
-
"mean_token_accuracy": 0.
|
| 1285 |
-
"num_tokens":
|
| 1286 |
"step": 6200
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
-
"entropy": 5.
|
| 1290 |
"epoch": 3.5981577432354634,
|
| 1291 |
-
"grad_norm":
|
| 1292 |
-
"learning_rate":
|
| 1293 |
-
"loss": 5.
|
| 1294 |
-
"mean_token_accuracy": 0.
|
| 1295 |
-
"num_tokens":
|
| 1296 |
"step": 6250
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
-
"entropy": 5.
|
| 1300 |
"epoch": 3.626943005181347,
|
| 1301 |
-
"grad_norm":
|
| 1302 |
-
"learning_rate":
|
| 1303 |
-
"loss": 5.
|
| 1304 |
-
"mean_token_accuracy": 0.
|
| 1305 |
-
"num_tokens":
|
| 1306 |
"step": 6300
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
-
"entropy": 5.
|
| 1310 |
"epoch": 3.655728267127231,
|
| 1311 |
-
"grad_norm": 1.
|
| 1312 |
-
"learning_rate":
|
| 1313 |
-
"loss": 5.
|
| 1314 |
-
"mean_token_accuracy": 0.
|
| 1315 |
-
"num_tokens":
|
| 1316 |
"step": 6350
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
-
"entropy": 5.
|
| 1320 |
"epoch": 3.6845135290731146,
|
| 1321 |
-
"grad_norm":
|
| 1322 |
-
"learning_rate":
|
| 1323 |
-
"loss": 5.
|
| 1324 |
-
"mean_token_accuracy": 0.
|
| 1325 |
-
"num_tokens":
|
| 1326 |
"step": 6400
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
-
"entropy": 5.
|
| 1330 |
"epoch": 3.7132987910189983,
|
| 1331 |
-
"grad_norm":
|
| 1332 |
-
"learning_rate":
|
| 1333 |
-
"loss": 5.
|
| 1334 |
-
"mean_token_accuracy": 0.
|
| 1335 |
-
"num_tokens":
|
| 1336 |
"step": 6450
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
-
"entropy": 5.
|
| 1340 |
"epoch": 3.742084052964882,
|
| 1341 |
-
"grad_norm":
|
| 1342 |
-
"learning_rate":
|
| 1343 |
-
"loss": 5.
|
| 1344 |
-
"mean_token_accuracy": 0.
|
| 1345 |
-
"num_tokens":
|
| 1346 |
"step": 6500
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
-
"entropy": 5.
|
| 1350 |
"epoch": 3.7708693149107657,
|
| 1351 |
-
"grad_norm": 2.
|
| 1352 |
-
"learning_rate":
|
| 1353 |
-
"loss": 5.
|
| 1354 |
-
"mean_token_accuracy": 0.
|
| 1355 |
-
"num_tokens":
|
| 1356 |
"step": 6550
|
| 1357 |
},
|
| 1358 |
{
|
| 1359 |
-
"entropy": 5.
|
| 1360 |
"epoch": 3.7996545768566494,
|
| 1361 |
-
"grad_norm":
|
| 1362 |
-
"learning_rate":
|
| 1363 |
-
"loss": 5.
|
| 1364 |
-
"mean_token_accuracy": 0.
|
| 1365 |
-
"num_tokens":
|
| 1366 |
"step": 6600
|
| 1367 |
},
|
| 1368 |
{
|
| 1369 |
-
"entropy": 5.
|
| 1370 |
"epoch": 3.828439838802533,
|
| 1371 |
-
"grad_norm": 2.
|
| 1372 |
-
"learning_rate":
|
| 1373 |
-
"loss": 5.
|
| 1374 |
-
"mean_token_accuracy": 0.
|
| 1375 |
-
"num_tokens":
|
| 1376 |
"step": 6650
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
-
"entropy": 5.
|
| 1380 |
"epoch": 3.857225100748417,
|
| 1381 |
-
"grad_norm":
|
| 1382 |
-
"learning_rate":
|
| 1383 |
-
"loss": 5.
|
| 1384 |
-
"mean_token_accuracy": 0.
|
| 1385 |
-
"num_tokens":
|
| 1386 |
"step": 6700
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
-
"entropy": 5.
|
| 1390 |
"epoch": 3.8860103626943006,
|
| 1391 |
-
"grad_norm":
|
| 1392 |
-
"learning_rate":
|
| 1393 |
-
"loss": 5.
|
| 1394 |
-
"mean_token_accuracy": 0.
|
| 1395 |
-
"num_tokens":
|
| 1396 |
"step": 6750
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
-
"entropy": 5.
|
| 1400 |
"epoch": 3.9147956246401843,
|
| 1401 |
-
"grad_norm": 1.
|
| 1402 |
-
"learning_rate":
|
| 1403 |
-
"loss": 5.
|
| 1404 |
-
"mean_token_accuracy": 0.
|
| 1405 |
-
"num_tokens":
|
| 1406 |
"step": 6800
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
-
"entropy": 5.
|
| 1410 |
"epoch": 3.943580886586068,
|
| 1411 |
-
"grad_norm":
|
| 1412 |
-
"learning_rate":
|
| 1413 |
-
"loss": 5.
|
| 1414 |
-
"mean_token_accuracy": 0.
|
| 1415 |
-
"num_tokens":
|
| 1416 |
"step": 6850
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
-
"entropy": 5.
|
| 1420 |
"epoch": 3.9723661485319517,
|
| 1421 |
-
"grad_norm": 5.
|
| 1422 |
-
"learning_rate":
|
| 1423 |
-
"loss":
|
| 1424 |
-
"mean_token_accuracy": 0.
|
| 1425 |
-
"num_tokens":
|
| 1426 |
"step": 6900
|
| 1427 |
},
|
| 1428 |
{
|
| 1429 |
"epoch": 4.0,
|
| 1430 |
-
"eval_entropy": 5.
|
| 1431 |
-
"eval_loss": 5.
|
| 1432 |
-
"eval_mean_token_accuracy": 0.
|
| 1433 |
-
"eval_model_preparation_time": 0.
|
| 1434 |
-
"eval_num_tokens":
|
| 1435 |
-
"eval_runtime":
|
| 1436 |
-
"eval_samples_per_second":
|
| 1437 |
-
"eval_steps_per_second":
|
| 1438 |
"step": 6948
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
-
"entropy": 5.
|
| 1442 |
"epoch": 4.001151410477835,
|
| 1443 |
-
"grad_norm": 1.
|
| 1444 |
-
"learning_rate":
|
| 1445 |
-
"loss": 5.
|
| 1446 |
-
"mean_token_accuracy": 0.
|
| 1447 |
-
"num_tokens":
|
| 1448 |
"step": 6950
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
-
"entropy": 5.
|
| 1452 |
"epoch": 4.029936672423719,
|
| 1453 |
-
"grad_norm":
|
| 1454 |
-
"learning_rate":
|
| 1455 |
-
"loss":
|
| 1456 |
-
"mean_token_accuracy": 0.
|
| 1457 |
-
"num_tokens":
|
| 1458 |
"step": 7000
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
-
"entropy": 5.
|
| 1462 |
"epoch": 4.058721934369602,
|
| 1463 |
-
"grad_norm":
|
| 1464 |
-
"learning_rate":
|
| 1465 |
-
"loss": 5.
|
| 1466 |
-
"mean_token_accuracy": 0.
|
| 1467 |
-
"num_tokens":
|
| 1468 |
"step": 7050
|
| 1469 |
},
|
| 1470 |
{
|
| 1471 |
-
"entropy": 5.
|
| 1472 |
"epoch": 4.087507196315486,
|
| 1473 |
-
"grad_norm":
|
| 1474 |
-
"learning_rate":
|
| 1475 |
-
"loss": 5.
|
| 1476 |
-
"mean_token_accuracy": 0.
|
| 1477 |
-
"num_tokens":
|
| 1478 |
"step": 7100
|
| 1479 |
},
|
| 1480 |
{
|
| 1481 |
-
"entropy": 5.
|
| 1482 |
"epoch": 4.11629245826137,
|
| 1483 |
-
"grad_norm": 1.
|
| 1484 |
-
"learning_rate":
|
| 1485 |
-
"loss": 5.
|
| 1486 |
-
"mean_token_accuracy": 0.
|
| 1487 |
-
"num_tokens":
|
| 1488 |
"step": 7150
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
-
"entropy": 5.
|
| 1492 |
"epoch": 4.1450777202072535,
|
| 1493 |
-
"grad_norm": 1.
|
| 1494 |
-
"learning_rate":
|
| 1495 |
-
"loss": 5.
|
| 1496 |
-
"mean_token_accuracy": 0.
|
| 1497 |
-
"num_tokens":
|
| 1498 |
"step": 7200
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
-
"entropy": 5.
|
| 1502 |
"epoch": 4.173862982153137,
|
| 1503 |
-
"grad_norm": 1.
|
| 1504 |
-
"learning_rate":
|
| 1505 |
-
"loss": 5.
|
| 1506 |
-
"mean_token_accuracy": 0.
|
| 1507 |
-
"num_tokens":
|
| 1508 |
"step": 7250
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
-
"entropy": 5.
|
| 1512 |
"epoch": 4.202648244099021,
|
| 1513 |
-
"grad_norm":
|
| 1514 |
-
"learning_rate":
|
| 1515 |
-
"loss": 5.
|
| 1516 |
-
"mean_token_accuracy": 0.
|
| 1517 |
-
"num_tokens":
|
| 1518 |
"step": 7300
|
| 1519 |
},
|
| 1520 |
{
|
| 1521 |
-
"entropy": 5.
|
| 1522 |
"epoch": 4.231433506044905,
|
| 1523 |
-
"grad_norm": 1.
|
| 1524 |
-
"learning_rate":
|
| 1525 |
-
"loss": 5.
|
| 1526 |
-
"mean_token_accuracy": 0.
|
| 1527 |
-
"num_tokens":
|
| 1528 |
"step": 7350
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
-
"entropy": 5.
|
| 1532 |
"epoch": 4.260218767990788,
|
| 1533 |
-
"grad_norm":
|
| 1534 |
-
"learning_rate":
|
| 1535 |
-
"loss": 5.
|
| 1536 |
-
"mean_token_accuracy": 0.
|
| 1537 |
-
"num_tokens":
|
| 1538 |
"step": 7400
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
-
"entropy": 5.
|
| 1542 |
"epoch": 4.289004029936672,
|
| 1543 |
-
"grad_norm": 1.
|
| 1544 |
-
"learning_rate":
|
| 1545 |
-
"loss": 5.
|
| 1546 |
-
"mean_token_accuracy": 0.
|
| 1547 |
-
"num_tokens":
|
| 1548 |
"step": 7450
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
-
"entropy": 5.
|
| 1552 |
"epoch": 4.317789291882556,
|
| 1553 |
-
"grad_norm":
|
| 1554 |
-
"learning_rate":
|
| 1555 |
-
"loss": 5.
|
| 1556 |
-
"mean_token_accuracy": 0.
|
| 1557 |
-
"num_tokens":
|
| 1558 |
"step": 7500
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
-
"entropy": 5.
|
| 1562 |
"epoch": 4.3465745538284395,
|
| 1563 |
-
"grad_norm": 1.
|
| 1564 |
-
"learning_rate":
|
| 1565 |
-
"loss": 5.
|
| 1566 |
-
"mean_token_accuracy": 0.
|
| 1567 |
-
"num_tokens":
|
| 1568 |
"step": 7550
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
-
"entropy": 5.
|
| 1572 |
"epoch": 4.375359815774323,
|
| 1573 |
-
"grad_norm":
|
| 1574 |
-
"learning_rate":
|
| 1575 |
-
"loss": 5.
|
| 1576 |
-
"mean_token_accuracy": 0.
|
| 1577 |
-
"num_tokens":
|
| 1578 |
"step": 7600
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
-
"entropy": 5.
|
| 1582 |
"epoch": 4.404145077720207,
|
| 1583 |
-
"grad_norm": 1.
|
| 1584 |
-
"learning_rate":
|
| 1585 |
-
"loss": 5.
|
| 1586 |
-
"mean_token_accuracy": 0.
|
| 1587 |
-
"num_tokens":
|
| 1588 |
"step": 7650
|
| 1589 |
},
|
| 1590 |
{
|
| 1591 |
-
"entropy": 5.
|
| 1592 |
"epoch": 4.432930339666091,
|
| 1593 |
-
"grad_norm":
|
| 1594 |
-
"learning_rate":
|
| 1595 |
-
"loss": 5.
|
| 1596 |
-
"mean_token_accuracy": 0.
|
| 1597 |
-
"num_tokens":
|
| 1598 |
"step": 7700
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
-
"entropy": 5.
|
| 1602 |
"epoch": 4.461715601611974,
|
| 1603 |
-
"grad_norm":
|
| 1604 |
-
"learning_rate":
|
| 1605 |
-
"loss": 5.
|
| 1606 |
-
"mean_token_accuracy": 0.
|
| 1607 |
-
"num_tokens":
|
| 1608 |
"step": 7750
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
-
"entropy": 5.
|
| 1612 |
"epoch": 4.490500863557858,
|
| 1613 |
-
"grad_norm": 1.
|
| 1614 |
-
"learning_rate":
|
| 1615 |
-
"loss": 5.
|
| 1616 |
-
"mean_token_accuracy": 0.
|
| 1617 |
-
"num_tokens":
|
| 1618 |
"step": 7800
|
| 1619 |
},
|
| 1620 |
{
|
| 1621 |
-
"entropy": 5.
|
| 1622 |
"epoch": 4.519286125503742,
|
| 1623 |
-
"grad_norm":
|
| 1624 |
-
"learning_rate":
|
| 1625 |
-
"loss": 5.
|
| 1626 |
-
"mean_token_accuracy": 0.
|
| 1627 |
-
"num_tokens":
|
| 1628 |
"step": 7850
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
-
"entropy": 5.
|
| 1632 |
"epoch": 4.5480713874496255,
|
| 1633 |
-
"grad_norm":
|
| 1634 |
-
"learning_rate":
|
| 1635 |
-
"loss": 5.
|
| 1636 |
-
"mean_token_accuracy": 0.
|
| 1637 |
-
"num_tokens":
|
| 1638 |
"step": 7900
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
-
"entropy": 5.
|
| 1642 |
"epoch": 4.576856649395509,
|
| 1643 |
-
"grad_norm":
|
| 1644 |
-
"learning_rate":
|
| 1645 |
-
"loss": 5.
|
| 1646 |
-
"mean_token_accuracy": 0.
|
| 1647 |
-
"num_tokens":
|
| 1648 |
"step": 7950
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
-
"entropy": 5.
|
| 1652 |
"epoch": 4.605641911341393,
|
| 1653 |
-
"grad_norm":
|
| 1654 |
-
"learning_rate":
|
| 1655 |
-
"loss": 5.
|
| 1656 |
-
"mean_token_accuracy": 0.
|
| 1657 |
-
"num_tokens":
|
| 1658 |
"step": 8000
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
-
"entropy": 5.
|
| 1662 |
"epoch": 4.634427173287277,
|
| 1663 |
-
"grad_norm": 2.
|
| 1664 |
-
"learning_rate":
|
| 1665 |
-
"loss": 5.
|
| 1666 |
-
"mean_token_accuracy": 0.
|
| 1667 |
-
"num_tokens":
|
| 1668 |
"step": 8050
|
| 1669 |
},
|
| 1670 |
{
|
| 1671 |
-
"entropy": 5.
|
| 1672 |
"epoch": 4.66321243523316,
|
| 1673 |
-
"grad_norm": 1.
|
| 1674 |
-
"learning_rate":
|
| 1675 |
-
"loss": 5.
|
| 1676 |
-
"mean_token_accuracy": 0.
|
| 1677 |
-
"num_tokens":
|
| 1678 |
"step": 8100
|
| 1679 |
},
|
| 1680 |
{
|
| 1681 |
-
"entropy": 5.
|
| 1682 |
"epoch": 4.691997697179044,
|
| 1683 |
-
"grad_norm":
|
| 1684 |
-
"learning_rate":
|
| 1685 |
-
"loss": 5.
|
| 1686 |
-
"mean_token_accuracy": 0.
|
| 1687 |
-
"num_tokens":
|
| 1688 |
"step": 8150
|
| 1689 |
},
|
| 1690 |
{
|
| 1691 |
-
"entropy": 5.
|
| 1692 |
"epoch": 4.720782959124928,
|
| 1693 |
-
"grad_norm":
|
| 1694 |
-
"learning_rate":
|
| 1695 |
-
"loss": 5.
|
| 1696 |
-
"mean_token_accuracy": 0.
|
| 1697 |
-
"num_tokens":
|
| 1698 |
"step": 8200
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
-
"entropy": 5.
|
| 1702 |
"epoch": 4.7495682210708114,
|
| 1703 |
-
"grad_norm":
|
| 1704 |
-
"learning_rate":
|
| 1705 |
-
"loss": 5.
|
| 1706 |
-
"mean_token_accuracy": 0.
|
| 1707 |
-
"num_tokens":
|
| 1708 |
"step": 8250
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
-
"entropy": 5.
|
| 1712 |
"epoch": 4.778353483016695,
|
| 1713 |
-
"grad_norm":
|
| 1714 |
-
"learning_rate":
|
| 1715 |
-
"loss": 5.
|
| 1716 |
-
"mean_token_accuracy": 0.
|
| 1717 |
-
"num_tokens":
|
| 1718 |
"step": 8300
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
-
"entropy": 5.
|
| 1722 |
"epoch": 4.807138744962579,
|
| 1723 |
-
"grad_norm": 2.
|
| 1724 |
-
"learning_rate":
|
| 1725 |
-
"loss": 5.
|
| 1726 |
-
"mean_token_accuracy": 0.
|
| 1727 |
-
"num_tokens":
|
| 1728 |
"step": 8350
|
| 1729 |
},
|
| 1730 |
{
|
| 1731 |
-
"entropy": 5.
|
| 1732 |
"epoch": 4.835924006908463,
|
| 1733 |
-
"grad_norm": 1.
|
| 1734 |
-
"learning_rate":
|
| 1735 |
-
"loss": 5.
|
| 1736 |
-
"mean_token_accuracy": 0.
|
| 1737 |
-
"num_tokens":
|
| 1738 |
"step": 8400
|
| 1739 |
},
|
| 1740 |
{
|
| 1741 |
-
"entropy": 5.
|
| 1742 |
"epoch": 4.864709268854346,
|
| 1743 |
-
"grad_norm":
|
| 1744 |
-
"learning_rate":
|
| 1745 |
-
"loss": 5.
|
| 1746 |
-
"mean_token_accuracy": 0.
|
| 1747 |
-
"num_tokens":
|
| 1748 |
"step": 8450
|
| 1749 |
},
|
| 1750 |
{
|
| 1751 |
-
"entropy": 5.
|
| 1752 |
"epoch": 4.89349453080023,
|
| 1753 |
-
"grad_norm":
|
| 1754 |
-
"learning_rate":
|
| 1755 |
-
"loss": 5.
|
| 1756 |
-
"mean_token_accuracy": 0.
|
| 1757 |
-
"num_tokens":
|
| 1758 |
"step": 8500
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
-
"entropy": 5.
|
| 1762 |
"epoch": 4.922279792746114,
|
| 1763 |
-
"grad_norm":
|
| 1764 |
-
"learning_rate": 3.
|
| 1765 |
-
"loss": 5.
|
| 1766 |
-
"mean_token_accuracy": 0.
|
| 1767 |
-
"num_tokens":
|
| 1768 |
"step": 8550
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
-
"entropy": 5.
|
| 1772 |
"epoch": 4.951065054691997,
|
| 1773 |
-
"grad_norm": 1.
|
| 1774 |
-
"learning_rate":
|
| 1775 |
-
"loss": 5.
|
| 1776 |
-
"mean_token_accuracy": 0.
|
| 1777 |
-
"num_tokens":
|
| 1778 |
"step": 8600
|
| 1779 |
},
|
| 1780 |
{
|
| 1781 |
-
"entropy": 5.
|
| 1782 |
"epoch": 4.979850316637881,
|
| 1783 |
-
"grad_norm": 1.
|
| 1784 |
-
"learning_rate":
|
| 1785 |
-
"loss": 5.
|
| 1786 |
-
"mean_token_accuracy": 0.
|
| 1787 |
-
"num_tokens":
|
| 1788 |
"step": 8650
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 5.0,
|
| 1792 |
-
"eval_entropy": 5.
|
| 1793 |
-
"eval_loss": 5.
|
| 1794 |
-
"eval_mean_token_accuracy": 0.
|
| 1795 |
-
"eval_model_preparation_time": 0.
|
| 1796 |
-
"eval_num_tokens":
|
| 1797 |
-
"eval_runtime":
|
| 1798 |
-
"eval_samples_per_second":
|
| 1799 |
-
"eval_steps_per_second":
|
| 1800 |
"step": 8685
|
| 1801 |
}
|
| 1802 |
],
|
| 1803 |
"logging_steps": 50,
|
| 1804 |
-
"max_steps":
|
| 1805 |
"num_input_tokens_seen": 0,
|
| 1806 |
-
"num_train_epochs":
|
| 1807 |
"save_steps": 500,
|
| 1808 |
"stateful_callbacks": {
|
| 1809 |
"TrainerControl": {
|
|
@@ -1812,12 +1812,12 @@
|
|
| 1812 |
"should_evaluate": false,
|
| 1813 |
"should_log": false,
|
| 1814 |
"should_save": true,
|
| 1815 |
-
"should_training_stop":
|
| 1816 |
},
|
| 1817 |
"attributes": {}
|
| 1818 |
}
|
| 1819 |
},
|
| 1820 |
-
"total_flos": 1.
|
| 1821 |
"train_batch_size": 2,
|
| 1822 |
"trial_name": null,
|
| 1823 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 8685,
|
| 3 |
+
"best_metric": 5.523473739624023,
|
| 4 |
"best_model_checkpoint": "./output/checkpoint-8685",
|
| 5 |
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"entropy": 3.606692385673523,
|
| 14 |
"epoch": 0.028785261945883708,
|
| 15 |
+
"grad_norm": 3.2999913692474365,
|
| 16 |
"learning_rate": 4.9e-07,
|
| 17 |
+
"loss": 13.6598,
|
| 18 |
+
"mean_token_accuracy": 0.16028020828962325,
|
| 19 |
+
"num_tokens": 53993.0,
|
| 20 |
"step": 50
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"entropy": 3.618675880432129,
|
| 24 |
"epoch": 0.057570523891767415,
|
| 25 |
+
"grad_norm": 3.101252555847168,
|
| 26 |
"learning_rate": 9.9e-07,
|
| 27 |
+
"loss": 14.0188,
|
| 28 |
+
"mean_token_accuracy": 0.1508466500043869,
|
| 29 |
+
"num_tokens": 110134.0,
|
| 30 |
"step": 100
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"entropy": 3.5215235900878907,
|
| 34 |
"epoch": 0.08635578583765112,
|
| 35 |
+
"grad_norm": 3.513662815093994,
|
| 36 |
"learning_rate": 1.49e-06,
|
| 37 |
+
"loss": 12.8555,
|
| 38 |
+
"mean_token_accuracy": 0.18527640983462335,
|
| 39 |
+
"num_tokens": 160191.0,
|
| 40 |
"step": 150
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"entropy": 3.667909698486328,
|
| 44 |
"epoch": 0.11514104778353483,
|
| 45 |
+
"grad_norm": 4.327610492706299,
|
| 46 |
"learning_rate": 1.99e-06,
|
| 47 |
+
"loss": 13.5394,
|
| 48 |
+
"mean_token_accuracy": 0.157139780074358,
|
| 49 |
+
"num_tokens": 214993.0,
|
| 50 |
"step": 200
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"entropy": 3.768263258934021,
|
| 54 |
"epoch": 0.14392630972941853,
|
| 55 |
+
"grad_norm": 4.290107250213623,
|
| 56 |
+
"learning_rate": 1.988450206246317e-06,
|
| 57 |
+
"loss": 12.8912,
|
| 58 |
+
"mean_token_accuracy": 0.17374794125556947,
|
| 59 |
+
"num_tokens": 268184.0,
|
| 60 |
"step": 250
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"entropy": 3.990619196891785,
|
| 64 |
"epoch": 0.17271157167530224,
|
| 65 |
+
"grad_norm": 4.444278717041016,
|
| 66 |
+
"learning_rate": 1.976664702416028e-06,
|
| 67 |
+
"loss": 12.455,
|
| 68 |
+
"mean_token_accuracy": 0.17780130118131637,
|
| 69 |
+
"num_tokens": 319458.0,
|
| 70 |
"step": 300
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"entropy": 4.162646284103394,
|
| 74 |
"epoch": 0.20149683362118595,
|
| 75 |
+
"grad_norm": 5.615262508392334,
|
| 76 |
+
"learning_rate": 1.9648791985857395e-06,
|
| 77 |
+
"loss": 12.0893,
|
| 78 |
+
"mean_token_accuracy": 0.18191319867968558,
|
| 79 |
+
"num_tokens": 373337.0,
|
| 80 |
"step": 350
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"entropy": 4.532100868225098,
|
| 84 |
"epoch": 0.23028209556706966,
|
| 85 |
+
"grad_norm": 10.074016571044922,
|
| 86 |
+
"learning_rate": 1.9530936947554507e-06,
|
| 87 |
+
"loss": 11.9261,
|
| 88 |
+
"mean_token_accuracy": 0.169477596282959,
|
| 89 |
+
"num_tokens": 427526.0,
|
| 90 |
"step": 400
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"entropy": 4.923871030807495,
|
| 94 |
"epoch": 0.25906735751295334,
|
| 95 |
+
"grad_norm": 16.220163345336914,
|
| 96 |
+
"learning_rate": 1.9413081909251622e-06,
|
| 97 |
+
"loss": 11.0048,
|
| 98 |
+
"mean_token_accuracy": 0.1704501649737358,
|
| 99 |
+
"num_tokens": 480528.0,
|
| 100 |
"step": 450
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"entropy": 5.521005854606629,
|
| 104 |
"epoch": 0.28785261945883706,
|
| 105 |
+
"grad_norm": 29.904008865356445,
|
| 106 |
+
"learning_rate": 1.9295226870948733e-06,
|
| 107 |
+
"loss": 9.6524,
|
| 108 |
+
"mean_token_accuracy": 0.16450899541378022,
|
| 109 |
+
"num_tokens": 535314.0,
|
| 110 |
"step": 500
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"entropy": 6.092623329162597,
|
| 114 |
"epoch": 0.31663788140472077,
|
| 115 |
+
"grad_norm": 17.821575164794922,
|
| 116 |
+
"learning_rate": 1.9177371832645845e-06,
|
| 117 |
+
"loss": 8.1054,
|
| 118 |
+
"mean_token_accuracy": 0.17205011785030366,
|
| 119 |
+
"num_tokens": 588410.0,
|
| 120 |
"step": 550
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"entropy": 6.385262680053711,
|
| 124 |
"epoch": 0.3454231433506045,
|
| 125 |
+
"grad_norm": 5.502202987670898,
|
| 126 |
+
"learning_rate": 1.9059516794342958e-06,
|
| 127 |
+
"loss": 7.4313,
|
| 128 |
+
"mean_token_accuracy": 0.1734227080643177,
|
| 129 |
+
"num_tokens": 641736.0,
|
| 130 |
"step": 600
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"entropy": 6.278562617301941,
|
| 134 |
"epoch": 0.3742084052964882,
|
| 135 |
+
"grad_norm": 5.4657697677612305,
|
| 136 |
+
"learning_rate": 1.8941661756040071e-06,
|
| 137 |
+
"loss": 6.9266,
|
| 138 |
+
"mean_token_accuracy": 0.18680249139666558,
|
| 139 |
+
"num_tokens": 692200.0,
|
| 140 |
"step": 650
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"entropy": 6.553266277313233,
|
| 144 |
"epoch": 0.4029936672423719,
|
| 145 |
+
"grad_norm": 4.955812931060791,
|
| 146 |
+
"learning_rate": 1.8823806717737183e-06,
|
| 147 |
+
"loss": 6.9847,
|
| 148 |
+
"mean_token_accuracy": 0.16679802387952805,
|
| 149 |
+
"num_tokens": 745830.0,
|
| 150 |
"step": 700
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"entropy": 6.470935583114624,
|
| 154 |
"epoch": 0.4317789291882556,
|
| 155 |
+
"grad_norm": 4.198381423950195,
|
| 156 |
+
"learning_rate": 1.8705951679434296e-06,
|
| 157 |
+
"loss": 6.7277,
|
| 158 |
+
"mean_token_accuracy": 0.17847734570503235,
|
| 159 |
+
"num_tokens": 798872.0,
|
| 160 |
"step": 750
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"entropy": 6.5620588779449465,
|
| 164 |
"epoch": 0.4605641911341393,
|
| 165 |
+
"grad_norm": 3.1793746948242188,
|
| 166 |
+
"learning_rate": 1.8588096641131407e-06,
|
| 167 |
+
"loss": 6.7032,
|
| 168 |
+
"mean_token_accuracy": 0.17336134731769562,
|
| 169 |
+
"num_tokens": 853045.0,
|
| 170 |
"step": 800
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"entropy": 6.532204885482788,
|
| 174 |
"epoch": 0.48934945308002303,
|
| 175 |
+
"grad_norm": 3.824537515640259,
|
| 176 |
+
"learning_rate": 1.847024160282852e-06,
|
| 177 |
+
"loss": 6.5762,
|
| 178 |
+
"mean_token_accuracy": 0.1805124071240425,
|
| 179 |
+
"num_tokens": 907679.0,
|
| 180 |
"step": 850
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"entropy": 6.535988225936889,
|
| 184 |
"epoch": 0.5181347150259067,
|
| 185 |
+
"grad_norm": 4.350001811981201,
|
| 186 |
+
"learning_rate": 1.8352386564525632e-06,
|
| 187 |
+
"loss": 6.505,
|
| 188 |
+
"mean_token_accuracy": 0.1842605724930763,
|
| 189 |
+
"num_tokens": 964170.0,
|
| 190 |
"step": 900
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"entropy": 6.204533562660218,
|
| 194 |
"epoch": 0.5469199769717904,
|
| 195 |
+
"grad_norm": 2.193660020828247,
|
| 196 |
+
"learning_rate": 1.8234531526222745e-06,
|
| 197 |
+
"loss": 6.1211,
|
| 198 |
+
"mean_token_accuracy": 0.21968430042266845,
|
| 199 |
+
"num_tokens": 1015909.0,
|
| 200 |
"step": 950
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"entropy": 6.308737449645996,
|
| 204 |
"epoch": 0.5757052389176741,
|
| 205 |
+
"grad_norm": 2.325622320175171,
|
| 206 |
+
"learning_rate": 1.8116676487919857e-06,
|
| 207 |
+
"loss": 6.1653,
|
| 208 |
+
"mean_token_accuracy": 0.21636426240205764,
|
| 209 |
+
"num_tokens": 1068859.0,
|
| 210 |
"step": 1000
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"entropy": 6.332560749053955,
|
| 214 |
"epoch": 0.6044905008635578,
|
| 215 |
+
"grad_norm": 2.0439090728759766,
|
| 216 |
+
"learning_rate": 1.799882144961697e-06,
|
| 217 |
+
"loss": 6.1559,
|
| 218 |
+
"mean_token_accuracy": 0.21859725564718246,
|
| 219 |
+
"num_tokens": 1123202.0,
|
| 220 |
"step": 1050
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"entropy": 6.042124252319336,
|
| 224 |
"epoch": 0.6332757628094415,
|
| 225 |
+
"grad_norm": 3.621903657913208,
|
| 226 |
+
"learning_rate": 1.7880966411314081e-06,
|
| 227 |
+
"loss": 5.8441,
|
| 228 |
+
"mean_token_accuracy": 0.24906315237283708,
|
| 229 |
+
"num_tokens": 1173403.0,
|
| 230 |
"step": 1100
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"entropy": 5.921343173980713,
|
| 234 |
"epoch": 0.6620610247553252,
|
| 235 |
+
"grad_norm": 5.658033847808838,
|
| 236 |
+
"learning_rate": 1.7763111373011195e-06,
|
| 237 |
+
"loss": 5.7104,
|
| 238 |
+
"mean_token_accuracy": 0.2625067520141602,
|
| 239 |
+
"num_tokens": 1225026.0,
|
| 240 |
"step": 1150
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"entropy": 6.093586492538452,
|
| 244 |
"epoch": 0.690846286701209,
|
| 245 |
+
"grad_norm": 2.4292995929718018,
|
| 246 |
+
"learning_rate": 1.7645256334708308e-06,
|
| 247 |
+
"loss": 5.8658,
|
| 248 |
+
"mean_token_accuracy": 0.24842385441064835,
|
| 249 |
+
"num_tokens": 1279013.0,
|
| 250 |
"step": 1200
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"entropy": 6.119112596511841,
|
| 254 |
"epoch": 0.7196315486470927,
|
| 255 |
+
"grad_norm": 3.369384288787842,
|
| 256 |
+
"learning_rate": 1.752740129640542e-06,
|
| 257 |
+
"loss": 5.8784,
|
| 258 |
+
"mean_token_accuracy": 0.24857850253582,
|
| 259 |
+
"num_tokens": 1332547.0,
|
| 260 |
"step": 1250
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"entropy": 6.025163550376892,
|
| 264 |
"epoch": 0.7484168105929764,
|
| 265 |
+
"grad_norm": 2.5110116004943848,
|
| 266 |
+
"learning_rate": 1.7409546258102533e-06,
|
| 267 |
+
"loss": 5.7769,
|
| 268 |
+
"mean_token_accuracy": 0.25835376888513567,
|
| 269 |
+
"num_tokens": 1385192.0,
|
| 270 |
"step": 1300
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"entropy": 5.877259612083435,
|
| 274 |
"epoch": 0.7772020725388601,
|
| 275 |
+
"grad_norm": 2.4179303646087646,
|
| 276 |
+
"learning_rate": 1.7291691219799646e-06,
|
| 277 |
+
"loss": 5.6284,
|
| 278 |
+
"mean_token_accuracy": 0.2756252554059029,
|
| 279 |
+
"num_tokens": 1437071.0,
|
| 280 |
"step": 1350
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"entropy": 6.002246947288513,
|
| 284 |
"epoch": 0.8059873344847438,
|
| 285 |
+
"grad_norm": 3.494359016418457,
|
| 286 |
+
"learning_rate": 1.717383618149676e-06,
|
| 287 |
+
"loss": 5.747,
|
| 288 |
+
"mean_token_accuracy": 0.26462210685014725,
|
| 289 |
+
"num_tokens": 1490818.0,
|
| 290 |
"step": 1400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"entropy": 5.991955623626709,
|
| 294 |
"epoch": 0.8347725964306275,
|
| 295 |
+
"grad_norm": 2.340975761413574,
|
| 296 |
+
"learning_rate": 1.705598114319387e-06,
|
| 297 |
+
"loss": 5.7379,
|
| 298 |
+
"mean_token_accuracy": 0.26444981098175047,
|
| 299 |
+
"num_tokens": 1544997.0,
|
| 300 |
"step": 1450
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"entropy": 5.91768889427185,
|
| 304 |
"epoch": 0.8635578583765112,
|
| 305 |
+
"grad_norm": 2.2394514083862305,
|
| 306 |
+
"learning_rate": 1.6938126104890984e-06,
|
| 307 |
+
"loss": 5.6564,
|
| 308 |
+
"mean_token_accuracy": 0.2730415526032448,
|
| 309 |
+
"num_tokens": 1598302.0,
|
| 310 |
"step": 1500
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"entropy": 5.982716989517212,
|
| 314 |
"epoch": 0.8923431203223949,
|
| 315 |
+
"grad_norm": 1.876839518547058,
|
| 316 |
+
"learning_rate": 1.6820271066588098e-06,
|
| 317 |
+
"loss": 5.7215,
|
| 318 |
+
"mean_token_accuracy": 0.26642445534467696,
|
| 319 |
+
"num_tokens": 1655267.0,
|
| 320 |
"step": 1550
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"entropy": 5.820467872619629,
|
| 324 |
"epoch": 0.9211283822682786,
|
| 325 |
+
"grad_norm": 2.219966173171997,
|
| 326 |
+
"learning_rate": 1.6702416028285209e-06,
|
| 327 |
+
"loss": 5.5555,
|
| 328 |
+
"mean_token_accuracy": 0.2856418335437775,
|
| 329 |
+
"num_tokens": 1709199.0,
|
| 330 |
"step": 1600
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"entropy": 5.996349005699158,
|
| 334 |
"epoch": 0.9499136442141624,
|
| 335 |
+
"grad_norm": 2.247213840484619,
|
| 336 |
+
"learning_rate": 1.6584560989982322e-06,
|
| 337 |
+
"loss": 5.7283,
|
| 338 |
+
"mean_token_accuracy": 0.2696125540137291,
|
| 339 |
+
"num_tokens": 1765443.0,
|
| 340 |
"step": 1650
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"entropy": 5.696683068275451,
|
| 344 |
"epoch": 0.9786989061600461,
|
| 345 |
+
"grad_norm": 2.8499979972839355,
|
| 346 |
+
"learning_rate": 1.6466705951679433e-06,
|
| 347 |
+
"loss": 5.4335,
|
| 348 |
+
"mean_token_accuracy": 0.29918427973985673,
|
| 349 |
+
"num_tokens": 1817494.0,
|
| 350 |
"step": 1700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 1.0,
|
| 354 |
+
"eval_entropy": 5.993559589034401,
|
| 355 |
+
"eval_loss": 5.737204551696777,
|
| 356 |
+
"eval_mean_token_accuracy": 0.2618687468739699,
|
| 357 |
+
"eval_model_preparation_time": 0.0045,
|
| 358 |
+
"eval_num_tokens": 1856362.0,
|
| 359 |
+
"eval_runtime": 50.5332,
|
| 360 |
+
"eval_samples_per_second": 8.588,
|
| 361 |
+
"eval_steps_per_second": 4.294,
|
| 362 |
"step": 1737
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"entropy": 5.746842083930969,
|
| 366 |
"epoch": 1.0074841681059297,
|
| 367 |
+
"grad_norm": 2.33052921295166,
|
| 368 |
+
"learning_rate": 1.6348850913376547e-06,
|
| 369 |
+
"loss": 5.4796,
|
| 370 |
+
"mean_token_accuracy": 0.2966849410533905,
|
| 371 |
+
"num_tokens": 1870353.0,
|
| 372 |
"step": 1750
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"entropy": 5.859029049873352,
|
| 376 |
"epoch": 1.0362694300518134,
|
| 377 |
+
"grad_norm": 1.6248886585235596,
|
| 378 |
+
"learning_rate": 1.6230995875073658e-06,
|
| 379 |
+
"loss": 5.5975,
|
| 380 |
+
"mean_token_accuracy": 0.2838129925727844,
|
| 381 |
+
"num_tokens": 1926205.0,
|
| 382 |
"step": 1800
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"entropy": 5.731445336341858,
|
| 386 |
"epoch": 1.065054691997697,
|
| 387 |
+
"grad_norm": 1.6941566467285156,
|
| 388 |
+
"learning_rate": 1.6113140836770771e-06,
|
| 389 |
+
"loss": 5.476,
|
| 390 |
+
"mean_token_accuracy": 0.2992346465587616,
|
| 391 |
+
"num_tokens": 1979821.0,
|
| 392 |
"step": 1850
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"entropy": 5.6993954515457155,
|
| 396 |
"epoch": 1.0938399539435808,
|
| 397 |
+
"grad_norm": 1.1746597290039062,
|
| 398 |
+
"learning_rate": 1.5995285798467883e-06,
|
| 399 |
+
"loss": 5.4608,
|
| 400 |
+
"mean_token_accuracy": 0.3000726142525673,
|
| 401 |
+
"num_tokens": 2034373.0,
|
| 402 |
"step": 1900
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"entropy": 5.668873124122619,
|
| 406 |
"epoch": 1.1226252158894645,
|
| 407 |
+
"grad_norm": 1.728211760520935,
|
| 408 |
+
"learning_rate": 1.5877430760164996e-06,
|
| 409 |
+
"loss": 5.4347,
|
| 410 |
+
"mean_token_accuracy": 0.3033922725915909,
|
| 411 |
+
"num_tokens": 2087339.0,
|
| 412 |
"step": 1950
|
| 413 |
},
|
| 414 |
{
|
| 415 |
+
"entropy": 5.624621086120605,
|
| 416 |
"epoch": 1.1514104778353482,
|
| 417 |
+
"grad_norm": 1.4078539609909058,
|
| 418 |
+
"learning_rate": 1.5759575721862107e-06,
|
| 419 |
+
"loss": 5.3954,
|
| 420 |
+
"mean_token_accuracy": 0.30784171640872954,
|
| 421 |
+
"num_tokens": 2139520.0,
|
| 422 |
"step": 2000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"entropy": 5.7141213130950925,
|
| 426 |
"epoch": 1.180195739781232,
|
| 427 |
+
"grad_norm": 2.186459541320801,
|
| 428 |
+
"learning_rate": 1.564172068355922e-06,
|
| 429 |
+
"loss": 5.4847,
|
| 430 |
+
"mean_token_accuracy": 0.29594049394130706,
|
| 431 |
+
"num_tokens": 2193987.0,
|
| 432 |
"step": 2050
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"entropy": 5.632415266036987,
|
| 436 |
"epoch": 1.2089810017271156,
|
| 437 |
+
"grad_norm": 1.3601349592208862,
|
| 438 |
+
"learning_rate": 1.5523865645256334e-06,
|
| 439 |
+
"loss": 5.4135,
|
| 440 |
+
"mean_token_accuracy": 0.30366597563028336,
|
| 441 |
+
"num_tokens": 2249616.0,
|
| 442 |
"step": 2100
|
| 443 |
},
|
| 444 |
{
|
| 445 |
+
"entropy": 5.510904269218445,
|
| 446 |
"epoch": 1.2377662636729994,
|
| 447 |
+
"grad_norm": 2.065760612487793,
|
| 448 |
+
"learning_rate": 1.5406010606953445e-06,
|
| 449 |
+
"loss": 5.2904,
|
| 450 |
+
"mean_token_accuracy": 0.3211754837632179,
|
| 451 |
+
"num_tokens": 2300863.0,
|
| 452 |
"step": 2150
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"entropy": 5.703383626937867,
|
| 456 |
"epoch": 1.266551525618883,
|
| 457 |
+
"grad_norm": 1.1172698736190796,
|
| 458 |
+
"learning_rate": 1.5288155568650559e-06,
|
| 459 |
+
"loss": 5.4802,
|
| 460 |
+
"mean_token_accuracy": 0.29713701367378237,
|
| 461 |
+
"num_tokens": 2356029.0,
|
| 462 |
"step": 2200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
+
"entropy": 5.565930342674255,
|
| 466 |
"epoch": 1.2953367875647668,
|
| 467 |
+
"grad_norm": 1.7528513669967651,
|
| 468 |
+
"learning_rate": 1.5170300530347672e-06,
|
| 469 |
+
"loss": 5.3518,
|
| 470 |
+
"mean_token_accuracy": 0.31301232606172563,
|
| 471 |
+
"num_tokens": 2408957.0,
|
| 472 |
"step": 2250
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"entropy": 5.496430187225342,
|
| 476 |
"epoch": 1.3241220495106505,
|
| 477 |
+
"grad_norm": 1.892640233039856,
|
| 478 |
+
"learning_rate": 1.5052445492044786e-06,
|
| 479 |
+
"loss": 5.2967,
|
| 480 |
+
"mean_token_accuracy": 0.3181899458169937,
|
| 481 |
+
"num_tokens": 2462569.0,
|
| 482 |
"step": 2300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"entropy": 5.725150098800659,
|
| 486 |
"epoch": 1.3529073114565342,
|
| 487 |
+
"grad_norm": 1.774940848350525,
|
| 488 |
+
"learning_rate": 1.4934590453741897e-06,
|
| 489 |
+
"loss": 5.5215,
|
| 490 |
+
"mean_token_accuracy": 0.29055028676986694,
|
| 491 |
+
"num_tokens": 2518544.0,
|
| 492 |
"step": 2350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"entropy": 5.4884827613830565,
|
| 496 |
"epoch": 1.381692573402418,
|
| 497 |
+
"grad_norm": 2.2167599201202393,
|
| 498 |
+
"learning_rate": 1.481673541543901e-06,
|
| 499 |
+
"loss": 5.2917,
|
| 500 |
+
"mean_token_accuracy": 0.31803421139717103,
|
| 501 |
+
"num_tokens": 2570863.0,
|
| 502 |
"step": 2400
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"entropy": 5.697079472541809,
|
| 506 |
"epoch": 1.4104778353483016,
|
| 507 |
+
"grad_norm": 1.6489030122756958,
|
| 508 |
+
"learning_rate": 1.4698880377136124e-06,
|
| 509 |
+
"loss": 5.4982,
|
| 510 |
+
"mean_token_accuracy": 0.2925163987278938,
|
| 511 |
+
"num_tokens": 2626998.0,
|
| 512 |
"step": 2450
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"entropy": 5.46209939956665,
|
| 516 |
"epoch": 1.4392630972941853,
|
| 517 |
+
"grad_norm": 1.153914451599121,
|
| 518 |
+
"learning_rate": 1.4581025338833235e-06,
|
| 519 |
+
"loss": 5.2736,
|
| 520 |
+
"mean_token_accuracy": 0.3182168474793434,
|
| 521 |
+
"num_tokens": 2681568.0,
|
| 522 |
"step": 2500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"entropy": 5.4405768728256225,
|
| 526 |
"epoch": 1.468048359240069,
|
| 527 |
+
"grad_norm": 3.6614978313446045,
|
| 528 |
+
"learning_rate": 1.4463170300530348e-06,
|
| 529 |
+
"loss": 5.2515,
|
| 530 |
+
"mean_token_accuracy": 0.3218736210465431,
|
| 531 |
+
"num_tokens": 2733587.0,
|
| 532 |
"step": 2550
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"entropy": 5.528175053596496,
|
| 536 |
"epoch": 1.4968336211859528,
|
| 537 |
+
"grad_norm": 1.0849746465682983,
|
| 538 |
+
"learning_rate": 1.434531526222746e-06,
|
| 539 |
+
"loss": 5.3378,
|
| 540 |
+
"mean_token_accuracy": 0.31061659604310987,
|
| 541 |
+
"num_tokens": 2787003.0,
|
| 542 |
"step": 2600
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"entropy": 5.46110897064209,
|
| 546 |
"epoch": 1.5256188831318365,
|
| 547 |
+
"grad_norm": 1.8315683603286743,
|
| 548 |
+
"learning_rate": 1.4227460223924573e-06,
|
| 549 |
+
"loss": 5.2782,
|
| 550 |
+
"mean_token_accuracy": 0.31781029611825945,
|
| 551 |
+
"num_tokens": 2840263.0,
|
| 552 |
"step": 2650
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"entropy": 5.455560960769653,
|
| 556 |
"epoch": 1.5544041450777202,
|
| 557 |
+
"grad_norm": 1.1859091520309448,
|
| 558 |
+
"learning_rate": 1.4109605185621684e-06,
|
| 559 |
+
"loss": 5.2735,
|
| 560 |
+
"mean_token_accuracy": 0.3194814011454582,
|
| 561 |
+
"num_tokens": 2894186.0,
|
| 562 |
"step": 2700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"entropy": 5.430496115684509,
|
| 566 |
"epoch": 1.583189407023604,
|
| 567 |
+
"grad_norm": 2.3500001430511475,
|
| 568 |
+
"learning_rate": 1.3991750147318797e-06,
|
| 569 |
+
"loss": 5.2464,
|
| 570 |
+
"mean_token_accuracy": 0.32140792965888976,
|
| 571 |
+
"num_tokens": 2948171.0,
|
| 572 |
"step": 2750
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"entropy": 5.588023023605347,
|
| 576 |
"epoch": 1.6119746689694876,
|
| 577 |
+
"grad_norm": 1.727825403213501,
|
| 578 |
+
"learning_rate": 1.3873895109015909e-06,
|
| 579 |
+
"loss": 5.4028,
|
| 580 |
+
"mean_token_accuracy": 0.3039530631899834,
|
| 581 |
+
"num_tokens": 3002678.0,
|
| 582 |
"step": 2800
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"entropy": 5.410525422096253,
|
| 586 |
"epoch": 1.6407599309153713,
|
| 587 |
+
"grad_norm": 1.3401474952697754,
|
| 588 |
+
"learning_rate": 1.3756040070713022e-06,
|
| 589 |
+
"loss": 5.2298,
|
| 590 |
+
"mean_token_accuracy": 0.324065263569355,
|
| 591 |
+
"num_tokens": 3055844.0,
|
| 592 |
"step": 2850
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"entropy": 5.36959942817688,
|
| 596 |
"epoch": 1.669545192861255,
|
| 597 |
+
"grad_norm": 1.1892589330673218,
|
| 598 |
+
"learning_rate": 1.3638185032410133e-06,
|
| 599 |
+
"loss": 5.1956,
|
| 600 |
+
"mean_token_accuracy": 0.32639502108097074,
|
| 601 |
+
"num_tokens": 3108636.0,
|
| 602 |
"step": 2900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"entropy": 5.53826907157898,
|
| 606 |
"epoch": 1.6983304548071387,
|
| 607 |
+
"grad_norm": 1.2652360200881958,
|
| 608 |
+
"learning_rate": 1.3520329994107247e-06,
|
| 609 |
+
"loss": 5.3583,
|
| 610 |
+
"mean_token_accuracy": 0.3074926760792732,
|
| 611 |
+
"num_tokens": 3162627.0,
|
| 612 |
"step": 2950
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"entropy": 5.417449145317078,
|
| 616 |
"epoch": 1.7271157167530224,
|
| 617 |
+
"grad_norm": 1.584312915802002,
|
| 618 |
+
"learning_rate": 1.340247495580436e-06,
|
| 619 |
+
"loss": 5.2388,
|
| 620 |
+
"mean_token_accuracy": 0.32019727885723115,
|
| 621 |
+
"num_tokens": 3216409.0,
|
| 622 |
"step": 3000
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"entropy": 5.241390740871429,
|
| 626 |
"epoch": 1.7559009786989062,
|
| 627 |
+
"grad_norm": 1.5219439268112183,
|
| 628 |
+
"learning_rate": 1.3284619917501471e-06,
|
| 629 |
+
"loss": 5.0645,
|
| 630 |
+
"mean_token_accuracy": 0.3445430138707161,
|
| 631 |
+
"num_tokens": 3266967.0,
|
| 632 |
"step": 3050
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"entropy": 5.405424036979675,
|
| 636 |
"epoch": 1.7846862406447899,
|
| 637 |
+
"grad_norm": 2.1165153980255127,
|
| 638 |
+
"learning_rate": 1.3166764879198585e-06,
|
| 639 |
+
"loss": 5.232,
|
| 640 |
+
"mean_token_accuracy": 0.32085000157356264,
|
| 641 |
+
"num_tokens": 3319877.0,
|
| 642 |
"step": 3100
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"entropy": 5.123006024360657,
|
| 646 |
"epoch": 1.8134715025906736,
|
| 647 |
+
"grad_norm": 1.2189785242080688,
|
| 648 |
+
"learning_rate": 1.3048909840895698e-06,
|
| 649 |
+
"loss": 4.9582,
|
| 650 |
+
"mean_token_accuracy": 0.356108532845974,
|
| 651 |
+
"num_tokens": 3368569.0,
|
| 652 |
"step": 3150
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"entropy": 5.417610831260681,
|
| 656 |
"epoch": 1.8422567645365573,
|
| 657 |
+
"grad_norm": 1.5157604217529297,
|
| 658 |
+
"learning_rate": 1.2931054802592812e-06,
|
| 659 |
+
"loss": 5.2454,
|
| 660 |
+
"mean_token_accuracy": 0.31976755023002623,
|
| 661 |
+
"num_tokens": 3422449.0,
|
| 662 |
"step": 3200
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"entropy": 5.409690895080566,
|
| 666 |
"epoch": 1.871042026482441,
|
| 667 |
+
"grad_norm": 1.3088161945343018,
|
| 668 |
+
"learning_rate": 1.2813199764289923e-06,
|
| 669 |
+
"loss": 5.2348,
|
| 670 |
+
"mean_token_accuracy": 0.32325415283441544,
|
| 671 |
+
"num_tokens": 3474399.0,
|
| 672 |
"step": 3250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"entropy": 5.44662567615509,
|
| 676 |
"epoch": 1.8998272884283247,
|
| 677 |
+
"grad_norm": 2.178372621536255,
|
| 678 |
+
"learning_rate": 1.2695344725987036e-06,
|
| 679 |
+
"loss": 5.2661,
|
| 680 |
+
"mean_token_accuracy": 0.3182847076654434,
|
| 681 |
+
"num_tokens": 3527726.0,
|
| 682 |
"step": 3300
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"entropy": 5.512614865303039,
|
| 686 |
"epoch": 1.9286125503742084,
|
| 687 |
+
"grad_norm": 1.3050425052642822,
|
| 688 |
+
"learning_rate": 1.2577489687684147e-06,
|
| 689 |
+
"loss": 5.3416,
|
| 690 |
+
"mean_token_accuracy": 0.3084403133392334,
|
| 691 |
+
"num_tokens": 3581980.0,
|
| 692 |
"step": 3350
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"entropy": 5.379772834777832,
|
| 696 |
"epoch": 1.9573978123200921,
|
| 697 |
+
"grad_norm": 1.4584404230117798,
|
| 698 |
+
"learning_rate": 1.245963464938126e-06,
|
| 699 |
+
"loss": 5.2087,
|
| 700 |
+
"mean_token_accuracy": 0.32388432770967485,
|
| 701 |
+
"num_tokens": 3635393.0,
|
| 702 |
"step": 3400
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"entropy": 5.483665924072266,
|
| 706 |
"epoch": 1.9861830742659758,
|
| 707 |
+
"grad_norm": 1.2157734632492065,
|
| 708 |
+
"learning_rate": 1.2341779611078374e-06,
|
| 709 |
+
"loss": 5.3101,
|
| 710 |
+
"mean_token_accuracy": 0.3121953472495079,
|
| 711 |
+
"num_tokens": 3689894.0,
|
| 712 |
"step": 3450
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"epoch": 2.0,
|
| 716 |
+
"eval_entropy": 5.711394641805904,
|
| 717 |
+
"eval_loss": 5.55628776550293,
|
| 718 |
+
"eval_mean_token_accuracy": 0.2764948787777105,
|
| 719 |
+
"eval_model_preparation_time": 0.0045,
|
| 720 |
+
"eval_num_tokens": 3712724.0,
|
| 721 |
+
"eval_runtime": 50.187,
|
| 722 |
+
"eval_samples_per_second": 8.648,
|
| 723 |
+
"eval_steps_per_second": 4.324,
|
| 724 |
"step": 3474
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"entropy": 5.349283556938172,
|
| 728 |
"epoch": 2.0149683362118593,
|
| 729 |
+
"grad_norm": 1.1696771383285522,
|
| 730 |
+
"learning_rate": 1.2223924572775486e-06,
|
| 731 |
+
"loss": 5.1782,
|
| 732 |
+
"mean_token_accuracy": 0.33028870791196824,
|
| 733 |
+
"num_tokens": 3740861.0,
|
| 734 |
"step": 3500
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"entropy": 5.4721107006073,
|
| 738 |
"epoch": 2.043753598157743,
|
| 739 |
+
"grad_norm": 1.8449370861053467,
|
| 740 |
+
"learning_rate": 1.2106069534472599e-06,
|
| 741 |
+
"loss": 5.2978,
|
| 742 |
+
"mean_token_accuracy": 0.31511022299528124,
|
| 743 |
+
"num_tokens": 3794869.0,
|
| 744 |
"step": 3550
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"entropy": 5.404226851463318,
|
| 748 |
"epoch": 2.0725388601036268,
|
| 749 |
+
"grad_norm": 3.789496660232544,
|
| 750 |
+
"learning_rate": 1.198821449616971e-06,
|
| 751 |
+
"loss": 5.2371,
|
| 752 |
+
"mean_token_accuracy": 0.32092176616191864,
|
| 753 |
+
"num_tokens": 3848573.0,
|
| 754 |
"step": 3600
|
| 755 |
},
|
| 756 |
{
|
| 757 |
+
"entropy": 5.435445628166199,
|
| 758 |
"epoch": 2.1013241220495105,
|
| 759 |
+
"grad_norm": 2.2847959995269775,
|
| 760 |
+
"learning_rate": 1.1870359457866824e-06,
|
| 761 |
+
"loss": 5.2662,
|
| 762 |
+
"mean_token_accuracy": 0.3186633634567261,
|
| 763 |
+
"num_tokens": 3901204.0,
|
| 764 |
"step": 3650
|
| 765 |
},
|
| 766 |
{
|
| 767 |
+
"entropy": 5.4066293334960935,
|
| 768 |
"epoch": 2.130109383995394,
|
| 769 |
+
"grad_norm": 1.0950902700424194,
|
| 770 |
+
"learning_rate": 1.1752504419563935e-06,
|
| 771 |
+
"loss": 5.2345,
|
| 772 |
+
"mean_token_accuracy": 0.32156052827835085,
|
| 773 |
+
"num_tokens": 3953753.0,
|
| 774 |
"step": 3700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
+
"entropy": 5.272332944869995,
|
| 778 |
"epoch": 2.158894645941278,
|
| 779 |
+
"grad_norm": 2.1477339267730713,
|
| 780 |
+
"learning_rate": 1.1634649381261048e-06,
|
| 781 |
+
"loss": 5.1091,
|
| 782 |
+
"mean_token_accuracy": 0.3380983591079712,
|
| 783 |
+
"num_tokens": 4005481.0,
|
| 784 |
"step": 3750
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"entropy": 5.4118804311752315,
|
| 788 |
"epoch": 2.1876799078871616,
|
| 789 |
+
"grad_norm": 1.4509484767913818,
|
| 790 |
+
"learning_rate": 1.151679434295816e-06,
|
| 791 |
+
"loss": 5.2448,
|
| 792 |
+
"mean_token_accuracy": 0.3208243528008461,
|
| 793 |
+
"num_tokens": 4058829.0,
|
| 794 |
"step": 3800
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"entropy": 5.4763900089263915,
|
| 798 |
"epoch": 2.2164651698330453,
|
| 799 |
+
"grad_norm": 1.0856804847717285,
|
| 800 |
+
"learning_rate": 1.1398939304655273e-06,
|
| 801 |
+
"loss": 5.3042,
|
| 802 |
+
"mean_token_accuracy": 0.31338351368904116,
|
| 803 |
+
"num_tokens": 4113326.0,
|
| 804 |
"step": 3850
|
| 805 |
},
|
| 806 |
{
|
| 807 |
+
"entropy": 5.328452725410461,
|
| 808 |
"epoch": 2.245250431778929,
|
| 809 |
+
"grad_norm": 3.2843880653381348,
|
| 810 |
+
"learning_rate": 1.1281084266352386e-06,
|
| 811 |
+
"loss": 5.1624,
|
| 812 |
+
"mean_token_accuracy": 0.3305218696594238,
|
| 813 |
+
"num_tokens": 4165454.0,
|
| 814 |
"step": 3900
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"entropy": 5.383157343864441,
|
| 818 |
"epoch": 2.2740356937248127,
|
| 819 |
+
"grad_norm": 2.207082748413086,
|
| 820 |
+
"learning_rate": 1.1163229228049497e-06,
|
| 821 |
+
"loss": 5.2163,
|
| 822 |
+
"mean_token_accuracy": 0.32331310987472534,
|
| 823 |
+
"num_tokens": 4219250.0,
|
| 824 |
"step": 3950
|
| 825 |
},
|
| 826 |
{
|
| 827 |
+
"entropy": 5.585261764526368,
|
| 828 |
"epoch": 2.3028209556706964,
|
| 829 |
+
"grad_norm": 2.7102835178375244,
|
| 830 |
+
"learning_rate": 1.104537418974661e-06,
|
| 831 |
+
"loss": 5.4137,
|
| 832 |
+
"mean_token_accuracy": 0.29959124475717547,
|
| 833 |
+
"num_tokens": 4274711.0,
|
| 834 |
"step": 4000
|
| 835 |
},
|
| 836 |
{
|
| 837 |
+
"entropy": 5.434073266983032,
|
| 838 |
"epoch": 2.33160621761658,
|
| 839 |
+
"grad_norm": 1.3775779008865356,
|
| 840 |
+
"learning_rate": 1.0927519151443724e-06,
|
| 841 |
+
"loss": 5.2644,
|
| 842 |
+
"mean_token_accuracy": 0.3175011593103409,
|
| 843 |
+
"num_tokens": 4328616.0,
|
| 844 |
"step": 4050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
+
"entropy": 5.462391858100891,
|
| 848 |
"epoch": 2.360391479562464,
|
| 849 |
+
"grad_norm": 1.4101024866104126,
|
| 850 |
+
"learning_rate": 1.0809664113140838e-06,
|
| 851 |
+
"loss": 5.2924,
|
| 852 |
+
"mean_token_accuracy": 0.3137941011786461,
|
| 853 |
+
"num_tokens": 4382416.0,
|
| 854 |
"step": 4100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
+
"entropy": 5.529892563819885,
|
| 858 |
"epoch": 2.3891767415083476,
|
| 859 |
+
"grad_norm": 1.2311837673187256,
|
| 860 |
+
"learning_rate": 1.0691809074837949e-06,
|
| 861 |
+
"loss": 5.364,
|
| 862 |
+
"mean_token_accuracy": 0.3046491605043411,
|
| 863 |
+
"num_tokens": 4437848.0,
|
| 864 |
"step": 4150
|
| 865 |
},
|
| 866 |
{
|
| 867 |
+
"entropy": 5.4370484542846675,
|
| 868 |
"epoch": 2.4179620034542313,
|
| 869 |
+
"grad_norm": 1.0929864645004272,
|
| 870 |
+
"learning_rate": 1.0573954036535062e-06,
|
| 871 |
+
"loss": 5.2734,
|
| 872 |
+
"mean_token_accuracy": 0.3169013774394989,
|
| 873 |
+
"num_tokens": 4491185.0,
|
| 874 |
"step": 4200
|
| 875 |
},
|
| 876 |
{
|
| 877 |
+
"entropy": 5.395377616882325,
|
| 878 |
"epoch": 2.446747265400115,
|
| 879 |
+
"grad_norm": 1.5457273721694946,
|
| 880 |
+
"learning_rate": 1.0456098998232174e-06,
|
| 881 |
+
"loss": 5.2276,
|
| 882 |
+
"mean_token_accuracy": 0.32221508473157884,
|
| 883 |
+
"num_tokens": 4544086.0,
|
| 884 |
"step": 4250
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"entropy": 5.443737335205078,
|
| 888 |
"epoch": 2.4755325273459987,
|
| 889 |
+
"grad_norm": 1.4844346046447754,
|
| 890 |
+
"learning_rate": 1.0338243959929287e-06,
|
| 891 |
+
"loss": 5.2786,
|
| 892 |
+
"mean_token_accuracy": 0.3157751387357712,
|
| 893 |
+
"num_tokens": 4597677.0,
|
| 894 |
"step": 4300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
+
"entropy": 5.419876251220703,
|
| 898 |
"epoch": 2.5043177892918824,
|
| 899 |
+
"grad_norm": 1.2481963634490967,
|
| 900 |
+
"learning_rate": 1.02203889216264e-06,
|
| 901 |
+
"loss": 5.2564,
|
| 902 |
+
"mean_token_accuracy": 0.31889803290367125,
|
| 903 |
+
"num_tokens": 4651343.0,
|
| 904 |
"step": 4350
|
| 905 |
},
|
| 906 |
{
|
| 907 |
+
"entropy": 5.578677978515625,
|
| 908 |
"epoch": 2.533103051237766,
|
| 909 |
+
"grad_norm": 2.0005414485931396,
|
| 910 |
+
"learning_rate": 1.0102533883323512e-06,
|
| 911 |
+
"loss": 5.4145,
|
| 912 |
+
"mean_token_accuracy": 0.30037090003490446,
|
| 913 |
+
"num_tokens": 4705985.0,
|
| 914 |
"step": 4400
|
| 915 |
},
|
| 916 |
{
|
| 917 |
+
"entropy": 5.279946126937866,
|
| 918 |
"epoch": 2.56188831318365,
|
| 919 |
+
"grad_norm": 1.080521821975708,
|
| 920 |
+
"learning_rate": 9.984678845020625e-07,
|
| 921 |
+
"loss": 5.1226,
|
| 922 |
+
"mean_token_accuracy": 0.3341303279995918,
|
| 923 |
+
"num_tokens": 4757741.0,
|
| 924 |
"step": 4450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"entropy": 5.551463279724121,
|
| 928 |
"epoch": 2.5906735751295336,
|
| 929 |
+
"grad_norm": 1.28898024559021,
|
| 930 |
+
"learning_rate": 9.866823806717736e-07,
|
| 931 |
+
"loss": 5.3832,
|
| 932 |
+
"mean_token_accuracy": 0.3028248634934425,
|
| 933 |
+
"num_tokens": 4812808.0,
|
| 934 |
"step": 4500
|
| 935 |
},
|
| 936 |
{
|
| 937 |
+
"entropy": 5.3787487554550175,
|
| 938 |
"epoch": 2.6194588370754173,
|
| 939 |
+
"grad_norm": 1.5697983503341675,
|
| 940 |
+
"learning_rate": 9.74896876841485e-07,
|
| 941 |
+
"loss": 5.2141,
|
| 942 |
+
"mean_token_accuracy": 0.3227942296862602,
|
| 943 |
+
"num_tokens": 4866572.0,
|
| 944 |
"step": 4550
|
| 945 |
},
|
| 946 |
{
|
| 947 |
+
"entropy": 5.460358958244324,
|
| 948 |
"epoch": 2.648244099021301,
|
| 949 |
+
"grad_norm": 1.3180441856384277,
|
| 950 |
+
"learning_rate": 9.63111373011196e-07,
|
| 951 |
+
"loss": 5.2954,
|
| 952 |
+
"mean_token_accuracy": 0.31269474506378175,
|
| 953 |
+
"num_tokens": 4921312.0,
|
| 954 |
"step": 4600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"entropy": 5.434084935188293,
|
| 958 |
"epoch": 2.6770293609671847,
|
| 959 |
+
"grad_norm": 1.2409590482711792,
|
| 960 |
+
"learning_rate": 9.513258691809074e-07,
|
| 961 |
+
"loss": 5.271,
|
| 962 |
+
"mean_token_accuracy": 0.3172155100107193,
|
| 963 |
+
"num_tokens": 4974289.0,
|
| 964 |
"step": 4650
|
| 965 |
},
|
| 966 |
{
|
| 967 |
+
"entropy": 5.406955418586731,
|
| 968 |
"epoch": 2.7058146229130684,
|
| 969 |
+
"grad_norm": 1.4782609939575195,
|
| 970 |
+
"learning_rate": 9.395403653506187e-07,
|
| 971 |
+
"loss": 5.2473,
|
| 972 |
+
"mean_token_accuracy": 0.32031788885593415,
|
| 973 |
+
"num_tokens": 5028149.0,
|
| 974 |
"step": 4700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
+
"entropy": 5.206603040695191,
|
| 978 |
"epoch": 2.734599884858952,
|
| 979 |
+
"grad_norm": 2.351633071899414,
|
| 980 |
+
"learning_rate": 9.2775486152033e-07,
|
| 981 |
+
"loss": 5.0478,
|
| 982 |
+
"mean_token_accuracy": 0.3428420132398605,
|
| 983 |
+
"num_tokens": 5079349.0,
|
| 984 |
"step": 4750
|
| 985 |
},
|
| 986 |
{
|
| 987 |
+
"entropy": 5.388812799453735,
|
| 988 |
"epoch": 2.763385146804836,
|
| 989 |
+
"grad_norm": 7.564618110656738,
|
| 990 |
+
"learning_rate": 9.159693576900412e-07,
|
| 991 |
+
"loss": 5.2281,
|
| 992 |
+
"mean_token_accuracy": 0.3222071170806885,
|
| 993 |
+
"num_tokens": 5132564.0,
|
| 994 |
"step": 4800
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"entropy": 5.374106278419495,
|
| 998 |
"epoch": 2.7921704087507195,
|
| 999 |
+
"grad_norm": 1.4734679460525513,
|
| 1000 |
+
"learning_rate": 9.041838538597525e-07,
|
| 1001 |
+
"loss": 5.2161,
|
| 1002 |
+
"mean_token_accuracy": 0.3219477406144142,
|
| 1003 |
+
"num_tokens": 5185921.0,
|
| 1004 |
"step": 4850
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
+
"entropy": 5.232998585700988,
|
| 1008 |
"epoch": 2.8209556706966032,
|
| 1009 |
+
"grad_norm": 1.4175471067428589,
|
| 1010 |
+
"learning_rate": 8.923983500294637e-07,
|
| 1011 |
+
"loss": 5.0769,
|
| 1012 |
+
"mean_token_accuracy": 0.3403926733136177,
|
| 1013 |
+
"num_tokens": 5237521.0,
|
| 1014 |
"step": 4900
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
+
"entropy": 5.394891719818116,
|
| 1018 |
"epoch": 2.849740932642487,
|
| 1019 |
+
"grad_norm": 4.951873779296875,
|
| 1020 |
+
"learning_rate": 8.806128461991749e-07,
|
| 1021 |
+
"loss": 5.2344,
|
| 1022 |
+
"mean_token_accuracy": 0.3213117456436157,
|
| 1023 |
+
"num_tokens": 5291104.0,
|
| 1024 |
"step": 4950
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"entropy": 5.413805012702942,
|
| 1028 |
"epoch": 2.8785261945883707,
|
| 1029 |
+
"grad_norm": 1.679518461227417,
|
| 1030 |
+
"learning_rate": 8.688273423688863e-07,
|
| 1031 |
+
"loss": 5.2597,
|
| 1032 |
+
"mean_token_accuracy": 0.3165634173154831,
|
| 1033 |
+
"num_tokens": 5345058.0,
|
| 1034 |
"step": 5000
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
+
"entropy": 5.256177935600281,
|
| 1038 |
"epoch": 2.9073114565342544,
|
| 1039 |
+
"grad_norm": 1.8892916440963745,
|
| 1040 |
+
"learning_rate": 8.570418385385975e-07,
|
| 1041 |
+
"loss": 5.1004,
|
| 1042 |
+
"mean_token_accuracy": 0.3369427987933159,
|
| 1043 |
+
"num_tokens": 5395918.0,
|
| 1044 |
"step": 5050
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
+
"entropy": 5.259814453125,
|
| 1048 |
"epoch": 2.936096718480138,
|
| 1049 |
+
"grad_norm": 1.3802675008773804,
|
| 1050 |
+
"learning_rate": 8.452563347083087e-07,
|
| 1051 |
+
"loss": 5.1057,
|
| 1052 |
+
"mean_token_accuracy": 0.3362414276599884,
|
| 1053 |
+
"num_tokens": 5448086.0,
|
| 1054 |
"step": 5100
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
+
"entropy": 5.416206178665161,
|
| 1058 |
"epoch": 2.964881980426022,
|
| 1059 |
+
"grad_norm": 1.7677236795425415,
|
| 1060 |
+
"learning_rate": 8.3347083087802e-07,
|
| 1061 |
+
"loss": 5.2562,
|
| 1062 |
+
"mean_token_accuracy": 0.31725785195827483,
|
| 1063 |
+
"num_tokens": 5501959.0,
|
| 1064 |
"step": 5150
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
+
"entropy": 5.507337794303894,
|
| 1068 |
"epoch": 2.9936672423719055,
|
| 1069 |
+
"grad_norm": 1.021727442741394,
|
| 1070 |
+
"learning_rate": 8.216853270477313e-07,
|
| 1071 |
+
"loss": 5.344,
|
| 1072 |
+
"mean_token_accuracy": 0.30679062128067014,
|
| 1073 |
+
"num_tokens": 5557908.0,
|
| 1074 |
"step": 5200
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 3.0,
|
| 1078 |
+
"eval_entropy": 5.682707933786278,
|
| 1079 |
+
"eval_loss": 5.53223991394043,
|
| 1080 |
+
"eval_mean_token_accuracy": 0.27747743456594404,
|
| 1081 |
+
"eval_model_preparation_time": 0.0045,
|
| 1082 |
+
"eval_num_tokens": 5569086.0,
|
| 1083 |
+
"eval_runtime": 49.9944,
|
| 1084 |
+
"eval_samples_per_second": 8.681,
|
| 1085 |
+
"eval_steps_per_second": 4.34,
|
| 1086 |
"step": 5211
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
+
"entropy": 5.209756035804748,
|
| 1090 |
"epoch": 3.0224525043177892,
|
| 1091 |
+
"grad_norm": 1.725786566734314,
|
| 1092 |
+
"learning_rate": 8.098998232174425e-07,
|
| 1093 |
+
"loss": 5.0541,
|
| 1094 |
+
"mean_token_accuracy": 0.34166110813617706,
|
| 1095 |
+
"num_tokens": 5608917.0,
|
| 1096 |
"step": 5250
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
+
"entropy": 5.396296281814575,
|
| 1100 |
"epoch": 3.051237766263673,
|
| 1101 |
+
"grad_norm": 0.7720207571983337,
|
| 1102 |
+
"learning_rate": 7.981143193871538e-07,
|
| 1103 |
+
"loss": 5.2337,
|
| 1104 |
+
"mean_token_accuracy": 0.32116260558366777,
|
| 1105 |
+
"num_tokens": 5662712.0,
|
| 1106 |
"step": 5300
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
+
"entropy": 5.341518473625183,
|
| 1110 |
"epoch": 3.0800230282095566,
|
| 1111 |
+
"grad_norm": 2.2686808109283447,
|
| 1112 |
+
"learning_rate": 7.86328815556865e-07,
|
| 1113 |
+
"loss": 5.1824,
|
| 1114 |
+
"mean_token_accuracy": 0.32726580530405047,
|
| 1115 |
+
"num_tokens": 5715921.0,
|
| 1116 |
"step": 5350
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
+
"entropy": 5.376176896095276,
|
| 1120 |
"epoch": 3.1088082901554404,
|
| 1121 |
+
"grad_norm": 1.2420796155929565,
|
| 1122 |
+
"learning_rate": 7.745433117265762e-07,
|
| 1123 |
+
"loss": 5.2162,
|
| 1124 |
+
"mean_token_accuracy": 0.32142678707838057,
|
| 1125 |
+
"num_tokens": 5769436.0,
|
| 1126 |
"step": 5400
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
+
"entropy": 5.4553061914443965,
|
| 1130 |
"epoch": 3.137593552101324,
|
| 1131 |
+
"grad_norm": 1.2402859926223755,
|
| 1132 |
+
"learning_rate": 7.627578078962876e-07,
|
| 1133 |
+
"loss": 5.2971,
|
| 1134 |
+
"mean_token_accuracy": 0.31396267503499986,
|
| 1135 |
+
"num_tokens": 5823126.0,
|
| 1136 |
"step": 5450
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"entropy": 5.385247969627381,
|
| 1140 |
"epoch": 3.166378814047208,
|
| 1141 |
+
"grad_norm": 1.112062931060791,
|
| 1142 |
+
"learning_rate": 7.509723040659988e-07,
|
| 1143 |
+
"loss": 5.2324,
|
| 1144 |
+
"mean_token_accuracy": 0.3207343602180481,
|
| 1145 |
+
"num_tokens": 5875751.0,
|
| 1146 |
"step": 5500
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
+
"entropy": 5.55422221660614,
|
| 1150 |
"epoch": 3.1951640759930915,
|
| 1151 |
+
"grad_norm": 1.5440446138381958,
|
| 1152 |
+
"learning_rate": 7.3918680023571e-07,
|
| 1153 |
+
"loss": 5.3902,
|
| 1154 |
+
"mean_token_accuracy": 0.3006985321640968,
|
| 1155 |
+
"num_tokens": 5932163.0,
|
| 1156 |
"step": 5550
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
+
"entropy": 5.403217372894287,
|
| 1160 |
"epoch": 3.223949337938975,
|
| 1161 |
+
"grad_norm": 0.8481096625328064,
|
| 1162 |
+
"learning_rate": 7.274012964054213e-07,
|
| 1163 |
+
"loss": 5.2417,
|
| 1164 |
+
"mean_token_accuracy": 0.3210747820138931,
|
| 1165 |
+
"num_tokens": 5985889.0,
|
| 1166 |
"step": 5600
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
+
"entropy": 5.388293180465698,
|
| 1170 |
"epoch": 3.252734599884859,
|
| 1171 |
+
"grad_norm": 0.9305989146232605,
|
| 1172 |
+
"learning_rate": 7.156157925751326e-07,
|
| 1173 |
+
"loss": 5.2319,
|
| 1174 |
+
"mean_token_accuracy": 0.3206030324101448,
|
| 1175 |
+
"num_tokens": 6040052.0,
|
| 1176 |
"step": 5650
|
| 1177 |
},
|
| 1178 |
{
|
| 1179 |
+
"entropy": 5.401709322929382,
|
| 1180 |
"epoch": 3.2815198618307426,
|
| 1181 |
+
"grad_norm": 0.8080459237098694,
|
| 1182 |
+
"learning_rate": 7.038302887448438e-07,
|
| 1183 |
+
"loss": 5.2438,
|
| 1184 |
+
"mean_token_accuracy": 0.3199671137332916,
|
| 1185 |
+
"num_tokens": 6092350.0,
|
| 1186 |
"step": 5700
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
+
"entropy": 5.4320423412323,
|
| 1190 |
"epoch": 3.3103051237766263,
|
| 1191 |
+
"grad_norm": 1.9186089038848877,
|
| 1192 |
+
"learning_rate": 6.920447849145551e-07,
|
| 1193 |
+
"loss": 5.2696,
|
| 1194 |
+
"mean_token_accuracy": 0.31657984614372253,
|
| 1195 |
+
"num_tokens": 6146112.0,
|
| 1196 |
"step": 5750
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
+
"entropy": 5.276471285820008,
|
| 1200 |
"epoch": 3.33909038572251,
|
| 1201 |
+
"grad_norm": 1.032879114151001,
|
| 1202 |
+
"learning_rate": 6.802592810842663e-07,
|
| 1203 |
+
"loss": 5.1224,
|
| 1204 |
+
"mean_token_accuracy": 0.3347566506266594,
|
| 1205 |
+
"num_tokens": 6197916.0,
|
| 1206 |
"step": 5800
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
+
"entropy": 5.122317051887512,
|
| 1210 |
"epoch": 3.3678756476683938,
|
| 1211 |
+
"grad_norm": 3.156858444213867,
|
| 1212 |
+
"learning_rate": 6.684737772539775e-07,
|
| 1213 |
+
"loss": 4.9706,
|
| 1214 |
+
"mean_token_accuracy": 0.35455317378044127,
|
| 1215 |
+
"num_tokens": 6247565.0,
|
| 1216 |
"step": 5850
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
+
"entropy": 5.346597375869751,
|
| 1220 |
"epoch": 3.3966609096142775,
|
| 1221 |
+
"grad_norm": 1.2619549036026,
|
| 1222 |
+
"learning_rate": 6.566882734236889e-07,
|
| 1223 |
+
"loss": 5.1902,
|
| 1224 |
+
"mean_token_accuracy": 0.3258721518516541,
|
| 1225 |
+
"num_tokens": 6300481.0,
|
| 1226 |
"step": 5900
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
+
"entropy": 5.413151068687439,
|
| 1230 |
"epoch": 3.425446171560161,
|
| 1231 |
+
"grad_norm": 1.801740050315857,
|
| 1232 |
+
"learning_rate": 6.449027695934001e-07,
|
| 1233 |
+
"loss": 5.2513,
|
| 1234 |
+
"mean_token_accuracy": 0.3187857499718666,
|
| 1235 |
+
"num_tokens": 6353098.0,
|
| 1236 |
"step": 5950
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
+
"entropy": 5.464186942577362,
|
| 1240 |
"epoch": 3.454231433506045,
|
| 1241 |
+
"grad_norm": 1.6306997537612915,
|
| 1242 |
+
"learning_rate": 6.331172657631113e-07,
|
| 1243 |
+
"loss": 5.3043,
|
| 1244 |
+
"mean_token_accuracy": 0.31154109388589857,
|
| 1245 |
+
"num_tokens": 6407984.0,
|
| 1246 |
"step": 6000
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
+
"entropy": 5.401795778274536,
|
| 1250 |
"epoch": 3.4830166954519286,
|
| 1251 |
+
"grad_norm": 1.1694583892822266,
|
| 1252 |
+
"learning_rate": 6.213317619328226e-07,
|
| 1253 |
+
"loss": 5.2427,
|
| 1254 |
+
"mean_token_accuracy": 0.31954523265361784,
|
| 1255 |
+
"num_tokens": 6461854.0,
|
| 1256 |
"step": 6050
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
+
"entropy": 5.317689285278321,
|
| 1260 |
"epoch": 3.5118019573978123,
|
| 1261 |
+
"grad_norm": 0.9361855387687683,
|
| 1262 |
+
"learning_rate": 6.095462581025339e-07,
|
| 1263 |
+
"loss": 5.1588,
|
| 1264 |
+
"mean_token_accuracy": 0.330586878657341,
|
| 1265 |
+
"num_tokens": 6514882.0,
|
| 1266 |
"step": 6100
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
+
"entropy": 5.478708257675171,
|
| 1270 |
"epoch": 3.540587219343696,
|
| 1271 |
+
"grad_norm": 1.05711030960083,
|
| 1272 |
+
"learning_rate": 5.977607542722451e-07,
|
| 1273 |
+
"loss": 5.321,
|
| 1274 |
+
"mean_token_accuracy": 0.3104448106884956,
|
| 1275 |
+
"num_tokens": 6569455.0,
|
| 1276 |
"step": 6150
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
+
"entropy": 5.309361801147461,
|
| 1280 |
"epoch": 3.5693724812895797,
|
| 1281 |
+
"grad_norm": 1.3499550819396973,
|
| 1282 |
+
"learning_rate": 5.859752504419564e-07,
|
| 1283 |
+
"loss": 5.153,
|
| 1284 |
+
"mean_token_accuracy": 0.331512533724308,
|
| 1285 |
+
"num_tokens": 6621734.0,
|
| 1286 |
"step": 6200
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
+
"entropy": 5.296572666168213,
|
| 1290 |
"epoch": 3.5981577432354634,
|
| 1291 |
+
"grad_norm": 1.940708875656128,
|
| 1292 |
+
"learning_rate": 5.741897466116676e-07,
|
| 1293 |
+
"loss": 5.14,
|
| 1294 |
+
"mean_token_accuracy": 0.3299832499027252,
|
| 1295 |
+
"num_tokens": 6674994.0,
|
| 1296 |
"step": 6250
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
+
"entropy": 5.544284400939941,
|
| 1300 |
"epoch": 3.626943005181347,
|
| 1301 |
+
"grad_norm": 1.8903827667236328,
|
| 1302 |
+
"learning_rate": 5.624042427813788e-07,
|
| 1303 |
+
"loss": 5.3885,
|
| 1304 |
+
"mean_token_accuracy": 0.3016947290301323,
|
| 1305 |
+
"num_tokens": 6730674.0,
|
| 1306 |
"step": 6300
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
+
"entropy": 5.333053431510925,
|
| 1310 |
"epoch": 3.655728267127231,
|
| 1311 |
+
"grad_norm": 1.1618578433990479,
|
| 1312 |
+
"learning_rate": 5.506187389510902e-07,
|
| 1313 |
+
"loss": 5.1781,
|
| 1314 |
+
"mean_token_accuracy": 0.3275001719594002,
|
| 1315 |
+
"num_tokens": 6784235.0,
|
| 1316 |
"step": 6350
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
+
"entropy": 5.4938449716568,
|
| 1320 |
"epoch": 3.6845135290731146,
|
| 1321 |
+
"grad_norm": 1.384329080581665,
|
| 1322 |
+
"learning_rate": 5.388332351208014e-07,
|
| 1323 |
+
"loss": 5.3399,
|
| 1324 |
+
"mean_token_accuracy": 0.3068840709328651,
|
| 1325 |
+
"num_tokens": 6839590.0,
|
| 1326 |
"step": 6400
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
+
"entropy": 5.277545223236084,
|
| 1330 |
"epoch": 3.7132987910189983,
|
| 1331 |
+
"grad_norm": 1.8918265104293823,
|
| 1332 |
+
"learning_rate": 5.270477312905126e-07,
|
| 1333 |
+
"loss": 5.1221,
|
| 1334 |
+
"mean_token_accuracy": 0.33364981949329375,
|
| 1335 |
+
"num_tokens": 6891301.0,
|
| 1336 |
"step": 6450
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
+
"entropy": 5.40100293636322,
|
| 1340 |
"epoch": 3.742084052964882,
|
| 1341 |
+
"grad_norm": 1.6968809366226196,
|
| 1342 |
+
"learning_rate": 5.152622274602239e-07,
|
| 1343 |
+
"loss": 5.2471,
|
| 1344 |
+
"mean_token_accuracy": 0.31912936180830004,
|
| 1345 |
+
"num_tokens": 6945510.0,
|
| 1346 |
"step": 6500
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
+
"entropy": 5.561220169067383,
|
| 1350 |
"epoch": 3.7708693149107657,
|
| 1351 |
+
"grad_norm": 2.066960573196411,
|
| 1352 |
+
"learning_rate": 5.034767236299352e-07,
|
| 1353 |
+
"loss": 5.4026,
|
| 1354 |
+
"mean_token_accuracy": 0.2984810543060303,
|
| 1355 |
+
"num_tokens": 7001870.0,
|
| 1356 |
"step": 6550
|
| 1357 |
},
|
| 1358 |
{
|
| 1359 |
+
"entropy": 5.3108087682724,
|
| 1360 |
"epoch": 3.7996545768566494,
|
| 1361 |
+
"grad_norm": 1.6065007448196411,
|
| 1362 |
+
"learning_rate": 4.916912197996464e-07,
|
| 1363 |
+
"loss": 5.155,
|
| 1364 |
+
"mean_token_accuracy": 0.3304683968424797,
|
| 1365 |
+
"num_tokens": 7053974.0,
|
| 1366 |
"step": 6600
|
| 1367 |
},
|
| 1368 |
{
|
| 1369 |
+
"entropy": 5.323807754516602,
|
| 1370 |
"epoch": 3.828439838802533,
|
| 1371 |
+
"grad_norm": 2.6806318759918213,
|
| 1372 |
+
"learning_rate": 4.799057159693577e-07,
|
| 1373 |
+
"loss": 5.1653,
|
| 1374 |
+
"mean_token_accuracy": 0.3294159671664238,
|
| 1375 |
+
"num_tokens": 7107061.0,
|
| 1376 |
"step": 6650
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
+
"entropy": 5.4716163873672485,
|
| 1380 |
"epoch": 3.857225100748417,
|
| 1381 |
+
"grad_norm": 1.8264856338500977,
|
| 1382 |
+
"learning_rate": 4.6812021213906895e-07,
|
| 1383 |
+
"loss": 5.3124,
|
| 1384 |
+
"mean_token_accuracy": 0.3109353107213974,
|
| 1385 |
+
"num_tokens": 7161697.0,
|
| 1386 |
"step": 6700
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
+
"entropy": 5.382365622520447,
|
| 1390 |
"epoch": 3.8860103626943006,
|
| 1391 |
+
"grad_norm": 0.9954923987388611,
|
| 1392 |
+
"learning_rate": 4.563347083087802e-07,
|
| 1393 |
+
"loss": 5.2237,
|
| 1394 |
+
"mean_token_accuracy": 0.32161149621009827,
|
| 1395 |
+
"num_tokens": 7215524.0,
|
| 1396 |
"step": 6750
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
+
"entropy": 5.277496585845947,
|
| 1400 |
"epoch": 3.9147956246401843,
|
| 1401 |
+
"grad_norm": 1.267786979675293,
|
| 1402 |
+
"learning_rate": 4.445492044784914e-07,
|
| 1403 |
+
"loss": 5.1265,
|
| 1404 |
+
"mean_token_accuracy": 0.3319795566797257,
|
| 1405 |
+
"num_tokens": 7267329.0,
|
| 1406 |
"step": 6800
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
+
"entropy": 5.550942025184631,
|
| 1410 |
"epoch": 3.943580886586068,
|
| 1411 |
+
"grad_norm": 0.9425063133239746,
|
| 1412 |
+
"learning_rate": 4.3276370064820265e-07,
|
| 1413 |
+
"loss": 5.3898,
|
| 1414 |
+
"mean_token_accuracy": 0.30050904959440233,
|
| 1415 |
+
"num_tokens": 7324070.0,
|
| 1416 |
"step": 6850
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
+
"entropy": 5.125799627304077,
|
| 1420 |
"epoch": 3.9723661485319517,
|
| 1421 |
+
"grad_norm": 5.447021007537842,
|
| 1422 |
+
"learning_rate": 4.20978196817914e-07,
|
| 1423 |
+
"loss": 4.9781,
|
| 1424 |
+
"mean_token_accuracy": 0.3520450854301453,
|
| 1425 |
+
"num_tokens": 7375083.0,
|
| 1426 |
"step": 6900
|
| 1427 |
},
|
| 1428 |
{
|
| 1429 |
"epoch": 4.0,
|
| 1430 |
+
"eval_entropy": 5.6681923492712905,
|
| 1431 |
+
"eval_loss": 5.525067329406738,
|
| 1432 |
+
"eval_mean_token_accuracy": 0.2779707208893816,
|
| 1433 |
+
"eval_model_preparation_time": 0.0045,
|
| 1434 |
+
"eval_num_tokens": 7425448.0,
|
| 1435 |
+
"eval_runtime": 49.7944,
|
| 1436 |
+
"eval_samples_per_second": 8.716,
|
| 1437 |
+
"eval_steps_per_second": 4.358,
|
| 1438 |
"step": 6948
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
+
"entropy": 5.38487633228302,
|
| 1442 |
"epoch": 4.001151410477835,
|
| 1443 |
+
"grad_norm": 1.2222857475280762,
|
| 1444 |
+
"learning_rate": 4.091926929876252e-07,
|
| 1445 |
+
"loss": 5.2243,
|
| 1446 |
+
"mean_token_accuracy": 0.32152373433113096,
|
| 1447 |
+
"num_tokens": 7427859.0,
|
| 1448 |
"step": 6950
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
+
"entropy": 5.151427125930786,
|
| 1452 |
"epoch": 4.029936672423719,
|
| 1453 |
+
"grad_norm": 1.804520606994629,
|
| 1454 |
+
"learning_rate": 3.9740718915733645e-07,
|
| 1455 |
+
"loss": 4.9966,
|
| 1456 |
+
"mean_token_accuracy": 0.3502719843387604,
|
| 1457 |
+
"num_tokens": 7478346.0,
|
| 1458 |
"step": 7000
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
+
"entropy": 5.550741171836853,
|
| 1462 |
"epoch": 4.058721934369602,
|
| 1463 |
+
"grad_norm": 0.9588176608085632,
|
| 1464 |
+
"learning_rate": 3.856216853270477e-07,
|
| 1465 |
+
"loss": 5.3924,
|
| 1466 |
+
"mean_token_accuracy": 0.30148445934057233,
|
| 1467 |
+
"num_tokens": 7533394.0,
|
| 1468 |
"step": 7050
|
| 1469 |
},
|
| 1470 |
{
|
| 1471 |
+
"entropy": 5.254554944038391,
|
| 1472 |
"epoch": 4.087507196315486,
|
| 1473 |
+
"grad_norm": 2.456322431564331,
|
| 1474 |
+
"learning_rate": 3.7383618149675897e-07,
|
| 1475 |
+
"loss": 5.1002,
|
| 1476 |
+
"mean_token_accuracy": 0.3362416020035744,
|
| 1477 |
+
"num_tokens": 7585238.0,
|
| 1478 |
"step": 7100
|
| 1479 |
},
|
| 1480 |
{
|
| 1481 |
+
"entropy": 5.400679998397827,
|
| 1482 |
"epoch": 4.11629245826137,
|
| 1483 |
+
"grad_norm": 1.2165497541427612,
|
| 1484 |
+
"learning_rate": 3.6205067766647026e-07,
|
| 1485 |
+
"loss": 5.2434,
|
| 1486 |
+
"mean_token_accuracy": 0.3191445592045784,
|
| 1487 |
+
"num_tokens": 7638693.0,
|
| 1488 |
"step": 7150
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
+
"entropy": 5.261249952316284,
|
| 1492 |
"epoch": 4.1450777202072535,
|
| 1493 |
+
"grad_norm": 1.1991236209869385,
|
| 1494 |
+
"learning_rate": 3.502651738361815e-07,
|
| 1495 |
+
"loss": 5.1129,
|
| 1496 |
+
"mean_token_accuracy": 0.33393161594867704,
|
| 1497 |
+
"num_tokens": 7691223.0,
|
| 1498 |
"step": 7200
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
+
"entropy": 5.33568838596344,
|
| 1502 |
"epoch": 4.173862982153137,
|
| 1503 |
+
"grad_norm": 1.2175132036209106,
|
| 1504 |
+
"learning_rate": 3.384796700058927e-07,
|
| 1505 |
+
"loss": 5.1725,
|
| 1506 |
+
"mean_token_accuracy": 0.32868497937917707,
|
| 1507 |
+
"num_tokens": 7744696.0,
|
| 1508 |
"step": 7250
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
+
"entropy": 5.25973867893219,
|
| 1512 |
"epoch": 4.202648244099021,
|
| 1513 |
+
"grad_norm": 1.4588052034378052,
|
| 1514 |
+
"learning_rate": 3.2669416617560395e-07,
|
| 1515 |
+
"loss": 5.105,
|
| 1516 |
+
"mean_token_accuracy": 0.3384602865576744,
|
| 1517 |
+
"num_tokens": 7796600.0,
|
| 1518 |
"step": 7300
|
| 1519 |
},
|
| 1520 |
{
|
| 1521 |
+
"entropy": 5.551795811653137,
|
| 1522 |
"epoch": 4.231433506044905,
|
| 1523 |
+
"grad_norm": 1.017745852470398,
|
| 1524 |
+
"learning_rate": 3.149086623453153e-07,
|
| 1525 |
+
"loss": 5.3996,
|
| 1526 |
+
"mean_token_accuracy": 0.29976913034915925,
|
| 1527 |
+
"num_tokens": 7851822.0,
|
| 1528 |
"step": 7350
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
+
"entropy": 5.379249300956726,
|
| 1532 |
"epoch": 4.260218767990788,
|
| 1533 |
+
"grad_norm": 1.7130656242370605,
|
| 1534 |
+
"learning_rate": 3.031231585150265e-07,
|
| 1535 |
+
"loss": 5.2267,
|
| 1536 |
+
"mean_token_accuracy": 0.3196685525774956,
|
| 1537 |
+
"num_tokens": 7905364.0,
|
| 1538 |
"step": 7400
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
+
"entropy": 5.402966260910034,
|
| 1542 |
"epoch": 4.289004029936672,
|
| 1543 |
+
"grad_norm": 1.3162118196487427,
|
| 1544 |
+
"learning_rate": 2.9133765468473776e-07,
|
| 1545 |
+
"loss": 5.2473,
|
| 1546 |
+
"mean_token_accuracy": 0.31836180537939074,
|
| 1547 |
+
"num_tokens": 7959154.0,
|
| 1548 |
"step": 7450
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
+
"entropy": 5.393479719161987,
|
| 1552 |
"epoch": 4.317789291882556,
|
| 1553 |
+
"grad_norm": 1.5286513566970825,
|
| 1554 |
+
"learning_rate": 2.79552150854449e-07,
|
| 1555 |
+
"loss": 5.2345,
|
| 1556 |
+
"mean_token_accuracy": 0.32134009718894957,
|
| 1557 |
+
"num_tokens": 8013160.0,
|
| 1558 |
"step": 7500
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
+
"entropy": 5.46117250919342,
|
| 1562 |
"epoch": 4.3465745538284395,
|
| 1563 |
+
"grad_norm": 1.3793199062347412,
|
| 1564 |
+
"learning_rate": 2.6776664702416027e-07,
|
| 1565 |
+
"loss": 5.3056,
|
| 1566 |
+
"mean_token_accuracy": 0.311002559363842,
|
| 1567 |
+
"num_tokens": 8068763.0,
|
| 1568 |
"step": 7550
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
+
"entropy": 5.331483993530274,
|
| 1572 |
"epoch": 4.375359815774323,
|
| 1573 |
+
"grad_norm": 2.3478105068206787,
|
| 1574 |
+
"learning_rate": 2.5598114319387156e-07,
|
| 1575 |
+
"loss": 5.1795,
|
| 1576 |
+
"mean_token_accuracy": 0.32778155684471133,
|
| 1577 |
+
"num_tokens": 8121013.0,
|
| 1578 |
"step": 7600
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
+
"entropy": 5.39360936164856,
|
| 1582 |
"epoch": 4.404145077720207,
|
| 1583 |
+
"grad_norm": 1.095144510269165,
|
| 1584 |
+
"learning_rate": 2.441956393635828e-07,
|
| 1585 |
+
"loss": 5.2346,
|
| 1586 |
+
"mean_token_accuracy": 0.3206799927353859,
|
| 1587 |
+
"num_tokens": 8174476.0,
|
| 1588 |
"step": 7650
|
| 1589 |
},
|
| 1590 |
{
|
| 1591 |
+
"entropy": 5.468272385597229,
|
| 1592 |
"epoch": 4.432930339666091,
|
| 1593 |
+
"grad_norm": 1.161205768585205,
|
| 1594 |
+
"learning_rate": 2.3241013553329402e-07,
|
| 1595 |
+
"loss": 5.3126,
|
| 1596 |
+
"mean_token_accuracy": 0.3104386702179909,
|
| 1597 |
+
"num_tokens": 8229471.0,
|
| 1598 |
"step": 7700
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
+
"entropy": 5.325088725090027,
|
| 1602 |
"epoch": 4.461715601611974,
|
| 1603 |
+
"grad_norm": 1.2813605070114136,
|
| 1604 |
+
"learning_rate": 2.206246317030053e-07,
|
| 1605 |
+
"loss": 5.1722,
|
| 1606 |
+
"mean_token_accuracy": 0.32928301066160204,
|
| 1607 |
+
"num_tokens": 8281578.0,
|
| 1608 |
"step": 7750
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
+
"entropy": 5.231600880622864,
|
| 1612 |
"epoch": 4.490500863557858,
|
| 1613 |
+
"grad_norm": 1.0202534198760986,
|
| 1614 |
+
"learning_rate": 2.0883912787271654e-07,
|
| 1615 |
+
"loss": 5.0794,
|
| 1616 |
+
"mean_token_accuracy": 0.3391410967707634,
|
| 1617 |
+
"num_tokens": 8333192.0,
|
| 1618 |
"step": 7800
|
| 1619 |
},
|
| 1620 |
{
|
| 1621 |
+
"entropy": 5.243114166259765,
|
| 1622 |
"epoch": 4.519286125503742,
|
| 1623 |
+
"grad_norm": 1.8178458213806152,
|
| 1624 |
+
"learning_rate": 1.9705362404242783e-07,
|
| 1625 |
+
"loss": 5.0944,
|
| 1626 |
+
"mean_token_accuracy": 0.33702247083187103,
|
| 1627 |
+
"num_tokens": 8385403.0,
|
| 1628 |
"step": 7850
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
+
"entropy": 5.509276785850525,
|
| 1632 |
"epoch": 4.5480713874496255,
|
| 1633 |
+
"grad_norm": 1.2947449684143066,
|
| 1634 |
+
"learning_rate": 1.8526812021213906e-07,
|
| 1635 |
+
"loss": 5.3504,
|
| 1636 |
+
"mean_token_accuracy": 0.30472870826721193,
|
| 1637 |
+
"num_tokens": 8441398.0,
|
| 1638 |
"step": 7900
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
+
"entropy": 5.47810329914093,
|
| 1642 |
"epoch": 4.576856649395509,
|
| 1643 |
+
"grad_norm": 1.3030011653900146,
|
| 1644 |
+
"learning_rate": 1.7348261638185032e-07,
|
| 1645 |
+
"loss": 5.3205,
|
| 1646 |
+
"mean_token_accuracy": 0.3097413584589958,
|
| 1647 |
+
"num_tokens": 8496488.0,
|
| 1648 |
"step": 7950
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
+
"entropy": 5.325540552139282,
|
| 1652 |
"epoch": 4.605641911341393,
|
| 1653 |
+
"grad_norm": 1.807919979095459,
|
| 1654 |
+
"learning_rate": 1.6169711255156158e-07,
|
| 1655 |
+
"loss": 5.1724,
|
| 1656 |
+
"mean_token_accuracy": 0.32793802350759504,
|
| 1657 |
+
"num_tokens": 8549935.0,
|
| 1658 |
"step": 8000
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
+
"entropy": 5.597971448898315,
|
| 1662 |
"epoch": 4.634427173287277,
|
| 1663 |
+
"grad_norm": 2.5746006965637207,
|
| 1664 |
+
"learning_rate": 1.499116087212728e-07,
|
| 1665 |
+
"loss": 5.4384,
|
| 1666 |
+
"mean_token_accuracy": 0.2962429064512253,
|
| 1667 |
+
"num_tokens": 8605715.0,
|
| 1668 |
"step": 8050
|
| 1669 |
},
|
| 1670 |
{
|
| 1671 |
+
"entropy": 5.385147652626038,
|
| 1672 |
"epoch": 4.66321243523316,
|
| 1673 |
+
"grad_norm": 1.3230502605438232,
|
| 1674 |
+
"learning_rate": 1.381261048909841e-07,
|
| 1675 |
+
"loss": 5.2287,
|
| 1676 |
+
"mean_token_accuracy": 0.32063129514455796,
|
| 1677 |
+
"num_tokens": 8659693.0,
|
| 1678 |
"step": 8100
|
| 1679 |
},
|
| 1680 |
{
|
| 1681 |
+
"entropy": 5.237564296722412,
|
| 1682 |
"epoch": 4.691997697179044,
|
| 1683 |
+
"grad_norm": 1.5461162328720093,
|
| 1684 |
+
"learning_rate": 1.2634060106069533e-07,
|
| 1685 |
+
"loss": 5.0868,
|
| 1686 |
+
"mean_token_accuracy": 0.3380086237192154,
|
| 1687 |
+
"num_tokens": 8711203.0,
|
| 1688 |
"step": 8150
|
| 1689 |
},
|
| 1690 |
{
|
| 1691 |
+
"entropy": 5.3754426288604735,
|
| 1692 |
"epoch": 4.720782959124928,
|
| 1693 |
+
"grad_norm": 1.540111780166626,
|
| 1694 |
+
"learning_rate": 1.145550972304066e-07,
|
| 1695 |
+
"loss": 5.2201,
|
| 1696 |
+
"mean_token_accuracy": 0.3232385951280594,
|
| 1697 |
+
"num_tokens": 8763164.0,
|
| 1698 |
"step": 8200
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
+
"entropy": 5.358118782043457,
|
| 1702 |
"epoch": 4.7495682210708114,
|
| 1703 |
+
"grad_norm": 1.160130500793457,
|
| 1704 |
+
"learning_rate": 1.0276959340011786e-07,
|
| 1705 |
+
"loss": 5.1986,
|
| 1706 |
+
"mean_token_accuracy": 0.32435122996568677,
|
| 1707 |
+
"num_tokens": 8816843.0,
|
| 1708 |
"step": 8250
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
+
"entropy": 5.382785997390747,
|
| 1712 |
"epoch": 4.778353483016695,
|
| 1713 |
+
"grad_norm": 1.19467031955719,
|
| 1714 |
+
"learning_rate": 9.09840895698291e-08,
|
| 1715 |
+
"loss": 5.2294,
|
| 1716 |
+
"mean_token_accuracy": 0.320581151843071,
|
| 1717 |
+
"num_tokens": 8869513.0,
|
| 1718 |
"step": 8300
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
+
"entropy": 5.213022809028626,
|
| 1722 |
"epoch": 4.807138744962579,
|
| 1723 |
+
"grad_norm": 2.9769742488861084,
|
| 1724 |
+
"learning_rate": 7.919858573954036e-08,
|
| 1725 |
+
"loss": 5.0619,
|
| 1726 |
+
"mean_token_accuracy": 0.3414875140786171,
|
| 1727 |
+
"num_tokens": 8920491.0,
|
| 1728 |
"step": 8350
|
| 1729 |
},
|
| 1730 |
{
|
| 1731 |
+
"entropy": 5.5911472082138065,
|
| 1732 |
"epoch": 4.835924006908463,
|
| 1733 |
+
"grad_norm": 1.083928108215332,
|
| 1734 |
+
"learning_rate": 6.741308190925162e-08,
|
| 1735 |
+
"loss": 5.4315,
|
| 1736 |
+
"mean_token_accuracy": 0.2953196579217911,
|
| 1737 |
+
"num_tokens": 8976698.0,
|
| 1738 |
"step": 8400
|
| 1739 |
},
|
| 1740 |
{
|
| 1741 |
+
"entropy": 5.360013060569763,
|
| 1742 |
"epoch": 4.864709268854346,
|
| 1743 |
+
"grad_norm": 0.9796308279037476,
|
| 1744 |
+
"learning_rate": 5.562757807896287e-08,
|
| 1745 |
+
"loss": 5.1997,
|
| 1746 |
+
"mean_token_accuracy": 0.32382034361362455,
|
| 1747 |
+
"num_tokens": 9030719.0,
|
| 1748 |
"step": 8450
|
| 1749 |
},
|
| 1750 |
{
|
| 1751 |
+
"entropy": 5.447426581382752,
|
| 1752 |
"epoch": 4.89349453080023,
|
| 1753 |
+
"grad_norm": 1.2438750267028809,
|
| 1754 |
+
"learning_rate": 4.384207424867413e-08,
|
| 1755 |
+
"loss": 5.2921,
|
| 1756 |
+
"mean_token_accuracy": 0.31339066684246064,
|
| 1757 |
+
"num_tokens": 9084912.0,
|
| 1758 |
"step": 8500
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
+
"entropy": 5.419556441307068,
|
| 1762 |
"epoch": 4.922279792746114,
|
| 1763 |
+
"grad_norm": 1.0203313827514648,
|
| 1764 |
+
"learning_rate": 3.205657041838539e-08,
|
| 1765 |
+
"loss": 5.2623,
|
| 1766 |
+
"mean_token_accuracy": 0.31564354538917544,
|
| 1767 |
+
"num_tokens": 9139056.0,
|
| 1768 |
"step": 8550
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
+
"entropy": 5.24355613231659,
|
| 1772 |
"epoch": 4.951065054691997,
|
| 1773 |
+
"grad_norm": 1.5694708824157715,
|
| 1774 |
+
"learning_rate": 2.027106658809664e-08,
|
| 1775 |
+
"loss": 5.0942,
|
| 1776 |
+
"mean_token_accuracy": 0.3373285266757011,
|
| 1777 |
+
"num_tokens": 9190520.0,
|
| 1778 |
"step": 8600
|
| 1779 |
},
|
| 1780 |
{
|
| 1781 |
+
"entropy": 5.4742378282547,
|
| 1782 |
"epoch": 4.979850316637881,
|
| 1783 |
+
"grad_norm": 1.4123504161834717,
|
| 1784 |
+
"learning_rate": 8.485562757807896e-09,
|
| 1785 |
+
"loss": 5.3169,
|
| 1786 |
+
"mean_token_accuracy": 0.3118377339839935,
|
| 1787 |
+
"num_tokens": 9245023.0,
|
| 1788 |
"step": 8650
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 5.0,
|
| 1792 |
+
"eval_entropy": 5.6646331444313995,
|
| 1793 |
+
"eval_loss": 5.523473739624023,
|
| 1794 |
+
"eval_mean_token_accuracy": 0.27800820795347064,
|
| 1795 |
+
"eval_model_preparation_time": 0.0045,
|
| 1796 |
+
"eval_num_tokens": 9281810.0,
|
| 1797 |
+
"eval_runtime": 49.9701,
|
| 1798 |
+
"eval_samples_per_second": 8.685,
|
| 1799 |
+
"eval_steps_per_second": 4.343,
|
| 1800 |
"step": 8685
|
| 1801 |
}
|
| 1802 |
],
|
| 1803 |
"logging_steps": 50,
|
| 1804 |
+
"max_steps": 8685,
|
| 1805 |
"num_input_tokens_seen": 0,
|
| 1806 |
+
"num_train_epochs": 5,
|
| 1807 |
"save_steps": 500,
|
| 1808 |
"stateful_callbacks": {
|
| 1809 |
"TrainerControl": {
|
|
|
|
| 1812 |
"should_evaluate": false,
|
| 1813 |
"should_log": false,
|
| 1814 |
"should_save": true,
|
| 1815 |
+
"should_training_stop": true
|
| 1816 |
},
|
| 1817 |
"attributes": {}
|
| 1818 |
}
|
| 1819 |
},
|
| 1820 |
+
"total_flos": 1.2710932022552064e+17,
|
| 1821 |
"train_batch_size": 2,
|
| 1822 |
"trial_name": null,
|
| 1823 |
"trial_params": null
|
checkpoint-8685/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|
runs/Dec04_11-47-13_129-213-84-8/events.out.tfevents.1764848895.129-213-84-8.25442.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d77d61fb6af4433636be0b2505f03048f2b7390f5cb67124dae3bb4156bd4898
|
| 3 |
+
size 77711
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a666397e6243ddba6f7279c90610ed552907ef4de0be511faece3826d13e618
|
| 3 |
size 6225
|