Retain baseline metrics without trainer-state files
Browse files- ARTIFACTS.json +19 -20
- README.md +2 -7
- acc-step680/README.md +3 -2
- acc-step680/trainer_state.json +0 -1193
- base-acc-step680/README.md +3 -2
- base-acc-step680/trainer_state.json +0 -1193
- base-longtrace-step680/README.md +3 -2
- base-longtrace-step680/trainer_state.json +0 -1193
ARTIFACTS.json
CHANGED
|
@@ -219,12 +219,6 @@
|
|
| 219 |
"sha256": "f495d69ca65a8324038e6dcdad3fb416578c5ae4bcb6d7081e42821515c2ff0d",
|
| 220 |
"kind": "recipe_evidence"
|
| 221 |
},
|
| 222 |
-
{
|
| 223 |
-
"path": "acc-step680/trainer_state.json",
|
| 224 |
-
"bytes": 31921,
|
| 225 |
-
"sha256": "b1f0106fa055a902f0bf8fd58637feaed11da01d59492ed6f9a6531ac66af8c5",
|
| 226 |
-
"kind": "training_log"
|
| 227 |
-
},
|
| 228 |
{
|
| 229 |
"path": "acc-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 230 |
"bytes": 1090,
|
|
@@ -243,12 +237,6 @@
|
|
| 243 |
"sha256": "ae6add674f7e39081bf533ee6e31b128cf222231dcd99063f3fe44a5cb78a271",
|
| 244 |
"kind": "recipe_evidence"
|
| 245 |
},
|
| 246 |
-
{
|
| 247 |
-
"path": "base-acc-step680/trainer_state.json",
|
| 248 |
-
"bytes": 31950,
|
| 249 |
-
"sha256": "2570656fd7abcb7345b6b1b287d1b844d1d1277a6901fe232c724af93d5c4841",
|
| 250 |
-
"kind": "training_log"
|
| 251 |
-
},
|
| 252 |
{
|
| 253 |
"path": "base-acc-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 254 |
"bytes": 1095,
|
|
@@ -267,12 +255,6 @@
|
|
| 267 |
"sha256": "6e5e54c0f332fa4eb209893b29304480760d132047d21548287695b535f9b1d0",
|
| 268 |
"kind": "recipe_evidence"
|
| 269 |
},
|
| 270 |
-
{
|
| 271 |
-
"path": "base-longtrace-step680/trainer_state.json",
|
| 272 |
-
"bytes": 31573,
|
| 273 |
-
"sha256": "9cc249c6cbb4a93a627027306647d26a3dbf6936bc0eca98d4ec4d2077ae198c",
|
| 274 |
-
"kind": "training_log"
|
| 275 |
-
},
|
| 276 |
{
|
| 277 |
"path": "base-longtrace-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 278 |
"bytes": 1101,
|
|
@@ -304,7 +286,7 @@
|
|
| 304 |
"training_log",
|
| 305 |
"restore_map"
|
| 306 |
],
|
| 307 |
-
"note": "Standalone Xnhyacinth/Qwen3.5-4B-{ACC,Base-ACC,Base-LongTrace}-128k-SFT repositories were removed after their weight/loading files were confirmed byte-identical to these subfolders; each subfolder now also carries that run's README, args.json
|
| 308 |
},
|
| 309 |
"consolidation": {
|
| 310 |
"standalone_repos_removed": true,
|
|
@@ -314,5 +296,22 @@
|
|
| 314 |
"Xnhyacinth/Qwen3.5-4B-Base-LongTrace-128k-SFT"
|
| 315 |
]
|
| 316 |
},
|
| 317 |
-
"research_selected_manifest": "research-selected/MANIFEST.json"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 318 |
}
|
|
|
|
| 219 |
"sha256": "f495d69ca65a8324038e6dcdad3fb416578c5ae4bcb6d7081e42821515c2ff0d",
|
| 220 |
"kind": "recipe_evidence"
|
| 221 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 222 |
{
|
| 223 |
"path": "acc-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 224 |
"bytes": 1090,
|
|
|
|
| 237 |
"sha256": "ae6add674f7e39081bf533ee6e31b128cf222231dcd99063f3fe44a5cb78a271",
|
| 238 |
"kind": "recipe_evidence"
|
| 239 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 240 |
{
|
| 241 |
"path": "base-acc-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 242 |
"bytes": 1095,
|
|
|
|
| 255 |
"sha256": "6e5e54c0f332fa4eb209893b29304480760d132047d21548287695b535f9b1d0",
|
| 256 |
"kind": "recipe_evidence"
|
| 257 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 258 |
{
|
| 259 |
"path": "base-longtrace-step680/CLEANUP_20260930_RESTORE_MAP.json",
|
| 260 |
"bytes": 1101,
|
|
|
|
| 286 |
"training_log",
|
| 287 |
"restore_map"
|
| 288 |
],
|
| 289 |
+
"note": "Standalone Xnhyacinth/Qwen3.5-4B-{ACC,Base-ACC,Base-LongTrace}-128k-SFT repositories were removed after their weight/loading files were confirmed byte-identical to these subfolders; each subfolder now also carries that run's README, args.json and restore map; numeric metric histories were preserved in private GitHub and current trainer-state files were retired. Paths in args.json are sanitized to /workspace/<user>/...; no credentials are present."
|
| 290 |
},
|
| 291 |
"consolidation": {
|
| 292 |
"standalone_repos_removed": true,
|
|
|
|
| 296 |
"Xnhyacinth/Qwen3.5-4B-Base-LongTrace-128k-SFT"
|
| 297 |
]
|
| 298 |
},
|
| 299 |
+
"research_selected_manifest": "research-selected/MANIFEST.json",
|
| 300 |
+
"retired_trainer_state": {
|
| 301 |
+
"source_revision": "5643e6d81bdadfe56d22d76a75751144d092d8a2",
|
| 302 |
+
"paths": [
|
| 303 |
+
"acc-step680/trainer_state.json",
|
| 304 |
+
"base-acc-step680/trainer_state.json",
|
| 305 |
+
"base-longtrace-step680/trainer_state.json"
|
| 306 |
+
],
|
| 307 |
+
"numeric_metrics_repository": "Xnhyacinth/longworld",
|
| 308 |
+
"numeric_metrics_revision": "c1e6d05f574bd735c379147943c10619cde2f39b",
|
| 309 |
+
"numeric_metrics_path": "docs/research/baseline-training-metrics-20261030.json",
|
| 310 |
+
"source_sha256": {
|
| 311 |
+
"acc-step680": "b1f0106fa055a902f0bf8fd58637feaed11da01d59492ed6f9a6531ac66af8c5",
|
| 312 |
+
"base-acc-step680": "2570656fd7abcb7345b6b1b287d1b844d1d1277a6901fe232c724af93d5c4841",
|
| 313 |
+
"base-longtrace-step680": "9cc249c6cbb4a93a627027306647d26a3dbf6936bc0eca98d4ec4d2077ae198c"
|
| 314 |
+
},
|
| 315 |
+
"scope": "Only numeric log histories retained; current trainer-state files removed. Historical Git blobs may remain."
|
| 316 |
+
}
|
| 317 |
}
|
README.md
CHANGED
|
@@ -46,9 +46,7 @@ validation covers the complete safetensors headers, shard index and loading
|
|
| 46 |
files; inference has not been executed as part of this publication review.
|
| 47 |
`ARTIFACTS.json` records the exact published paths, sizes and SHA-256 identities.
|
| 48 |
|
| 49 |
-
Each subfolder carries the weight and loading files plus
|
| 50 |
-
training recipe and log (see [Standalone model cards](#standalone-model-cards)
|
| 51 |
-
below). Optimizer/RNG state, historical working snapshots and the removed
|
| 52 |
`training_args.bin` pickles are excluded. The original file hashes are retained
|
| 53 |
for the selected exports. Shared tokenizer files appear in each subfolder so
|
| 54 |
each variant can be downloaded independently; no storage-saving claim is made.
|
|
@@ -60,10 +58,7 @@ Apache-2.0. The companion training datasets retain their own terms.
|
|
| 60 |
## Standalone model cards
|
| 61 |
|
| 62 |
Each subfolder is a complete, independently loadable model card: it carries the
|
| 63 |
-
variant's own `args.json` (full training recipe), `
|
| 64 |
-
train loss and the seven periodic eval-loss points), `CLEANUP_20260930_RESTORE_MAP.json`
|
| 65 |
-
(restore map for the intentionally removed `training_args.bin` pickle) and a
|
| 66 |
-
`README.md` with the run summary. Loading a subfolder needs no other repository.
|
| 67 |
|
| 68 |
```python
|
| 69 |
from transformers import AutoProcessor, AutoModelForCausalLM
|
|
|
|
| 46 |
files; inference has not been executed as part of this publication review.
|
| 47 |
`ARTIFACTS.json` records the exact published paths, sizes and SHA-256 identities.
|
| 48 |
|
| 49 |
+
Each baseline subfolder carries the weight and loading files plus its training recipe (see [Standalone model cards](#standalone-model-cards) below). Numeric training metrics are retained in a private research record. Optimizer/RNG state, historical working snapshots and the removed
|
|
|
|
|
|
|
| 50 |
`training_args.bin` pickles are excluded. The original file hashes are retained
|
| 51 |
for the selected exports. Shared tokenizer files appear in each subfolder so
|
| 52 |
each variant can be downloaded independently; no storage-saving claim is made.
|
|
|
|
| 58 |
## Standalone model cards
|
| 59 |
|
| 60 |
Each subfolder is a complete, independently loadable model card: it carries the
|
| 61 |
+
variant's own `args.json` (full training recipe), `CLEANUP_20260930_RESTORE_MAP.json` (restore map for the removed `training_args.bin` pickle) and a `README.md` with the run summary. Per-step numeric training and evaluation histories are retained in [private research records](https://github.com/Xnhyacinth/longworld/blob/worlds/docs/research/baseline-training-metrics-20261030.json); current model folders no longer contain `trainer_state.json`. Loading a subfolder needs no other repository.
|
|
|
|
|
|
|
|
|
|
| 62 |
|
| 63 |
```python
|
| 64 |
from transformers import AutoProcessor, AutoModelForCausalLM
|
acc-step680/README.md
CHANGED
|
@@ -48,7 +48,7 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 48 |
|
| 49 |
### Train loss vs eval loss
|
| 50 |
|
| 51 |
-
|
| 52 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 53 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 54 |
interchangeable, so both are stated here for ACC:
|
|
@@ -59,4 +59,5 @@ interchangeable, so both are stated here for ACC:
|
|
| 59 |
| Train loss, final step 680 | 0.3645 |
|
| 60 |
| Eval loss, step 680 | 0.37456 |
|
| 61 |
|
| 62 |
-
The single `Train loss` figure in the table above (0.365) is the final step-680 train-loss value (0.3645), i.e. the last logged step, not a run mean.
|
|
|
|
|
|
| 48 |
|
| 49 |
### Train loss vs eval loss
|
| 50 |
|
| 51 |
+
The preserved numeric training metrics record per-step train loss
|
| 52 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 53 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 54 |
interchangeable, so both are stated here for ACC:
|
|
|
|
| 59 |
| Train loss, final step 680 | 0.3645 |
|
| 60 |
| Eval loss, step 680 | 0.37456 |
|
| 61 |
|
| 62 |
+
The single `Train loss` figure in the table above (0.365) is the final step-680 train-loss value (0.3645), i.e. the last logged step, not a run mean.
|
| 63 |
+
The numeric metric history is preserved in [private research records](https://github.com/Xnhyacinth/longworld/blob/worlds/docs/research/baseline-training-metrics-20261030.json). The current model folder does not include the trainer-state file.
|
acc-step680/trainer_state.json
DELETED
|
@@ -1,1193 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": 600,
|
| 3 |
-
"best_metric": 0.37431228,
|
| 4 |
-
"best_model_checkpoint": "/workspace/<user>/longworld/data/sft/swift_ext_acc/v6-20260825-153450/checkpoint-600",
|
| 5 |
-
"epoch": 1.0089136490250696,
|
| 6 |
-
"eval_steps": 100,
|
| 7 |
-
"global_step": 680,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0014856081708449396,
|
| 14 |
-
"grad_norm": 8.125,
|
| 15 |
-
"learning_rate": 2.9411764705882356e-07,
|
| 16 |
-
"loss": 0.6485965251922607,
|
| 17 |
-
"step": 1,
|
| 18 |
-
"token_acc": 0.8125101719741767
|
| 19 |
-
},
|
| 20 |
-
{
|
| 21 |
-
"epoch": 0.007428040854224698,
|
| 22 |
-
"grad_norm": 8.0625,
|
| 23 |
-
"learning_rate": 1.4705882352941177e-06,
|
| 24 |
-
"loss": 0.5434028506278992,
|
| 25 |
-
"step": 5,
|
| 26 |
-
"token_acc": 0.8689615769175832
|
| 27 |
-
},
|
| 28 |
-
{
|
| 29 |
-
"epoch": 0.014856081708449397,
|
| 30 |
-
"grad_norm": 4.625,
|
| 31 |
-
"learning_rate": 2.9411764705882355e-06,
|
| 32 |
-
"loss": 0.6126736164093017,
|
| 33 |
-
"step": 10,
|
| 34 |
-
"token_acc": 0.8189102151230194
|
| 35 |
-
},
|
| 36 |
-
{
|
| 37 |
-
"epoch": 0.022284122562674095,
|
| 38 |
-
"grad_norm": 3.28125,
|
| 39 |
-
"learning_rate": 4.411764705882353e-06,
|
| 40 |
-
"loss": 0.530322790145874,
|
| 41 |
-
"step": 15,
|
| 42 |
-
"token_acc": 0.8443691391892134
|
| 43 |
-
},
|
| 44 |
-
{
|
| 45 |
-
"epoch": 0.029712163416898793,
|
| 46 |
-
"grad_norm": 6.09375,
|
| 47 |
-
"learning_rate": 5.882352941176471e-06,
|
| 48 |
-
"loss": 0.5225050926208497,
|
| 49 |
-
"step": 20,
|
| 50 |
-
"token_acc": 0.8445983017635532
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"epoch": 0.03714020427112349,
|
| 54 |
-
"grad_norm": 5.4375,
|
| 55 |
-
"learning_rate": 7.352941176470589e-06,
|
| 56 |
-
"loss": 0.46790456771850586,
|
| 57 |
-
"step": 25,
|
| 58 |
-
"token_acc": 0.8630275589057381
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.04456824512534819,
|
| 62 |
-
"grad_norm": 4.5,
|
| 63 |
-
"learning_rate": 8.823529411764707e-06,
|
| 64 |
-
"loss": 0.47516918182373047,
|
| 65 |
-
"step": 30,
|
| 66 |
-
"token_acc": 0.8613025520524219
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.051996285979572884,
|
| 70 |
-
"grad_norm": 1.90625,
|
| 71 |
-
"learning_rate": 9.999940874631278e-06,
|
| 72 |
-
"loss": 0.4786482810974121,
|
| 73 |
-
"step": 35,
|
| 74 |
-
"token_acc": 0.8575864787432461
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 0.059424326833797586,
|
| 78 |
-
"grad_norm": 2.015625,
|
| 79 |
-
"learning_rate": 9.997871633546257e-06,
|
| 80 |
-
"loss": 0.4074864387512207,
|
| 81 |
-
"step": 40,
|
| 82 |
-
"token_acc": 0.8824653513135472
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 0.06685236768802229,
|
| 86 |
-
"grad_norm": 7.3125,
|
| 87 |
-
"learning_rate": 9.992847522200132e-06,
|
| 88 |
-
"loss": 0.40807528495788575,
|
| 89 |
-
"step": 45,
|
| 90 |
-
"token_acc": 0.8890855919954285
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.07428040854224698,
|
| 94 |
-
"grad_norm": 3.046875,
|
| 95 |
-
"learning_rate": 9.98487151097676e-06,
|
| 96 |
-
"loss": 0.44022340774536134,
|
| 97 |
-
"step": 50,
|
| 98 |
-
"token_acc": 0.8636080914230544
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 0.08170844939647168,
|
| 102 |
-
"grad_norm": 2.46875,
|
| 103 |
-
"learning_rate": 9.973948315499126e-06,
|
| 104 |
-
"loss": 0.4931690216064453,
|
| 105 |
-
"step": 55,
|
| 106 |
-
"token_acc": 0.8451892724877588
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 0.08913649025069638,
|
| 110 |
-
"grad_norm": 2.890625,
|
| 111 |
-
"learning_rate": 9.960084393841355e-06,
|
| 112 |
-
"loss": 0.4535623550415039,
|
| 113 |
-
"step": 60,
|
| 114 |
-
"token_acc": 0.8550124446188642
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.09656453110492108,
|
| 118 |
-
"grad_norm": 2.90625,
|
| 119 |
-
"learning_rate": 9.943287942710527e-06,
|
| 120 |
-
"loss": 0.4557951927185059,
|
| 121 |
-
"step": 65,
|
| 122 |
-
"token_acc": 0.8678312039960508
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.10399257195914577,
|
| 126 |
-
"grad_norm": 2.375,
|
| 127 |
-
"learning_rate": 9.923568892600579e-06,
|
| 128 |
-
"loss": 0.4425069332122803,
|
| 129 |
-
"step": 70,
|
| 130 |
-
"token_acc": 0.8687427846855542
|
| 131 |
-
},
|
| 132 |
-
{
|
| 133 |
-
"epoch": 0.11142061281337047,
|
| 134 |
-
"grad_norm": 3.015625,
|
| 135 |
-
"learning_rate": 9.90093890192113e-06,
|
| 136 |
-
"loss": 0.47688751220703124,
|
| 137 |
-
"step": 75,
|
| 138 |
-
"token_acc": 0.8499013507360753
|
| 139 |
-
},
|
| 140 |
-
{
|
| 141 |
-
"epoch": 0.11884865366759517,
|
| 142 |
-
"grad_norm": 4.25,
|
| 143 |
-
"learning_rate": 9.875411350104745e-06,
|
| 144 |
-
"loss": 0.4012148857116699,
|
| 145 |
-
"step": 80,
|
| 146 |
-
"token_acc": 0.8724830914129571
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.12627669452181986,
|
| 150 |
-
"grad_norm": 2.203125,
|
| 151 |
-
"learning_rate": 9.847001329696653e-06,
|
| 152 |
-
"loss": 0.4277793407440186,
|
| 153 |
-
"step": 85,
|
| 154 |
-
"token_acc": 0.8715421870512148
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 0.13370473537604458,
|
| 158 |
-
"grad_norm": 3.953125,
|
| 159 |
-
"learning_rate": 9.815725637431663e-06,
|
| 160 |
-
"loss": 0.4845582962036133,
|
| 161 |
-
"step": 90,
|
| 162 |
-
"token_acc": 0.8600017738134456
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"epoch": 0.14113277623026926,
|
| 166 |
-
"grad_norm": 2.5,
|
| 167 |
-
"learning_rate": 9.781602764303488e-06,
|
| 168 |
-
"loss": 0.4465078353881836,
|
| 169 |
-
"step": 95,
|
| 170 |
-
"token_acc": 0.8624471714445746
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.14856081708449395,
|
| 174 |
-
"grad_norm": 3.546875,
|
| 175 |
-
"learning_rate": 9.744652884632406e-06,
|
| 176 |
-
"loss": 0.43661985397338865,
|
| 177 |
-
"step": 100,
|
| 178 |
-
"token_acc": 0.8652658710147364
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.14856081708449395,
|
| 182 |
-
"eval_loss": 0.40439748764038086,
|
| 183 |
-
"eval_runtime": 22.9438,
|
| 184 |
-
"eval_samples_per_second": 1.395,
|
| 185 |
-
"eval_steps_per_second": 0.174,
|
| 186 |
-
"eval_token_acc": 0.8682214105085698,
|
| 187 |
-
"step": 100
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 0.15598885793871867,
|
| 191 |
-
"grad_norm": 3.546875,
|
| 192 |
-
"learning_rate": 9.704897844137674e-06,
|
| 193 |
-
"loss": 0.48310060501098634,
|
| 194 |
-
"step": 105,
|
| 195 |
-
"token_acc": 0.8515976357906306
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.16341689879294335,
|
| 199 |
-
"grad_norm": 3.21875,
|
| 200 |
-
"learning_rate": 9.66236114702178e-06,
|
| 201 |
-
"loss": 0.48354787826538087,
|
| 202 |
-
"step": 110,
|
| 203 |
-
"token_acc": 0.8476185747446034
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.17084493964716807,
|
| 207 |
-
"grad_norm": 1.8046875,
|
| 208 |
-
"learning_rate": 9.617067942074155e-06,
|
| 209 |
-
"loss": 0.3864017963409424,
|
| 210 |
-
"step": 115,
|
| 211 |
-
"token_acc": 0.8848053648499795
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.17827298050139276,
|
| 215 |
-
"grad_norm": 3.875,
|
| 216 |
-
"learning_rate": 9.569045007802558e-06,
|
| 217 |
-
"loss": 0.4695082187652588,
|
| 218 |
-
"step": 120,
|
| 219 |
-
"token_acc": 0.8506263786573438
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.18570102135561745,
|
| 223 |
-
"grad_norm": 1.1171875,
|
| 224 |
-
"learning_rate": 9.518320736600943e-06,
|
| 225 |
-
"loss": 0.39969921112060547,
|
| 226 |
-
"step": 125,
|
| 227 |
-
"token_acc": 0.8941288817879639
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.19312906220984216,
|
| 231 |
-
"grad_norm": 2.15625,
|
| 232 |
-
"learning_rate": 9.464925117963133e-06,
|
| 233 |
-
"loss": 0.4764643669128418,
|
| 234 |
-
"step": 130,
|
| 235 |
-
"token_acc": 0.8507799732580598
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.20055710306406685,
|
| 239 |
-
"grad_norm": 3.546875,
|
| 240 |
-
"learning_rate": 9.408889720752265e-06,
|
| 241 |
-
"loss": 0.466965389251709,
|
| 242 |
-
"step": 135,
|
| 243 |
-
"token_acc": 0.8652596513144197
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.20798514391829154,
|
| 247 |
-
"grad_norm": 1.875,
|
| 248 |
-
"learning_rate": 9.35024767453647e-06,
|
| 249 |
-
"loss": 0.3879223823547363,
|
| 250 |
-
"step": 140,
|
| 251 |
-
"token_acc": 0.8817286828316135
|
| 252 |
-
},
|
| 253 |
-
{
|
| 254 |
-
"epoch": 0.21541318477251625,
|
| 255 |
-
"grad_norm": 2.828125,
|
| 256 |
-
"learning_rate": 9.289033650001817e-06,
|
| 257 |
-
"loss": 0.38793110847473145,
|
| 258 |
-
"step": 145,
|
| 259 |
-
"token_acc": 0.8770929127136565
|
| 260 |
-
},
|
| 261 |
-
{
|
| 262 |
-
"epoch": 0.22284122562674094,
|
| 263 |
-
"grad_norm": 3.59375,
|
| 264 |
-
"learning_rate": 9.225283838454111e-06,
|
| 265 |
-
"loss": 0.42479572296142576,
|
| 266 |
-
"step": 150,
|
| 267 |
-
"token_acc": 0.8670651468816607
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.23026926648096566,
|
| 271 |
-
"grad_norm": 3.421875,
|
| 272 |
-
"learning_rate": 9.159035930421658e-06,
|
| 273 |
-
"loss": 0.42569942474365235,
|
| 274 |
-
"step": 155,
|
| 275 |
-
"token_acc": 0.8611438756520143
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.23769730733519034,
|
| 279 |
-
"grad_norm": 3.171875,
|
| 280 |
-
"learning_rate": 9.090329093371667e-06,
|
| 281 |
-
"loss": 0.44130630493164064,
|
| 282 |
-
"step": 160,
|
| 283 |
-
"token_acc": 0.8677522290967669
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.24512534818941503,
|
| 287 |
-
"grad_norm": 1.5078125,
|
| 288 |
-
"learning_rate": 9.019203948553422e-06,
|
| 289 |
-
"loss": 0.4441518306732178,
|
| 290 |
-
"step": 165,
|
| 291 |
-
"token_acc": 0.8792056153612526
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.2525533890436397,
|
| 295 |
-
"grad_norm": 3.453125,
|
| 296 |
-
"learning_rate": 8.94570254698197e-06,
|
| 297 |
-
"loss": 0.4152205467224121,
|
| 298 |
-
"step": 170,
|
| 299 |
-
"token_acc": 0.8796524413432761
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 0.25998142989786444,
|
| 303 |
-
"grad_norm": 2.546875,
|
| 304 |
-
"learning_rate": 8.86986834457646e-06,
|
| 305 |
-
"loss": 0.41381039619445803,
|
| 306 |
-
"step": 175,
|
| 307 |
-
"token_acc": 0.8745695138220967
|
| 308 |
-
},
|
| 309 |
-
{
|
| 310 |
-
"epoch": 0.26740947075208915,
|
| 311 |
-
"grad_norm": 3.703125,
|
| 312 |
-
"learning_rate": 8.791746176467908e-06,
|
| 313 |
-
"loss": 0.43212070465087893,
|
| 314 |
-
"step": 180,
|
| 315 |
-
"token_acc": 0.8632481703738306
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.2748375116063138,
|
| 319 |
-
"grad_norm": 3.03125,
|
| 320 |
-
"learning_rate": 8.711382230491494e-06,
|
| 321 |
-
"loss": 0.46705055236816406,
|
| 322 |
-
"step": 185,
|
| 323 |
-
"token_acc": 0.8539889993439975
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 0.2822655524605385,
|
| 327 |
-
"grad_norm": 4.375,
|
| 328 |
-
"learning_rate": 8.628824019879137e-06,
|
| 329 |
-
"loss": 0.43026342391967776,
|
| 330 |
-
"step": 190,
|
| 331 |
-
"token_acc": 0.874124122164896
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.28969359331476324,
|
| 335 |
-
"grad_norm": 2.09375,
|
| 336 |
-
"learning_rate": 8.544120355168451e-06,
|
| 337 |
-
"loss": 0.4596715927124023,
|
| 338 |
-
"step": 195,
|
| 339 |
-
"token_acc": 0.8622756806930693
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.2971216341689879,
|
| 343 |
-
"grad_norm": 4.09375,
|
| 344 |
-
"learning_rate": 8.457321315344695e-06,
|
| 345 |
-
"loss": 0.4292288780212402,
|
| 346 |
-
"step": 200,
|
| 347 |
-
"token_acc": 0.8666689452457448
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 0.2971216341689879,
|
| 351 |
-
"eval_loss": 0.3911648392677307,
|
| 352 |
-
"eval_runtime": 21.5029,
|
| 353 |
-
"eval_samples_per_second": 1.488,
|
| 354 |
-
"eval_steps_per_second": 0.186,
|
| 355 |
-
"eval_token_acc": 0.8711872873154132,
|
| 356 |
-
"step": 200
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 0.3045496750232126,
|
| 360 |
-
"grad_norm": 1.7421875,
|
| 361 |
-
"learning_rate": 8.368478218232787e-06,
|
| 362 |
-
"loss": 0.4239497184753418,
|
| 363 |
-
"step": 205,
|
| 364 |
-
"token_acc": 0.8727043710697872
|
| 365 |
-
},
|
| 366 |
-
{
|
| 367 |
-
"epoch": 0.31197771587743733,
|
| 368 |
-
"grad_norm": 5.15625,
|
| 369 |
-
"learning_rate": 8.277643590156893e-06,
|
| 370 |
-
"loss": 0.4996319770812988,
|
| 371 |
-
"step": 210,
|
| 372 |
-
"token_acc": 0.8516739090519616
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 0.31940575673166205,
|
| 376 |
-
"grad_norm": 3.0,
|
| 377 |
-
"learning_rate": 8.184871134885512e-06,
|
| 378 |
-
"loss": 0.4748319149017334,
|
| 379 |
-
"step": 215,
|
| 380 |
-
"token_acc": 0.8464925076249834
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.3268337975858867,
|
| 384 |
-
"grad_norm": 4.125,
|
| 385 |
-
"learning_rate": 8.090215701880418e-06,
|
| 386 |
-
"loss": 0.3902214527130127,
|
| 387 |
-
"step": 220,
|
| 388 |
-
"token_acc": 0.8806779202458137
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.3342618384401114,
|
| 392 |
-
"grad_norm": 1.9296875,
|
| 393 |
-
"learning_rate": 7.993733253868256e-06,
|
| 394 |
-
"loss": 0.41803436279296874,
|
| 395 |
-
"step": 225,
|
| 396 |
-
"token_acc": 0.8740076214671324
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 0.34168987929433614,
|
| 400 |
-
"grad_norm": 3.03125,
|
| 401 |
-
"learning_rate": 7.895480833753942e-06,
|
| 402 |
-
"loss": 0.43894367218017577,
|
| 403 |
-
"step": 230,
|
| 404 |
-
"token_acc": 0.8568047520341553
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.3491179201485608,
|
| 408 |
-
"grad_norm": 2.921875,
|
| 409 |
-
"learning_rate": 7.795516530895414e-06,
|
| 410 |
-
"loss": 0.4782839775085449,
|
| 411 |
-
"step": 235,
|
| 412 |
-
"token_acc": 0.8487684729064039
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 0.3565459610027855,
|
| 416 |
-
"grad_norm": 2.8125,
|
| 417 |
-
"learning_rate": 7.693899446759727e-06,
|
| 418 |
-
"loss": 0.4114743709564209,
|
| 419 |
-
"step": 240,
|
| 420 |
-
"token_acc": 0.8697130786016064
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 0.36397400185701023,
|
| 424 |
-
"grad_norm": 2.796875,
|
| 425 |
-
"learning_rate": 7.59068965998074e-06,
|
| 426 |
-
"loss": 0.44709053039550783,
|
| 427 |
-
"step": 245,
|
| 428 |
-
"token_acc": 0.8643808575341004
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 0.3714020427112349,
|
| 432 |
-
"grad_norm": 2.640625,
|
| 433 |
-
"learning_rate": 7.485948190839076e-06,
|
| 434 |
-
"loss": 0.4201956748962402,
|
| 435 |
-
"step": 250,
|
| 436 |
-
"token_acc": 0.8747247929394497
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.3788300835654596,
|
| 440 |
-
"grad_norm": 3.453125,
|
| 441 |
-
"learning_rate": 7.379736965185369e-06,
|
| 442 |
-
"loss": 0.3354783296585083,
|
| 443 |
-
"step": 255,
|
| 444 |
-
"token_acc": 0.9012714725694616
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.3862581244196843,
|
| 448 |
-
"grad_norm": 2.6875,
|
| 449 |
-
"learning_rate": 7.272118777828109e-06,
|
| 450 |
-
"loss": 0.40779767036437986,
|
| 451 |
-
"step": 260,
|
| 452 |
-
"token_acc": 0.8715497990174185
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 0.393686165273909,
|
| 456 |
-
"grad_norm": 2.4375,
|
| 457 |
-
"learning_rate": 7.163157255407732e-06,
|
| 458 |
-
"loss": 0.45515804290771483,
|
| 459 |
-
"step": 265,
|
| 460 |
-
"token_acc": 0.855320992215668
|
| 461 |
-
},
|
| 462 |
-
{
|
| 463 |
-
"epoch": 0.4011142061281337,
|
| 464 |
-
"grad_norm": 3.1875,
|
| 465 |
-
"learning_rate": 7.052916818778918e-06,
|
| 466 |
-
"loss": 0.48943204879760743,
|
| 467 |
-
"step": 270,
|
| 468 |
-
"token_acc": 0.8431622712876934
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.4085422469823584,
|
| 472 |
-
"grad_norm": 4.125,
|
| 473 |
-
"learning_rate": 6.941462644923318e-06,
|
| 474 |
-
"loss": 0.4512819290161133,
|
| 475 |
-
"step": 275,
|
| 476 |
-
"token_acc": 0.8618729787380445
|
| 477 |
-
},
|
| 478 |
-
{
|
| 479 |
-
"epoch": 0.4159702878365831,
|
| 480 |
-
"grad_norm": 2.453125,
|
| 481 |
-
"learning_rate": 6.8288606284152535e-06,
|
| 482 |
-
"loss": 0.4588315486907959,
|
| 483 |
-
"step": 280,
|
| 484 |
-
"token_acc": 0.8563293003567494
|
| 485 |
-
},
|
| 486 |
-
{
|
| 487 |
-
"epoch": 0.4233983286908078,
|
| 488 |
-
"grad_norm": 2.21875,
|
| 489 |
-
"learning_rate": 6.715177342463145e-06,
|
| 490 |
-
"loss": 0.44280128479003905,
|
| 491 |
-
"step": 285,
|
| 492 |
-
"token_acc": 0.863991293231524
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.4308263695450325,
|
| 496 |
-
"grad_norm": 2.234375,
|
| 497 |
-
"learning_rate": 6.600479999549721e-06,
|
| 498 |
-
"loss": 0.43950786590576174,
|
| 499 |
-
"step": 290,
|
| 500 |
-
"token_acc": 0.8633808420105856
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.4382544103992572,
|
| 504 |
-
"grad_norm": 1.7578125,
|
| 505 |
-
"learning_rate": 6.4848364116942665e-06,
|
| 506 |
-
"loss": 0.39799890518188474,
|
| 507 |
-
"step": 295,
|
| 508 |
-
"token_acc": 0.8769681946723458
|
| 509 |
-
},
|
| 510 |
-
{
|
| 511 |
-
"epoch": 0.4456824512534819,
|
| 512 |
-
"grad_norm": 2.28125,
|
| 513 |
-
"learning_rate": 6.368314950360416e-06,
|
| 514 |
-
"loss": 0.4450007438659668,
|
| 515 |
-
"step": 300,
|
| 516 |
-
"token_acc": 0.8609575344077167
|
| 517 |
-
},
|
| 518 |
-
{
|
| 519 |
-
"epoch": 0.4456824512534819,
|
| 520 |
-
"eval_loss": 0.38438016176223755,
|
| 521 |
-
"eval_runtime": 26.7866,
|
| 522 |
-
"eval_samples_per_second": 1.195,
|
| 523 |
-
"eval_steps_per_second": 0.149,
|
| 524 |
-
"eval_token_acc": 0.8722487590147046,
|
| 525 |
-
"step": 300
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 0.4531104921077066,
|
| 529 |
-
"grad_norm": 4.34375,
|
| 530 |
-
"learning_rate": 6.250984506033183e-06,
|
| 531 |
-
"loss": 0.4448129653930664,
|
| 532 |
-
"step": 305,
|
| 533 |
-
"token_acc": 0.8685314355735696
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.4605385329619313,
|
| 537 |
-
"grad_norm": 2.875,
|
| 538 |
-
"learning_rate": 6.132914447489137e-06,
|
| 539 |
-
"loss": 0.41407175064086915,
|
| 540 |
-
"step": 310,
|
| 541 |
-
"token_acc": 0.8718791625397042
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.467966573816156,
|
| 545 |
-
"grad_norm": 3.34375,
|
| 546 |
-
"learning_rate": 6.014174580783794e-06,
|
| 547 |
-
"loss": 0.46277647018432616,
|
| 548 |
-
"step": 315,
|
| 549 |
-
"token_acc": 0.8528253948338326
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.4753946146703807,
|
| 553 |
-
"grad_norm": 3.3125,
|
| 554 |
-
"learning_rate": 5.8948351079804875e-06,
|
| 555 |
-
"loss": 0.42051358222961427,
|
| 556 |
-
"step": 320,
|
| 557 |
-
"token_acc": 0.8765277678507613
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 0.4828226555246054,
|
| 561 |
-
"grad_norm": 2.78125,
|
| 562 |
-
"learning_rate": 5.774966585645092e-06,
|
| 563 |
-
"loss": 0.3983988046646118,
|
| 564 |
-
"step": 325,
|
| 565 |
-
"token_acc": 0.8770934109348802
|
| 566 |
-
},
|
| 567 |
-
{
|
| 568 |
-
"epoch": 0.49025069637883006,
|
| 569 |
-
"grad_norm": 1.5,
|
| 570 |
-
"learning_rate": 5.6546398831311774e-06,
|
| 571 |
-
"loss": 0.4116243362426758,
|
| 572 |
-
"step": 330,
|
| 573 |
-
"token_acc": 0.8746705736262362
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 0.4976787372330548,
|
| 577 |
-
"grad_norm": 3.109375,
|
| 578 |
-
"learning_rate": 5.533926140680222e-06,
|
| 579 |
-
"loss": 0.4553972244262695,
|
| 580 |
-
"step": 335,
|
| 581 |
-
"token_acc": 0.8610639411318659
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 0.5051067780872794,
|
| 585 |
-
"grad_norm": 2.796875,
|
| 586 |
-
"learning_rate": 5.412896727361663e-06,
|
| 587 |
-
"loss": 0.5090607643127442,
|
| 588 |
-
"step": 340,
|
| 589 |
-
"token_acc": 0.8447516118558375
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 0.5125348189415042,
|
| 593 |
-
"grad_norm": 2.4375,
|
| 594 |
-
"learning_rate": 5.29162319887768e-06,
|
| 595 |
-
"loss": 0.43277683258056643,
|
| 596 |
-
"step": 345,
|
| 597 |
-
"token_acc": 0.870130086031337
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.5199628597957289,
|
| 601 |
-
"grad_norm": 2.421875,
|
| 602 |
-
"learning_rate": 5.170177255257618e-06,
|
| 603 |
-
"loss": 0.4166229724884033,
|
| 604 |
-
"step": 350,
|
| 605 |
-
"token_acc": 0.868445220773608
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.5273909006499535,
|
| 609 |
-
"grad_norm": 1.7890625,
|
| 610 |
-
"learning_rate": 5.048630698467081e-06,
|
| 611 |
-
"loss": 0.43834514617919923,
|
| 612 |
-
"step": 355,
|
| 613 |
-
"token_acc": 0.8757114039229151
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 0.5348189415041783,
|
| 617 |
-
"grad_norm": 2.375,
|
| 618 |
-
"learning_rate": 4.927055389956768e-06,
|
| 619 |
-
"loss": 0.40874271392822265,
|
| 620 |
-
"step": 360,
|
| 621 |
-
"token_acc": 0.8767359613559906
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 0.542246982358403,
|
| 625 |
-
"grad_norm": 2.625,
|
| 626 |
-
"learning_rate": 4.805523208176139e-06,
|
| 627 |
-
"loss": 0.4303419589996338,
|
| 628 |
-
"step": 365,
|
| 629 |
-
"token_acc": 0.8604305484366991
|
| 630 |
-
},
|
| 631 |
-
{
|
| 632 |
-
"epoch": 0.5496750232126276,
|
| 633 |
-
"grad_norm": 1.4140625,
|
| 634 |
-
"learning_rate": 4.684106006077015e-06,
|
| 635 |
-
"loss": 0.36066598892211915,
|
| 636 |
-
"step": 370,
|
| 637 |
-
"token_acc": 0.8917032083241562
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 0.5571030640668524,
|
| 641 |
-
"grad_norm": 2.65625,
|
| 642 |
-
"learning_rate": 4.562875568632278e-06,
|
| 643 |
-
"loss": 0.4652695655822754,
|
| 644 |
-
"step": 375,
|
| 645 |
-
"token_acc": 0.8548486945122173
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 0.564531104921077,
|
| 649 |
-
"grad_norm": 1.8515625,
|
| 650 |
-
"learning_rate": 4.441903570394739e-06,
|
| 651 |
-
"loss": 0.3517768383026123,
|
| 652 |
-
"step": 380,
|
| 653 |
-
"token_acc": 0.9004434081266006
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.5719591457753017,
|
| 657 |
-
"grad_norm": 3.53125,
|
| 658 |
-
"learning_rate": 4.321261533121303e-06,
|
| 659 |
-
"loss": 0.42608466148376467,
|
| 660 |
-
"step": 385,
|
| 661 |
-
"token_acc": 0.8685075932061436
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.5793871866295265,
|
| 665 |
-
"grad_norm": 2.234375,
|
| 666 |
-
"learning_rate": 4.201020783487465e-06,
|
| 667 |
-
"loss": 0.4557151794433594,
|
| 668 |
-
"step": 390,
|
| 669 |
-
"token_acc": 0.8716891814654679
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.5868152274837511,
|
| 673 |
-
"grad_norm": 2.0625,
|
| 674 |
-
"learning_rate": 4.081252410917148e-06,
|
| 675 |
-
"loss": 0.42630910873413086,
|
| 676 |
-
"step": 395,
|
| 677 |
-
"token_acc": 0.8680198375860367
|
| 678 |
-
},
|
| 679 |
-
{
|
| 680 |
-
"epoch": 0.5942432683379758,
|
| 681 |
-
"grad_norm": 3.46875,
|
| 682 |
-
"learning_rate": 3.962027225552807e-06,
|
| 683 |
-
"loss": 0.4209193229675293,
|
| 684 |
-
"step": 400,
|
| 685 |
-
"token_acc": 0.8673176147867546
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 0.5942432683379758,
|
| 689 |
-
"eval_loss": 0.375412255525589,
|
| 690 |
-
"eval_runtime": 21.517,
|
| 691 |
-
"eval_samples_per_second": 1.487,
|
| 692 |
-
"eval_steps_per_second": 0.186,
|
| 693 |
-
"eval_token_acc": 0.8742156036339795,
|
| 694 |
-
"step": 400
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 0.6016713091922006,
|
| 698 |
-
"grad_norm": 4.90625,
|
| 699 |
-
"learning_rate": 3.843415716390644e-06,
|
| 700 |
-
"loss": 0.4258245944976807,
|
| 701 |
-
"step": 405,
|
| 702 |
-
"token_acc": 0.8705012881344171
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 0.6090993500464252,
|
| 706 |
-
"grad_norm": 2.765625,
|
| 707 |
-
"learning_rate": 3.725488009605708e-06,
|
| 708 |
-
"loss": 0.43317670822143556,
|
| 709 |
-
"step": 410,
|
| 710 |
-
"token_acc": 0.8678888940070938
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.61652739090065,
|
| 714 |
-
"grad_norm": 1.3671875,
|
| 715 |
-
"learning_rate": 3.6083138270914932e-06,
|
| 716 |
-
"loss": 0.47187175750732424,
|
| 717 |
-
"step": 415,
|
| 718 |
-
"token_acc": 0.8574750947928688
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 0.6239554317548747,
|
| 722 |
-
"grad_norm": 4.3125,
|
| 723 |
-
"learning_rate": 3.491962445238569e-06,
|
| 724 |
-
"loss": 0.43051533699035643,
|
| 725 |
-
"step": 420,
|
| 726 |
-
"token_acc": 0.8631317038492902
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 0.6313834726090993,
|
| 730 |
-
"grad_norm": 3.625,
|
| 731 |
-
"learning_rate": 3.3765026539765832e-06,
|
| 732 |
-
"loss": 0.421526575088501,
|
| 733 |
-
"step": 425,
|
| 734 |
-
"token_acc": 0.8785516804903234
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 0.6388115134633241,
|
| 738 |
-
"grad_norm": 3.140625,
|
| 739 |
-
"learning_rate": 3.2620027161038975e-06,
|
| 740 |
-
"loss": 0.4537972450256348,
|
| 741 |
-
"step": 430,
|
| 742 |
-
"token_acc": 0.8588126953866535
|
| 743 |
-
},
|
| 744 |
-
{
|
| 745 |
-
"epoch": 0.6462395543175488,
|
| 746 |
-
"grad_norm": 2.375,
|
| 747 |
-
"learning_rate": 3.1485303269288603e-06,
|
| 748 |
-
"loss": 0.39894444942474366,
|
| 749 |
-
"step": 435,
|
| 750 |
-
"token_acc": 0.8804231536926148
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 0.6536675951717734,
|
| 754 |
-
"grad_norm": 1.8515625,
|
| 755 |
-
"learning_rate": 3.0361525742465975e-06,
|
| 756 |
-
"loss": 0.3826523542404175,
|
| 757 |
-
"step": 440,
|
| 758 |
-
"token_acc": 0.8821679109280474
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.6610956360259982,
|
| 762 |
-
"grad_norm": 2.96875,
|
| 763 |
-
"learning_rate": 2.9249358986749922e-06,
|
| 764 |
-
"loss": 0.35647809505462646,
|
| 765 |
-
"step": 445,
|
| 766 |
-
"token_acc": 0.8984789623874334
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.6685236768802229,
|
| 770 |
-
"grad_norm": 2.859375,
|
| 771 |
-
"learning_rate": 2.8149460543732666e-06,
|
| 772 |
-
"loss": 0.4285717010498047,
|
| 773 |
-
"step": 450,
|
| 774 |
-
"token_acc": 0.8741057516235797
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 0.6759517177344475,
|
| 778 |
-
"grad_norm": 2.78125,
|
| 779 |
-
"learning_rate": 2.706248070166449e-06,
|
| 780 |
-
"loss": 0.5106130599975586,
|
| 781 |
-
"step": 455,
|
| 782 |
-
"token_acc": 0.8288131613957896
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 0.6833797585886723,
|
| 786 |
-
"grad_norm": 2.203125,
|
| 787 |
-
"learning_rate": 2.598906211098643e-06,
|
| 788 |
-
"loss": 0.4468533515930176,
|
| 789 |
-
"step": 460,
|
| 790 |
-
"token_acc": 0.8640778353326063
|
| 791 |
-
},
|
| 792 |
-
{
|
| 793 |
-
"epoch": 0.6908077994428969,
|
| 794 |
-
"grad_norm": 3.265625,
|
| 795 |
-
"learning_rate": 2.4929839404378935e-06,
|
| 796 |
-
"loss": 0.40039472579956054,
|
| 797 |
-
"step": 465,
|
| 798 |
-
"token_acc": 0.8920586328362599
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 0.6982358402971216,
|
| 802 |
-
"grad_norm": 3.3125,
|
| 803 |
-
"learning_rate": 2.388543882155067e-06,
|
| 804 |
-
"loss": 0.5190415382385254,
|
| 805 |
-
"step": 470,
|
| 806 |
-
"token_acc": 0.8275356471595438
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 0.7056638811513464,
|
| 810 |
-
"grad_norm": 1.9375,
|
| 811 |
-
"learning_rate": 2.2856477838989455e-06,
|
| 812 |
-
"loss": 0.38921270370483396,
|
| 813 |
-
"step": 475,
|
| 814 |
-
"token_acc": 0.8814649504444595
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 0.713091922005571,
|
| 818 |
-
"grad_norm": 3.625,
|
| 819 |
-
"learning_rate": 2.1843564804894316e-06,
|
| 820 |
-
"loss": 0.39909958839416504,
|
| 821 |
-
"step": 480,
|
| 822 |
-
"token_acc": 0.8813185592931304
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.7205199628597957,
|
| 826 |
-
"grad_norm": 2.765625,
|
| 827 |
-
"learning_rate": 2.0847298579504344e-06,
|
| 828 |
-
"loss": 0.4496356964111328,
|
| 829 |
-
"step": 485,
|
| 830 |
-
"token_acc": 0.8778593038216782
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 0.7279480037140205,
|
| 834 |
-
"grad_norm": 2.1875,
|
| 835 |
-
"learning_rate": 1.9868268181037186e-06,
|
| 836 |
-
"loss": 0.3959397077560425,
|
| 837 |
-
"step": 490,
|
| 838 |
-
"token_acc": 0.8756656570679445
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 0.7353760445682451,
|
| 842 |
-
"grad_norm": 3.515625,
|
| 843 |
-
"learning_rate": 1.8907052437446271e-06,
|
| 844 |
-
"loss": 0.4474188804626465,
|
| 845 |
-
"step": 495,
|
| 846 |
-
"token_acc": 0.8597172096702428
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"epoch": 0.7428040854224698,
|
| 850 |
-
"grad_norm": 1.7734375,
|
| 851 |
-
"learning_rate": 1.7964219644202852e-06,
|
| 852 |
-
"loss": 0.3955270528793335,
|
| 853 |
-
"step": 500,
|
| 854 |
-
"token_acc": 0.8754822239802951
|
| 855 |
-
},
|
| 856 |
-
{
|
| 857 |
-
"epoch": 0.7428040854224698,
|
| 858 |
-
"eval_loss": 0.37557822465896606,
|
| 859 |
-
"eval_runtime": 21.5987,
|
| 860 |
-
"eval_samples_per_second": 1.482,
|
| 861 |
-
"eval_steps_per_second": 0.185,
|
| 862 |
-
"eval_token_acc": 0.8747151197277637,
|
| 863 |
-
"step": 500
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.7502321262766946,
|
| 867 |
-
"grad_norm": 3.703125,
|
| 868 |
-
"learning_rate": 1.704032722830512e-06,
|
| 869 |
-
"loss": 0.4031815528869629,
|
| 870 |
-
"step": 505,
|
| 871 |
-
"token_acc": 0.8899086291669767
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.7576601671309192,
|
| 875 |
-
"grad_norm": 2.59375,
|
| 876 |
-
"learning_rate": 1.6135921418712959e-06,
|
| 877 |
-
"loss": 0.3656553983688354,
|
| 878 |
-
"step": 510,
|
| 879 |
-
"token_acc": 0.8893852921175185
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 0.7650882079851439,
|
| 883 |
-
"grad_norm": 3.234375,
|
| 884 |
-
"learning_rate": 1.5251536923403427e-06,
|
| 885 |
-
"loss": 0.4692857265472412,
|
| 886 |
-
"step": 515,
|
| 887 |
-
"token_acc": 0.8574473676526411
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 0.7725162488393686,
|
| 891 |
-
"grad_norm": 2.421875,
|
| 892 |
-
"learning_rate": 1.4387696613237612e-06,
|
| 893 |
-
"loss": 0.3626351594924927,
|
| 894 |
-
"step": 520,
|
| 895 |
-
"token_acc": 0.9045857183276383
|
| 896 |
-
},
|
| 897 |
-
{
|
| 898 |
-
"epoch": 0.7799442896935933,
|
| 899 |
-
"grad_norm": 2.6875,
|
| 900 |
-
"learning_rate": 1.3544911212825907e-06,
|
| 901 |
-
"loss": 0.3447706699371338,
|
| 902 |
-
"step": 525,
|
| 903 |
-
"token_acc": 0.9002377381370266
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 0.787372330547818,
|
| 907 |
-
"grad_norm": 2.421875,
|
| 908 |
-
"learning_rate": 1.2723678998574512e-06,
|
| 909 |
-
"loss": 0.3924096584320068,
|
| 910 |
-
"step": 530,
|
| 911 |
-
"token_acc": 0.8750451620838362
|
| 912 |
-
},
|
| 913 |
-
{
|
| 914 |
-
"epoch": 0.7948003714020427,
|
| 915 |
-
"grad_norm": 2.296875,
|
| 916 |
-
"learning_rate": 1.1924485504091565e-06,
|
| 917 |
-
"loss": 0.3919008493423462,
|
| 918 |
-
"step": 535,
|
| 919 |
-
"token_acc": 0.8835618533054692
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.8022284122562674,
|
| 923 |
-
"grad_norm": 3.03125,
|
| 924 |
-
"learning_rate": 1.1147803233127241e-06,
|
| 925 |
-
"loss": 0.4453620910644531,
|
| 926 |
-
"step": 540,
|
| 927 |
-
"token_acc": 0.85780114808503
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.8096564531104921,
|
| 931 |
-
"grad_norm": 1.375,
|
| 932 |
-
"learning_rate": 1.0394091380217354e-06,
|
| 933 |
-
"loss": 0.3620155334472656,
|
| 934 |
-
"step": 545,
|
| 935 |
-
"token_acc": 0.9008442736532335
|
| 936 |
-
},
|
| 937 |
-
{
|
| 938 |
-
"epoch": 0.8170844939647168,
|
| 939 |
-
"grad_norm": 2.640625,
|
| 940 |
-
"learning_rate": 9.663795559195733e-07,
|
| 941 |
-
"loss": 0.42974562644958497,
|
| 942 |
-
"step": 550,
|
| 943 |
-
"token_acc": 0.864184678274324
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 0.8245125348189415,
|
| 947 |
-
"grad_norm": 2.140625,
|
| 948 |
-
"learning_rate": 8.957347539735872e-07,
|
| 949 |
-
"loss": 0.3691257953643799,
|
| 950 |
-
"step": 555,
|
| 951 |
-
"token_acc": 0.8839409856326278
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 0.8319405756731661,
|
| 955 |
-
"grad_norm": 4.09375,
|
| 956 |
-
"learning_rate": 8.275164992077555e-07,
|
| 957 |
-
"loss": 0.4493074417114258,
|
| 958 |
-
"step": 560,
|
| 959 |
-
"token_acc": 0.8598278829604131
|
| 960 |
-
},
|
| 961 |
-
{
|
| 962 |
-
"epoch": 0.8393686165273909,
|
| 963 |
-
"grad_norm": 2.25,
|
| 964 |
-
"learning_rate": 7.617651240089546e-07,
|
| 965 |
-
"loss": 0.372517991065979,
|
| 966 |
-
"step": 565,
|
| 967 |
-
"token_acc": 0.881947067049085
|
| 968 |
-
},
|
| 969 |
-
{
|
| 970 |
-
"epoch": 0.8467966573816156,
|
| 971 |
-
"grad_norm": 3.171875,
|
| 972 |
-
"learning_rate": 6.985195022814068e-07,
|
| 973 |
-
"loss": 0.3768314838409424,
|
| 974 |
-
"step": 570,
|
| 975 |
-
"token_acc": 0.8848727047231888
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.8542246982358404,
|
| 979 |
-
"grad_norm": 2.859375,
|
| 980 |
-
"learning_rate": 6.37817026463432e-07,
|
| 981 |
-
"loss": 0.4177062511444092,
|
| 982 |
-
"step": 575,
|
| 983 |
-
"token_acc": 0.8703034841210413
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.861652739090065,
|
| 987 |
-
"grad_norm": 3.046875,
|
| 988 |
-
"learning_rate": 5.796935854200764e-07,
|
| 989 |
-
"loss": 0.41587390899658205,
|
| 990 |
-
"step": 580,
|
| 991 |
-
"token_acc": 0.8682770947300006
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 0.8690807799442897,
|
| 995 |
-
"grad_norm": 3.015625,
|
| 996 |
-
"learning_rate": 5.241835432246888e-07,
|
| 997 |
-
"loss": 0.4005917549133301,
|
| 998 |
-
"step": 585,
|
| 999 |
-
"token_acc": 0.8771229681463664
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.8765088207985144,
|
| 1003 |
-
"grad_norm": 2.484375,
|
| 1004 |
-
"learning_rate": 4.713197188420027e-07,
|
| 1005 |
-
"loss": 0.41893763542175294,
|
| 1006 |
-
"step": 590,
|
| 1007 |
-
"token_acc": 0.8923310283292438
|
| 1008 |
-
},
|
| 1009 |
-
{
|
| 1010 |
-
"epoch": 0.8839368616527391,
|
| 1011 |
-
"grad_norm": 1.53125,
|
| 1012 |
-
"learning_rate": 4.211333667247125e-07,
|
| 1013 |
-
"loss": 0.34410154819488525,
|
| 1014 |
-
"step": 595,
|
| 1015 |
-
"token_acc": 0.891663522449709
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 0.8913649025069638,
|
| 1019 |
-
"grad_norm": 2.484375,
|
| 1020 |
-
"learning_rate": 3.736541583350473e-07,
|
| 1021 |
-
"loss": 0.3968889474868774,
|
| 1022 |
-
"step": 600,
|
| 1023 |
-
"token_acc": 0.882126174798807
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 0.8913649025069638,
|
| 1027 |
-
"eval_loss": 0.37431228160858154,
|
| 1028 |
-
"eval_runtime": 21.5806,
|
| 1029 |
-
"eval_samples_per_second": 1.483,
|
| 1030 |
-
"eval_steps_per_second": 0.185,
|
| 1031 |
-
"eval_token_acc": 0.8744965814367331,
|
| 1032 |
-
"step": 600
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.8987929433611885,
|
| 1036 |
-
"grad_norm": 3.03125,
|
| 1037 |
-
"learning_rate": 3.289101646022297e-07,
|
| 1038 |
-
"loss": 0.44832868576049806,
|
| 1039 |
-
"step": 605,
|
| 1040 |
-
"token_acc": 0.855783849043105
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 0.9062209842154132,
|
| 1044 |
-
"grad_norm": 2.59375,
|
| 1045 |
-
"learning_rate": 2.869278393262226e-07,
|
| 1046 |
-
"loss": 0.43210883140563966,
|
| 1047 |
-
"step": 610,
|
| 1048 |
-
"token_acc": 0.8670671125118955
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 0.9136490250696379,
|
| 1052 |
-
"grad_norm": 1.734375,
|
| 1053 |
-
"learning_rate": 2.47732003537568e-07,
|
| 1054 |
-
"loss": 0.3760815143585205,
|
| 1055 |
-
"step": 615,
|
| 1056 |
-
"token_acc": 0.8866364087849024
|
| 1057 |
-
},
|
| 1058 |
-
{
|
| 1059 |
-
"epoch": 0.9210770659238626,
|
| 1060 |
-
"grad_norm": 4.90625,
|
| 1061 |
-
"learning_rate": 2.113458308225458e-07,
|
| 1062 |
-
"loss": 0.45517740249633787,
|
| 1063 |
-
"step": 620,
|
| 1064 |
-
"token_acc": 0.8552380306677062
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 0.9285051067780873,
|
| 1068 |
-
"grad_norm": 5.1875,
|
| 1069 |
-
"learning_rate": 1.777908336223655e-07,
|
| 1070 |
-
"loss": 0.4327892303466797,
|
| 1071 |
-
"step": 625,
|
| 1072 |
-
"token_acc": 0.8757005381025744
|
| 1073 |
-
},
|
| 1074 |
-
{
|
| 1075 |
-
"epoch": 0.935933147632312,
|
| 1076 |
-
"grad_norm": 2.0625,
|
| 1077 |
-
"learning_rate": 1.4708685051444515e-07,
|
| 1078 |
-
"loss": 0.38370471000671386,
|
| 1079 |
-
"step": 630,
|
| 1080 |
-
"token_acc": 0.881797869801948
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 0.9433611884865367,
|
| 1084 |
-
"grad_norm": 2.828125,
|
| 1085 |
-
"learning_rate": 1.19252034483342e-07,
|
| 1086 |
-
"loss": 0.4558698654174805,
|
| 1087 |
-
"step": 635,
|
| 1088 |
-
"token_acc": 0.8566193042626163
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.9507892293407614,
|
| 1092 |
-
"grad_norm": 2.34375,
|
| 1093 |
-
"learning_rate": 9.430284218824026e-08,
|
| 1094 |
-
"loss": 0.42174806594848635,
|
| 1095 |
-
"step": 640,
|
| 1096 |
-
"token_acc": 0.8646135032503057
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 0.958217270194986,
|
| 1100 |
-
"grad_norm": 1.3671875,
|
| 1101 |
-
"learning_rate": 7.225402423334694e-08,
|
| 1102 |
-
"loss": 0.39884974956512453,
|
| 1103 |
-
"step": 645,
|
| 1104 |
-
"token_acc": 0.8791346469512318
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 0.9656453110492108,
|
| 1108 |
-
"grad_norm": 4.0,
|
| 1109 |
-
"learning_rate": 5.3118616446960484e-08,
|
| 1110 |
-
"loss": 0.4783008575439453,
|
| 1111 |
-
"step": 650,
|
| 1112 |
-
"token_acc": 0.8564445083456416
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 0.9730733519034355,
|
| 1116 |
-
"grad_norm": 2.25,
|
| 1117 |
-
"learning_rate": 3.690793217434985e-08,
|
| 1118 |
-
"loss": 0.39383721351623535,
|
| 1119 |
-
"step": 655,
|
| 1120 |
-
"token_acc": 0.876372252994461
|
| 1121 |
-
},
|
| 1122 |
-
{
|
| 1123 |
-
"epoch": 0.9805013927576601,
|
| 1124 |
-
"grad_norm": 3.171875,
|
| 1125 |
-
"learning_rate": 2.363155558901542e-08,
|
| 1126 |
-
"loss": 0.4122488021850586,
|
| 1127 |
-
"step": 660,
|
| 1128 |
-
"token_acc": 0.8804762912590984
|
| 1129 |
-
},
|
| 1130 |
-
{
|
| 1131 |
-
"epoch": 0.9879294336118849,
|
| 1132 |
-
"grad_norm": 2.609375,
|
| 1133 |
-
"learning_rate": 1.3297336026280027e-08,
|
| 1134 |
-
"loss": 0.43268842697143556,
|
| 1135 |
-
"step": 665,
|
| 1136 |
-
"token_acc": 0.8618618618618619
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.9953574744661096,
|
| 1140 |
-
"grad_norm": 2.046875,
|
| 1141 |
-
"learning_rate": 5.911383342556143e-09,
|
| 1142 |
-
"loss": 0.4280752658843994,
|
| 1143 |
-
"step": 670,
|
| 1144 |
-
"token_acc": 0.8623605205260509
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 1.001485608170845,
|
| 1148 |
-
"grad_norm": 2.875,
|
| 1149 |
-
"learning_rate": 1.4780643030476439e-09,
|
| 1150 |
-
"loss": 0.42536125183105467,
|
| 1151 |
-
"step": 675,
|
| 1152 |
-
"token_acc": 0.8773729913164987
|
| 1153 |
-
},
|
| 1154 |
-
{
|
| 1155 |
-
"epoch": 1.0089136490250696,
|
| 1156 |
-
"grad_norm": 3.015625,
|
| 1157 |
-
"learning_rate": 0.0,
|
| 1158 |
-
"loss": 0.36453890800476074,
|
| 1159 |
-
"step": 680,
|
| 1160 |
-
"token_acc": 0.8981918139863314
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 1.0089136490250696,
|
| 1164 |
-
"eval_loss": 0.3745608925819397,
|
| 1165 |
-
"eval_runtime": 26.1049,
|
| 1166 |
-
"eval_samples_per_second": 1.226,
|
| 1167 |
-
"eval_steps_per_second": 0.153,
|
| 1168 |
-
"eval_token_acc": 0.8744965814367331,
|
| 1169 |
-
"step": 680
|
| 1170 |
-
}
|
| 1171 |
-
],
|
| 1172 |
-
"logging_steps": 5,
|
| 1173 |
-
"max_steps": 680,
|
| 1174 |
-
"num_input_tokens_seen": 0,
|
| 1175 |
-
"num_train_epochs": 2,
|
| 1176 |
-
"save_steps": 100,
|
| 1177 |
-
"stateful_callbacks": {
|
| 1178 |
-
"TrainerControl": {
|
| 1179 |
-
"args": {
|
| 1180 |
-
"should_epoch_stop": false,
|
| 1181 |
-
"should_evaluate": false,
|
| 1182 |
-
"should_log": false,
|
| 1183 |
-
"should_save": true,
|
| 1184 |
-
"should_training_stop": true
|
| 1185 |
-
},
|
| 1186 |
-
"attributes": {}
|
| 1187 |
-
}
|
| 1188 |
-
},
|
| 1189 |
-
"total_flos": 8.096724013710678e+19,
|
| 1190 |
-
"train_batch_size": 1,
|
| 1191 |
-
"trial_name": null,
|
| 1192 |
-
"trial_params": null
|
| 1193 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
base-acc-step680/README.md
CHANGED
|
@@ -47,7 +47,7 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 47 |
|
| 48 |
### Train loss vs eval loss
|
| 49 |
|
| 50 |
-
|
| 51 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 52 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 53 |
interchangeable, so both are stated here for Base-ACC:
|
|
@@ -58,4 +58,5 @@ interchangeable, so both are stated here for Base-ACC:
|
|
| 58 |
| Train loss, final step 680 | 0.3725 |
|
| 59 |
| Eval loss, step 680 | 0.37545 |
|
| 60 |
|
| 61 |
-
The single `Train loss` figure in the table above (0.437) corresponds to the mean over the 137 logged train-loss entries (0.4382, ≈0.438), not to the final step (0.3725). Use the precise values above when comparing runs.
|
|
|
|
|
|
| 47 |
|
| 48 |
### Train loss vs eval loss
|
| 49 |
|
| 50 |
+
The preserved numeric training metrics record per-step train loss
|
| 51 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 52 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 53 |
interchangeable, so both are stated here for Base-ACC:
|
|
|
|
| 58 |
| Train loss, final step 680 | 0.3725 |
|
| 59 |
| Eval loss, step 680 | 0.37545 |
|
| 60 |
|
| 61 |
+
The single `Train loss` figure in the table above (0.437) corresponds to the mean over the 137 logged train-loss entries (0.4382, ≈0.438), not to the final step (0.3725). Use the precise values above when comparing runs.
|
| 62 |
+
The numeric metric history is preserved in [private research records](https://github.com/Xnhyacinth/longworld/blob/worlds/docs/research/baseline-training-metrics-20261030.json). The current model folder does not include the trainer-state file.
|
base-acc-step680/trainer_state.json
DELETED
|
@@ -1,1193 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": 680,
|
| 3 |
-
"best_metric": 0.37545016,
|
| 4 |
-
"best_model_checkpoint": "/workspace/<user>/longworld/data/sft/swift_ext_acc_base/v1-20260903-163953/checkpoint-680",
|
| 5 |
-
"epoch": 1.0089136490250696,
|
| 6 |
-
"eval_steps": 100,
|
| 7 |
-
"global_step": 680,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0014856081708449396,
|
| 14 |
-
"grad_norm": 6.6875,
|
| 15 |
-
"learning_rate": 2.9411764705882356e-07,
|
| 16 |
-
"loss": 0.650237500667572,
|
| 17 |
-
"step": 1,
|
| 18 |
-
"token_acc": 0.8104486518743558
|
| 19 |
-
},
|
| 20 |
-
{
|
| 21 |
-
"epoch": 0.007428040854224698,
|
| 22 |
-
"grad_norm": 6.40625,
|
| 23 |
-
"learning_rate": 1.4705882352941177e-06,
|
| 24 |
-
"loss": 0.5539965629577637,
|
| 25 |
-
"step": 5,
|
| 26 |
-
"token_acc": 0.8675808217873637
|
| 27 |
-
},
|
| 28 |
-
{
|
| 29 |
-
"epoch": 0.014856081708449397,
|
| 30 |
-
"grad_norm": 3.28125,
|
| 31 |
-
"learning_rate": 2.9411764705882355e-06,
|
| 32 |
-
"loss": 0.6171924114227295,
|
| 33 |
-
"step": 10,
|
| 34 |
-
"token_acc": 0.8181759629009404
|
| 35 |
-
},
|
| 36 |
-
{
|
| 37 |
-
"epoch": 0.022284122562674095,
|
| 38 |
-
"grad_norm": 1.921875,
|
| 39 |
-
"learning_rate": 4.411764705882353e-06,
|
| 40 |
-
"loss": 0.5352516651153565,
|
| 41 |
-
"step": 15,
|
| 42 |
-
"token_acc": 0.8441988058738099
|
| 43 |
-
},
|
| 44 |
-
{
|
| 45 |
-
"epoch": 0.029712163416898793,
|
| 46 |
-
"grad_norm": 6.625,
|
| 47 |
-
"learning_rate": 5.882352941176471e-06,
|
| 48 |
-
"loss": 0.5350956439971923,
|
| 49 |
-
"step": 20,
|
| 50 |
-
"token_acc": 0.8428478118876551
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"epoch": 0.03714020427112349,
|
| 54 |
-
"grad_norm": 4.375,
|
| 55 |
-
"learning_rate": 7.352941176470589e-06,
|
| 56 |
-
"loss": 0.47702994346618655,
|
| 57 |
-
"step": 25,
|
| 58 |
-
"token_acc": 0.8602704035817724
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.04456824512534819,
|
| 62 |
-
"grad_norm": 3.0625,
|
| 63 |
-
"learning_rate": 8.823529411764707e-06,
|
| 64 |
-
"loss": 0.4857781410217285,
|
| 65 |
-
"step": 30,
|
| 66 |
-
"token_acc": 0.8594614890808009
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.051996285979572884,
|
| 70 |
-
"grad_norm": 1.5625,
|
| 71 |
-
"learning_rate": 9.999940874631278e-06,
|
| 72 |
-
"loss": 0.4885265350341797,
|
| 73 |
-
"step": 35,
|
| 74 |
-
"token_acc": 0.8559809256446248
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 0.059424326833797586,
|
| 78 |
-
"grad_norm": 1.7109375,
|
| 79 |
-
"learning_rate": 9.997871633546257e-06,
|
| 80 |
-
"loss": 0.41361289024353026,
|
| 81 |
-
"step": 40,
|
| 82 |
-
"token_acc": 0.8821993912469587
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 0.06685236768802229,
|
| 86 |
-
"grad_norm": 5.5625,
|
| 87 |
-
"learning_rate": 9.992847522200132e-06,
|
| 88 |
-
"loss": 0.4190025806427002,
|
| 89 |
-
"step": 45,
|
| 90 |
-
"token_acc": 0.8874833331465899
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.07428040854224698,
|
| 94 |
-
"grad_norm": 2.625,
|
| 95 |
-
"learning_rate": 9.98487151097676e-06,
|
| 96 |
-
"loss": 0.4505626201629639,
|
| 97 |
-
"step": 50,
|
| 98 |
-
"token_acc": 0.8612408210250745
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 0.08170844939647168,
|
| 102 |
-
"grad_norm": 2.171875,
|
| 103 |
-
"learning_rate": 9.973948315499126e-06,
|
| 104 |
-
"loss": 0.5010731220245361,
|
| 105 |
-
"step": 55,
|
| 106 |
-
"token_acc": 0.8440210772251634
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 0.08913649025069638,
|
| 110 |
-
"grad_norm": 2.578125,
|
| 111 |
-
"learning_rate": 9.960084393841355e-06,
|
| 112 |
-
"loss": 0.4615337371826172,
|
| 113 |
-
"step": 60,
|
| 114 |
-
"token_acc": 0.8533393921242602
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.09656453110492108,
|
| 118 |
-
"grad_norm": 2.640625,
|
| 119 |
-
"learning_rate": 9.943287942710527e-06,
|
| 120 |
-
"loss": 0.46202521324157714,
|
| 121 |
-
"step": 65,
|
| 122 |
-
"token_acc": 0.8668243712182676
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.10399257195914577,
|
| 126 |
-
"grad_norm": 2.0625,
|
| 127 |
-
"learning_rate": 9.923568892600579e-06,
|
| 128 |
-
"loss": 0.4475461483001709,
|
| 129 |
-
"step": 70,
|
| 130 |
-
"token_acc": 0.8687333219780844
|
| 131 |
-
},
|
| 132 |
-
{
|
| 133 |
-
"epoch": 0.11142061281337047,
|
| 134 |
-
"grad_norm": 2.71875,
|
| 135 |
-
"learning_rate": 9.90093890192113e-06,
|
| 136 |
-
"loss": 0.48204903602600097,
|
| 137 |
-
"step": 75,
|
| 138 |
-
"token_acc": 0.8499519401021906
|
| 139 |
-
},
|
| 140 |
-
{
|
| 141 |
-
"epoch": 0.11884865366759517,
|
| 142 |
-
"grad_norm": 3.734375,
|
| 143 |
-
"learning_rate": 9.875411350104745e-06,
|
| 144 |
-
"loss": 0.40892276763916013,
|
| 145 |
-
"step": 80,
|
| 146 |
-
"token_acc": 0.8711749772291235
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.12627669452181986,
|
| 150 |
-
"grad_norm": 1.6953125,
|
| 151 |
-
"learning_rate": 9.847001329696653e-06,
|
| 152 |
-
"loss": 0.43460965156555176,
|
| 153 |
-
"step": 85,
|
| 154 |
-
"token_acc": 0.8699079897392217
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 0.13370473537604458,
|
| 158 |
-
"grad_norm": 3.1875,
|
| 159 |
-
"learning_rate": 9.815725637431663e-06,
|
| 160 |
-
"loss": 0.45804848670959475,
|
| 161 |
-
"step": 90,
|
| 162 |
-
"token_acc": 0.8734954260953298
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"epoch": 0.14113277623026926,
|
| 166 |
-
"grad_norm": 2.09375,
|
| 167 |
-
"learning_rate": 9.781602764303488e-06,
|
| 168 |
-
"loss": 0.45301246643066406,
|
| 169 |
-
"step": 95,
|
| 170 |
-
"token_acc": 0.8610010693008809
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.14856081708449395,
|
| 174 |
-
"grad_norm": 3.1875,
|
| 175 |
-
"learning_rate": 9.744652884632406e-06,
|
| 176 |
-
"loss": 0.44422125816345215,
|
| 177 |
-
"step": 100,
|
| 178 |
-
"token_acc": 0.8636409480584972
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.14856081708449395,
|
| 182 |
-
"eval_loss": 0.4148000478744507,
|
| 183 |
-
"eval_runtime": 36.3786,
|
| 184 |
-
"eval_samples_per_second": 0.88,
|
| 185 |
-
"eval_steps_per_second": 0.22,
|
| 186 |
-
"eval_token_acc": 0.8672535980768631,
|
| 187 |
-
"step": 100
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 0.15598885793871867,
|
| 191 |
-
"grad_norm": 3.125,
|
| 192 |
-
"learning_rate": 9.704897844137674e-06,
|
| 193 |
-
"loss": 0.48714504241943357,
|
| 194 |
-
"step": 105,
|
| 195 |
-
"token_acc": 0.8499489698641813
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.16341689879294335,
|
| 199 |
-
"grad_norm": 2.671875,
|
| 200 |
-
"learning_rate": 9.66236114702178e-06,
|
| 201 |
-
"loss": 0.4881488800048828,
|
| 202 |
-
"step": 110,
|
| 203 |
-
"token_acc": 0.847469619294696
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.17084493964716807,
|
| 207 |
-
"grad_norm": 1.4140625,
|
| 208 |
-
"learning_rate": 9.617067942074155e-06,
|
| 209 |
-
"loss": 0.3898077249526978,
|
| 210 |
-
"step": 115,
|
| 211 |
-
"token_acc": 0.8851032581210555
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.17827298050139276,
|
| 215 |
-
"grad_norm": 3.390625,
|
| 216 |
-
"learning_rate": 9.569045007802558e-06,
|
| 217 |
-
"loss": 0.47558770179748533,
|
| 218 |
-
"step": 120,
|
| 219 |
-
"token_acc": 0.8496829955189301
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.18570102135561745,
|
| 223 |
-
"grad_norm": 1.1171875,
|
| 224 |
-
"learning_rate": 9.518320736600943e-06,
|
| 225 |
-
"loss": 0.4049226760864258,
|
| 226 |
-
"step": 125,
|
| 227 |
-
"token_acc": 0.8933943033441684
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.19312906220984216,
|
| 231 |
-
"grad_norm": 2.328125,
|
| 232 |
-
"learning_rate": 9.464925117963133e-06,
|
| 233 |
-
"loss": 0.48378310203552244,
|
| 234 |
-
"step": 130,
|
| 235 |
-
"token_acc": 0.8495320160451642
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.20055710306406685,
|
| 239 |
-
"grad_norm": 3.15625,
|
| 240 |
-
"learning_rate": 9.408889720752265e-06,
|
| 241 |
-
"loss": 0.47188844680786135,
|
| 242 |
-
"step": 135,
|
| 243 |
-
"token_acc": 0.8658623448064258
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.20798514391829154,
|
| 247 |
-
"grad_norm": 1.6640625,
|
| 248 |
-
"learning_rate": 9.35024767453647e-06,
|
| 249 |
-
"loss": 0.39438793659210203,
|
| 250 |
-
"step": 140,
|
| 251 |
-
"token_acc": 0.8809896055411847
|
| 252 |
-
},
|
| 253 |
-
{
|
| 254 |
-
"epoch": 0.21541318477251625,
|
| 255 |
-
"grad_norm": 2.453125,
|
| 256 |
-
"learning_rate": 9.289033650001817e-06,
|
| 257 |
-
"loss": 0.3956155300140381,
|
| 258 |
-
"step": 145,
|
| 259 |
-
"token_acc": 0.8743476320741905
|
| 260 |
-
},
|
| 261 |
-
{
|
| 262 |
-
"epoch": 0.22284122562674094,
|
| 263 |
-
"grad_norm": 3.46875,
|
| 264 |
-
"learning_rate": 9.225283838454111e-06,
|
| 265 |
-
"loss": 0.4306947708129883,
|
| 266 |
-
"step": 150,
|
| 267 |
-
"token_acc": 0.866333055323881
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.23026926648096566,
|
| 271 |
-
"grad_norm": 2.90625,
|
| 272 |
-
"learning_rate": 9.159035930421658e-06,
|
| 273 |
-
"loss": 0.4335611820220947,
|
| 274 |
-
"step": 155,
|
| 275 |
-
"token_acc": 0.8591517915650966
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.23769730733519034,
|
| 279 |
-
"grad_norm": 2.921875,
|
| 280 |
-
"learning_rate": 9.090329093371667e-06,
|
| 281 |
-
"loss": 0.4459502696990967,
|
| 282 |
-
"step": 160,
|
| 283 |
-
"token_acc": 0.8661823703840511
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.24512534818941503,
|
| 287 |
-
"grad_norm": 1.234375,
|
| 288 |
-
"learning_rate": 9.019203948553422e-06,
|
| 289 |
-
"loss": 0.4488051414489746,
|
| 290 |
-
"step": 165,
|
| 291 |
-
"token_acc": 0.8817028313029392
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.2525533890436397,
|
| 295 |
-
"grad_norm": 3.078125,
|
| 296 |
-
"learning_rate": 8.94570254698197e-06,
|
| 297 |
-
"loss": 0.42046666145324707,
|
| 298 |
-
"step": 170,
|
| 299 |
-
"token_acc": 0.8783048661518874
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 0.25998142989786444,
|
| 303 |
-
"grad_norm": 2.265625,
|
| 304 |
-
"learning_rate": 8.86986834457646e-06,
|
| 305 |
-
"loss": 0.416002082824707,
|
| 306 |
-
"step": 175,
|
| 307 |
-
"token_acc": 0.8755390772858428
|
| 308 |
-
},
|
| 309 |
-
{
|
| 310 |
-
"epoch": 0.26740947075208915,
|
| 311 |
-
"grad_norm": 3.140625,
|
| 312 |
-
"learning_rate": 8.791746176467908e-06,
|
| 313 |
-
"loss": 0.43740019798278806,
|
| 314 |
-
"step": 180,
|
| 315 |
-
"token_acc": 0.8627690538646735
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.2748375116063138,
|
| 319 |
-
"grad_norm": 2.96875,
|
| 320 |
-
"learning_rate": 8.711382230491494e-06,
|
| 321 |
-
"loss": 0.47383904457092285,
|
| 322 |
-
"step": 185,
|
| 323 |
-
"token_acc": 0.8521471463894635
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 0.2822655524605385,
|
| 327 |
-
"grad_norm": 3.53125,
|
| 328 |
-
"learning_rate": 8.628824019879137e-06,
|
| 329 |
-
"loss": 0.4318506717681885,
|
| 330 |
-
"step": 190,
|
| 331 |
-
"token_acc": 0.8731298295610237
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.28969359331476324,
|
| 335 |
-
"grad_norm": 1.8984375,
|
| 336 |
-
"learning_rate": 8.544120355168451e-06,
|
| 337 |
-
"loss": 0.47009100914001467,
|
| 338 |
-
"step": 195,
|
| 339 |
-
"token_acc": 0.8608060024752475
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.2971216341689879,
|
| 343 |
-
"grad_norm": 3.625,
|
| 344 |
-
"learning_rate": 8.457321315344695e-06,
|
| 345 |
-
"loss": 0.4356997013092041,
|
| 346 |
-
"step": 200,
|
| 347 |
-
"token_acc": 0.8650625469956935
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 0.2971216341689879,
|
| 351 |
-
"eval_loss": 0.3804284334182739,
|
| 352 |
-
"eval_runtime": 36.3904,
|
| 353 |
-
"eval_samples_per_second": 0.879,
|
| 354 |
-
"eval_steps_per_second": 0.22,
|
| 355 |
-
"eval_token_acc": 0.8699384970809528,
|
| 356 |
-
"step": 200
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 0.3045496750232126,
|
| 360 |
-
"grad_norm": 1.578125,
|
| 361 |
-
"learning_rate": 8.368478218232787e-06,
|
| 362 |
-
"loss": 0.42498369216918946,
|
| 363 |
-
"step": 205,
|
| 364 |
-
"token_acc": 0.8724639910770947
|
| 365 |
-
},
|
| 366 |
-
{
|
| 367 |
-
"epoch": 0.31197771587743733,
|
| 368 |
-
"grad_norm": 4.375,
|
| 369 |
-
"learning_rate": 8.277643590156893e-06,
|
| 370 |
-
"loss": 0.5058988571166992,
|
| 371 |
-
"step": 210,
|
| 372 |
-
"token_acc": 0.8499143154415815
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 0.31940575673166205,
|
| 376 |
-
"grad_norm": 2.875,
|
| 377 |
-
"learning_rate": 8.184871134885512e-06,
|
| 378 |
-
"loss": 0.48337273597717284,
|
| 379 |
-
"step": 215,
|
| 380 |
-
"token_acc": 0.8439464262034213
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.3268337975858867,
|
| 384 |
-
"grad_norm": 3.3125,
|
| 385 |
-
"learning_rate": 8.090215701880418e-06,
|
| 386 |
-
"loss": 0.39829819202423095,
|
| 387 |
-
"step": 220,
|
| 388 |
-
"token_acc": 0.8789773221377292
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.3342618384401114,
|
| 392 |
-
"grad_norm": 1.796875,
|
| 393 |
-
"learning_rate": 7.993733253868256e-06,
|
| 394 |
-
"loss": 0.4247430801391602,
|
| 395 |
-
"step": 225,
|
| 396 |
-
"token_acc": 0.8725433118097456
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 0.34168987929433614,
|
| 400 |
-
"grad_norm": 3.109375,
|
| 401 |
-
"learning_rate": 7.895480833753942e-06,
|
| 402 |
-
"loss": 0.44652495384216306,
|
| 403 |
-
"step": 230,
|
| 404 |
-
"token_acc": 0.8550258329981747
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.3491179201485608,
|
| 408 |
-
"grad_norm": 2.703125,
|
| 409 |
-
"learning_rate": 7.795516530895414e-06,
|
| 410 |
-
"loss": 0.4880677223205566,
|
| 411 |
-
"step": 235,
|
| 412 |
-
"token_acc": 0.8474095464582979
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 0.3565459610027855,
|
| 416 |
-
"grad_norm": 2.5625,
|
| 417 |
-
"learning_rate": 7.693899446759727e-06,
|
| 418 |
-
"loss": 0.4174833297729492,
|
| 419 |
-
"step": 240,
|
| 420 |
-
"token_acc": 0.8676238451181577
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 0.36397400185701023,
|
| 424 |
-
"grad_norm": 2.546875,
|
| 425 |
-
"learning_rate": 7.59068965998074e-06,
|
| 426 |
-
"loss": 0.44954814910888674,
|
| 427 |
-
"step": 245,
|
| 428 |
-
"token_acc": 0.8637673169603955
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 0.3714020427112349,
|
| 432 |
-
"grad_norm": 2.546875,
|
| 433 |
-
"learning_rate": 7.485948190839076e-06,
|
| 434 |
-
"loss": 0.4236954689025879,
|
| 435 |
-
"step": 250,
|
| 436 |
-
"token_acc": 0.873028278419001
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.3788300835654596,
|
| 440 |
-
"grad_norm": 3.21875,
|
| 441 |
-
"learning_rate": 7.379736965185369e-06,
|
| 442 |
-
"loss": 0.3399410009384155,
|
| 443 |
-
"step": 255,
|
| 444 |
-
"token_acc": 0.9000444686002272
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.3862581244196843,
|
| 448 |
-
"grad_norm": 2.46875,
|
| 449 |
-
"learning_rate": 7.272118777828109e-06,
|
| 450 |
-
"loss": 0.4131065845489502,
|
| 451 |
-
"step": 260,
|
| 452 |
-
"token_acc": 0.8704600267976775
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 0.393686165273909,
|
| 456 |
-
"grad_norm": 2.203125,
|
| 457 |
-
"learning_rate": 7.163157255407732e-06,
|
| 458 |
-
"loss": 0.4625247001647949,
|
| 459 |
-
"step": 265,
|
| 460 |
-
"token_acc": 0.8550916665604974
|
| 461 |
-
},
|
| 462 |
-
{
|
| 463 |
-
"epoch": 0.4011142061281337,
|
| 464 |
-
"grad_norm": 2.796875,
|
| 465 |
-
"learning_rate": 7.052916818778918e-06,
|
| 466 |
-
"loss": 0.4972819805145264,
|
| 467 |
-
"step": 270,
|
| 468 |
-
"token_acc": 0.842036533929724
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.4085422469823584,
|
| 472 |
-
"grad_norm": 3.6875,
|
| 473 |
-
"learning_rate": 6.941462644923318e-06,
|
| 474 |
-
"loss": 0.4599851131439209,
|
| 475 |
-
"step": 275,
|
| 476 |
-
"token_acc": 0.8602353264983141
|
| 477 |
-
},
|
| 478 |
-
{
|
| 479 |
-
"epoch": 0.4159702878365831,
|
| 480 |
-
"grad_norm": 2.234375,
|
| 481 |
-
"learning_rate": 6.8288606284152535e-06,
|
| 482 |
-
"loss": 0.46378469467163086,
|
| 483 |
-
"step": 280,
|
| 484 |
-
"token_acc": 0.8549655796903809
|
| 485 |
-
},
|
| 486 |
-
{
|
| 487 |
-
"epoch": 0.4233983286908078,
|
| 488 |
-
"grad_norm": 2.015625,
|
| 489 |
-
"learning_rate": 6.715177342463145e-06,
|
| 490 |
-
"loss": 0.44658217430114744,
|
| 491 |
-
"step": 285,
|
| 492 |
-
"token_acc": 0.8630843381819439
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.4308263695450325,
|
| 496 |
-
"grad_norm": 2.15625,
|
| 497 |
-
"learning_rate": 6.600479999549721e-06,
|
| 498 |
-
"loss": 0.44566779136657714,
|
| 499 |
-
"step": 290,
|
| 500 |
-
"token_acc": 0.8626803196457605
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.4382544103992572,
|
| 504 |
-
"grad_norm": 1.703125,
|
| 505 |
-
"learning_rate": 6.4848364116942665e-06,
|
| 506 |
-
"loss": 0.40390496253967284,
|
| 507 |
-
"step": 295,
|
| 508 |
-
"token_acc": 0.876235031246727
|
| 509 |
-
},
|
| 510 |
-
{
|
| 511 |
-
"epoch": 0.4456824512534819,
|
| 512 |
-
"grad_norm": 2.5,
|
| 513 |
-
"learning_rate": 6.368314950360416e-06,
|
| 514 |
-
"loss": 0.45004024505615237,
|
| 515 |
-
"step": 300,
|
| 516 |
-
"token_acc": 0.8598025900696175
|
| 517 |
-
},
|
| 518 |
-
{
|
| 519 |
-
"epoch": 0.4456824512534819,
|
| 520 |
-
"eval_loss": 0.38784611225128174,
|
| 521 |
-
"eval_runtime": 36.3092,
|
| 522 |
-
"eval_samples_per_second": 0.881,
|
| 523 |
-
"eval_steps_per_second": 0.22,
|
| 524 |
-
"eval_token_acc": 0.8708750897567981,
|
| 525 |
-
"step": 300
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 0.4531104921077066,
|
| 529 |
-
"grad_norm": 4.6875,
|
| 530 |
-
"learning_rate": 6.250984506033183e-06,
|
| 531 |
-
"loss": 0.45397129058837893,
|
| 532 |
-
"step": 305,
|
| 533 |
-
"token_acc": 0.8667876331416722
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.4605385329619313,
|
| 537 |
-
"grad_norm": 2.84375,
|
| 538 |
-
"learning_rate": 6.132914447489137e-06,
|
| 539 |
-
"loss": 0.41869430541992186,
|
| 540 |
-
"step": 310,
|
| 541 |
-
"token_acc": 0.8714266197616038
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.467966573816156,
|
| 545 |
-
"grad_norm": 2.671875,
|
| 546 |
-
"learning_rate": 6.014174580783794e-06,
|
| 547 |
-
"loss": 0.4706721305847168,
|
| 548 |
-
"step": 315,
|
| 549 |
-
"token_acc": 0.8511786201596587
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.4753946146703807,
|
| 553 |
-
"grad_norm": 3.03125,
|
| 554 |
-
"learning_rate": 5.8948351079804875e-06,
|
| 555 |
-
"loss": 0.42859907150268556,
|
| 556 |
-
"step": 320,
|
| 557 |
-
"token_acc": 0.8752680294475019
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 0.4828226555246054,
|
| 561 |
-
"grad_norm": 2.375,
|
| 562 |
-
"learning_rate": 5.774966585645092e-06,
|
| 563 |
-
"loss": 0.4039738655090332,
|
| 564 |
-
"step": 325,
|
| 565 |
-
"token_acc": 0.87506986066411
|
| 566 |
-
},
|
| 567 |
-
{
|
| 568 |
-
"epoch": 0.49025069637883006,
|
| 569 |
-
"grad_norm": 1.328125,
|
| 570 |
-
"learning_rate": 5.6546398831311774e-06,
|
| 571 |
-
"loss": 0.4178034782409668,
|
| 572 |
-
"step": 330,
|
| 573 |
-
"token_acc": 0.873616409230192
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 0.4976787372330548,
|
| 577 |
-
"grad_norm": 2.84375,
|
| 578 |
-
"learning_rate": 5.533926140680222e-06,
|
| 579 |
-
"loss": 0.46120491027832033,
|
| 580 |
-
"step": 335,
|
| 581 |
-
"token_acc": 0.860412461418517
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 0.5051067780872794,
|
| 585 |
-
"grad_norm": 2.59375,
|
| 586 |
-
"learning_rate": 5.412896727361663e-06,
|
| 587 |
-
"loss": 0.5120056629180908,
|
| 588 |
-
"step": 340,
|
| 589 |
-
"token_acc": 0.8444269214713113
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 0.5125348189415042,
|
| 593 |
-
"grad_norm": 2.234375,
|
| 594 |
-
"learning_rate": 5.29162319887768e-06,
|
| 595 |
-
"loss": 0.4420137405395508,
|
| 596 |
-
"step": 345,
|
| 597 |
-
"token_acc": 0.8672762375989369
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.5199628597957289,
|
| 601 |
-
"grad_norm": 2.203125,
|
| 602 |
-
"learning_rate": 5.170177255257618e-06,
|
| 603 |
-
"loss": 0.4228532791137695,
|
| 604 |
-
"step": 350,
|
| 605 |
-
"token_acc": 0.8667074839732046
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.5273909006499535,
|
| 609 |
-
"grad_norm": 1.6171875,
|
| 610 |
-
"learning_rate": 5.048630698467081e-06,
|
| 611 |
-
"loss": 0.44324636459350586,
|
| 612 |
-
"step": 355,
|
| 613 |
-
"token_acc": 0.8755654749130865
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 0.5348189415041783,
|
| 617 |
-
"grad_norm": 2.140625,
|
| 618 |
-
"learning_rate": 4.927055389956768e-06,
|
| 619 |
-
"loss": 0.4113778591156006,
|
| 620 |
-
"step": 360,
|
| 621 |
-
"token_acc": 0.8763391701889071
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 0.542246982358403,
|
| 625 |
-
"grad_norm": 2.453125,
|
| 626 |
-
"learning_rate": 4.805523208176139e-06,
|
| 627 |
-
"loss": 0.4342958450317383,
|
| 628 |
-
"step": 365,
|
| 629 |
-
"token_acc": 0.8598026652998463
|
| 630 |
-
},
|
| 631 |
-
{
|
| 632 |
-
"epoch": 0.5496750232126276,
|
| 633 |
-
"grad_norm": 1.0625,
|
| 634 |
-
"learning_rate": 4.684106006077015e-06,
|
| 635 |
-
"loss": 0.3641706705093384,
|
| 636 |
-
"step": 370,
|
| 637 |
-
"token_acc": 0.8914295630351935
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 0.5571030640668524,
|
| 641 |
-
"grad_norm": 2.46875,
|
| 642 |
-
"learning_rate": 4.562875568632278e-06,
|
| 643 |
-
"loss": 0.47191247940063474,
|
| 644 |
-
"step": 375,
|
| 645 |
-
"token_acc": 0.8525909471614289
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 0.564531104921077,
|
| 649 |
-
"grad_norm": 1.59375,
|
| 650 |
-
"learning_rate": 4.441903570394739e-06,
|
| 651 |
-
"loss": 0.358817982673645,
|
| 652 |
-
"step": 380,
|
| 653 |
-
"token_acc": 0.8989526394250984
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.5719591457753017,
|
| 657 |
-
"grad_norm": 3.140625,
|
| 658 |
-
"learning_rate": 4.321261533121303e-06,
|
| 659 |
-
"loss": 0.4361182689666748,
|
| 660 |
-
"step": 385,
|
| 661 |
-
"token_acc": 0.8649994412228528
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.5793871866295265,
|
| 665 |
-
"grad_norm": 2.09375,
|
| 666 |
-
"learning_rate": 4.201020783487465e-06,
|
| 667 |
-
"loss": 0.4602985382080078,
|
| 668 |
-
"step": 390,
|
| 669 |
-
"token_acc": 0.8708352105004807
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.5868152274837511,
|
| 673 |
-
"grad_norm": 2.0625,
|
| 674 |
-
"learning_rate": 4.081252410917148e-06,
|
| 675 |
-
"loss": 0.4332606315612793,
|
| 676 |
-
"step": 395,
|
| 677 |
-
"token_acc": 0.8665395455419792
|
| 678 |
-
},
|
| 679 |
-
{
|
| 680 |
-
"epoch": 0.5942432683379758,
|
| 681 |
-
"grad_norm": 3.515625,
|
| 682 |
-
"learning_rate": 3.962027225552807e-06,
|
| 683 |
-
"loss": 0.4250647068023682,
|
| 684 |
-
"step": 400,
|
| 685 |
-
"token_acc": 0.8664141249210807
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 0.5942432683379758,
|
| 689 |
-
"eval_loss": 0.3794129192829132,
|
| 690 |
-
"eval_runtime": 36.313,
|
| 691 |
-
"eval_samples_per_second": 0.881,
|
| 692 |
-
"eval_steps_per_second": 0.22,
|
| 693 |
-
"eval_token_acc": 0.8717804626767819,
|
| 694 |
-
"step": 400
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 0.6016713091922006,
|
| 698 |
-
"grad_norm": 4.5,
|
| 699 |
-
"learning_rate": 3.843415716390644e-06,
|
| 700 |
-
"loss": 0.4179537296295166,
|
| 701 |
-
"step": 405,
|
| 702 |
-
"token_acc": 0.8738840357108573
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 0.6090993500464252,
|
| 706 |
-
"grad_norm": 2.453125,
|
| 707 |
-
"learning_rate": 3.725488009605708e-06,
|
| 708 |
-
"loss": 0.4385994911193848,
|
| 709 |
-
"step": 410,
|
| 710 |
-
"token_acc": 0.8660233083053112
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.61652739090065,
|
| 714 |
-
"grad_norm": 1.1875,
|
| 715 |
-
"learning_rate": 3.6083138270914932e-06,
|
| 716 |
-
"loss": 0.4708572864532471,
|
| 717 |
-
"step": 415,
|
| 718 |
-
"token_acc": 0.8595829113774993
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 0.6239554317548747,
|
| 722 |
-
"grad_norm": 3.78125,
|
| 723 |
-
"learning_rate": 3.491962445238569e-06,
|
| 724 |
-
"loss": 0.43685154914855956,
|
| 725 |
-
"step": 420,
|
| 726 |
-
"token_acc": 0.8617621845401133
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 0.6313834726090993,
|
| 730 |
-
"grad_norm": 2.8125,
|
| 731 |
-
"learning_rate": 3.3765026539765832e-06,
|
| 732 |
-
"loss": 0.4290899753570557,
|
| 733 |
-
"step": 425,
|
| 734 |
-
"token_acc": 0.8770416264170636
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 0.6388115134633241,
|
| 738 |
-
"grad_norm": 2.890625,
|
| 739 |
-
"learning_rate": 3.2620027161038975e-06,
|
| 740 |
-
"loss": 0.45714583396911623,
|
| 741 |
-
"step": 430,
|
| 742 |
-
"token_acc": 0.8581179872852261
|
| 743 |
-
},
|
| 744 |
-
{
|
| 745 |
-
"epoch": 0.6462395543175488,
|
| 746 |
-
"grad_norm": 2.109375,
|
| 747 |
-
"learning_rate": 3.1485303269288603e-06,
|
| 748 |
-
"loss": 0.40250434875488283,
|
| 749 |
-
"step": 435,
|
| 750 |
-
"token_acc": 0.8794570858283434
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 0.6536675951717734,
|
| 754 |
-
"grad_norm": 1.7421875,
|
| 755 |
-
"learning_rate": 3.0361525742465975e-06,
|
| 756 |
-
"loss": 0.3919744253158569,
|
| 757 |
-
"step": 440,
|
| 758 |
-
"token_acc": 0.8788066010893423
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.6610956360259982,
|
| 762 |
-
"grad_norm": 2.8125,
|
| 763 |
-
"learning_rate": 2.9249358986749922e-06,
|
| 764 |
-
"loss": 0.3606393814086914,
|
| 765 |
-
"step": 445,
|
| 766 |
-
"token_acc": 0.8975810968183555
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.6685236768802229,
|
| 770 |
-
"grad_norm": 2.546875,
|
| 771 |
-
"learning_rate": 2.8149460543732666e-06,
|
| 772 |
-
"loss": 0.4342175006866455,
|
| 773 |
-
"step": 450,
|
| 774 |
-
"token_acc": 0.8730138871512574
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 0.6759517177344475,
|
| 778 |
-
"grad_norm": 2.421875,
|
| 779 |
-
"learning_rate": 2.706248070166449e-06,
|
| 780 |
-
"loss": 0.518229341506958,
|
| 781 |
-
"step": 455,
|
| 782 |
-
"token_acc": 0.8271784214324843
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 0.6833797585886723,
|
| 786 |
-
"grad_norm": 1.921875,
|
| 787 |
-
"learning_rate": 2.598906211098643e-06,
|
| 788 |
-
"loss": 0.4507512092590332,
|
| 789 |
-
"step": 460,
|
| 790 |
-
"token_acc": 0.8645038167938931
|
| 791 |
-
},
|
| 792 |
-
{
|
| 793 |
-
"epoch": 0.6908077994428969,
|
| 794 |
-
"grad_norm": 2.828125,
|
| 795 |
-
"learning_rate": 2.4929839404378935e-06,
|
| 796 |
-
"loss": 0.40171589851379397,
|
| 797 |
-
"step": 465,
|
| 798 |
-
"token_acc": 0.8937497075710685
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 0.6982358402971216,
|
| 802 |
-
"grad_norm": 2.984375,
|
| 803 |
-
"learning_rate": 2.388543882155067e-06,
|
| 804 |
-
"loss": 0.5253053665161133,
|
| 805 |
-
"step": 470,
|
| 806 |
-
"token_acc": 0.8259218075004182
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 0.7056638811513464,
|
| 810 |
-
"grad_norm": 1.7421875,
|
| 811 |
-
"learning_rate": 2.2856477838989455e-06,
|
| 812 |
-
"loss": 0.39436964988708495,
|
| 813 |
-
"step": 475,
|
| 814 |
-
"token_acc": 0.8809013923641638
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 0.713091922005571,
|
| 818 |
-
"grad_norm": 3.1875,
|
| 819 |
-
"learning_rate": 2.1843564804894316e-06,
|
| 820 |
-
"loss": 0.4038860321044922,
|
| 821 |
-
"step": 480,
|
| 822 |
-
"token_acc": 0.8802953750485815
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.7205199628597957,
|
| 826 |
-
"grad_norm": 2.59375,
|
| 827 |
-
"learning_rate": 2.0847298579504344e-06,
|
| 828 |
-
"loss": 0.4557987689971924,
|
| 829 |
-
"step": 485,
|
| 830 |
-
"token_acc": 0.876829989220016
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 0.7279480037140205,
|
| 834 |
-
"grad_norm": 2.078125,
|
| 835 |
-
"learning_rate": 1.9868268181037186e-06,
|
| 836 |
-
"loss": 0.4029983043670654,
|
| 837 |
-
"step": 490,
|
| 838 |
-
"token_acc": 0.874598797522716
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 0.7353760445682451,
|
| 842 |
-
"grad_norm": 3.75,
|
| 843 |
-
"learning_rate": 1.8907052437446271e-06,
|
| 844 |
-
"loss": 0.4528185367584229,
|
| 845 |
-
"step": 495,
|
| 846 |
-
"token_acc": 0.8583082019083403
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"epoch": 0.7428040854224698,
|
| 850 |
-
"grad_norm": 1.7890625,
|
| 851 |
-
"learning_rate": 1.7964219644202852e-06,
|
| 852 |
-
"loss": 0.4013031005859375,
|
| 853 |
-
"step": 500,
|
| 854 |
-
"token_acc": 0.8737368382418909
|
| 855 |
-
},
|
| 856 |
-
{
|
| 857 |
-
"epoch": 0.7428040854224698,
|
| 858 |
-
"eval_loss": 0.37805721163749695,
|
| 859 |
-
"eval_runtime": 36.2847,
|
| 860 |
-
"eval_samples_per_second": 0.882,
|
| 861 |
-
"eval_steps_per_second": 0.22,
|
| 862 |
-
"eval_token_acc": 0.8719053417002279,
|
| 863 |
-
"step": 500
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.7502321262766946,
|
| 867 |
-
"grad_norm": 3.40625,
|
| 868 |
-
"learning_rate": 1.704032722830512e-06,
|
| 869 |
-
"loss": 0.4088125705718994,
|
| 870 |
-
"step": 505,
|
| 871 |
-
"token_acc": 0.8885762787520544
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.7576601671309192,
|
| 875 |
-
"grad_norm": 2.4375,
|
| 876 |
-
"learning_rate": 1.6135921418712959e-06,
|
| 877 |
-
"loss": 0.37178170680999756,
|
| 878 |
-
"step": 510,
|
| 879 |
-
"token_acc": 0.8881887594797235
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 0.7650882079851439,
|
| 883 |
-
"grad_norm": 2.921875,
|
| 884 |
-
"learning_rate": 1.5251536923403427e-06,
|
| 885 |
-
"loss": 0.47313766479492186,
|
| 886 |
-
"step": 515,
|
| 887 |
-
"token_acc": 0.8571553712733962
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 0.7725162488393686,
|
| 891 |
-
"grad_norm": 2.21875,
|
| 892 |
-
"learning_rate": 1.4387696613237612e-06,
|
| 893 |
-
"loss": 0.36481239795684817,
|
| 894 |
-
"step": 520,
|
| 895 |
-
"token_acc": 0.9051443329633418
|
| 896 |
-
},
|
| 897 |
-
{
|
| 898 |
-
"epoch": 0.7799442896935933,
|
| 899 |
-
"grad_norm": 2.328125,
|
| 900 |
-
"learning_rate": 1.3544911212825907e-06,
|
| 901 |
-
"loss": 0.3487920045852661,
|
| 902 |
-
"step": 525,
|
| 903 |
-
"token_acc": 0.8989453361840636
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 0.787372330547818,
|
| 907 |
-
"grad_norm": 2.109375,
|
| 908 |
-
"learning_rate": 1.2723678998574512e-06,
|
| 909 |
-
"loss": 0.39643423557281493,
|
| 910 |
-
"step": 530,
|
| 911 |
-
"token_acc": 0.8743610045584878
|
| 912 |
-
},
|
| 913 |
-
{
|
| 914 |
-
"epoch": 0.7948003714020427,
|
| 915 |
-
"grad_norm": 2.109375,
|
| 916 |
-
"learning_rate": 1.1924485504091565e-06,
|
| 917 |
-
"loss": 0.39751591682434084,
|
| 918 |
-
"step": 535,
|
| 919 |
-
"token_acc": 0.8818900045873898
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.8022284122562674,
|
| 923 |
-
"grad_norm": 2.703125,
|
| 924 |
-
"learning_rate": 1.1147803233127241e-06,
|
| 925 |
-
"loss": 0.4552616596221924,
|
| 926 |
-
"step": 540,
|
| 927 |
-
"token_acc": 0.8558503004753789
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.8096564531104921,
|
| 931 |
-
"grad_norm": 1.2265625,
|
| 932 |
-
"learning_rate": 1.0394091380217354e-06,
|
| 933 |
-
"loss": 0.3665018558502197,
|
| 934 |
-
"step": 545,
|
| 935 |
-
"token_acc": 0.9009014156940615
|
| 936 |
-
},
|
| 937 |
-
{
|
| 938 |
-
"epoch": 0.8170844939647168,
|
| 939 |
-
"grad_norm": 2.578125,
|
| 940 |
-
"learning_rate": 9.663795559195733e-07,
|
| 941 |
-
"loss": 0.4344160079956055,
|
| 942 |
-
"step": 550,
|
| 943 |
-
"token_acc": 0.8642133312957851
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 0.8245125348189415,
|
| 947 |
-
"grad_norm": 2.09375,
|
| 948 |
-
"learning_rate": 8.957347539735872e-07,
|
| 949 |
-
"loss": 0.37365174293518066,
|
| 950 |
-
"step": 555,
|
| 951 |
-
"token_acc": 0.8828788815078016
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 0.8319405756731661,
|
| 955 |
-
"grad_norm": 3.640625,
|
| 956 |
-
"learning_rate": 8.275164992077555e-07,
|
| 957 |
-
"loss": 0.45587959289550783,
|
| 958 |
-
"step": 560,
|
| 959 |
-
"token_acc": 0.8582994836488812
|
| 960 |
-
},
|
| 961 |
-
{
|
| 962 |
-
"epoch": 0.8393686165273909,
|
| 963 |
-
"grad_norm": 2.125,
|
| 964 |
-
"learning_rate": 7.617651240089546e-07,
|
| 965 |
-
"loss": 0.37633504867553713,
|
| 966 |
-
"step": 565,
|
| 967 |
-
"token_acc": 0.8808038949506497
|
| 968 |
-
},
|
| 969 |
-
{
|
| 970 |
-
"epoch": 0.8467966573816156,
|
| 971 |
-
"grad_norm": 2.984375,
|
| 972 |
-
"learning_rate": 6.985195022814068e-07,
|
| 973 |
-
"loss": 0.3819861888885498,
|
| 974 |
-
"step": 570,
|
| 975 |
-
"token_acc": 0.8838667447548834
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.8542246982358404,
|
| 979 |
-
"grad_norm": 2.78125,
|
| 980 |
-
"learning_rate": 6.37817026463432e-07,
|
| 981 |
-
"loss": 0.4230340480804443,
|
| 982 |
-
"step": 575,
|
| 983 |
-
"token_acc": 0.8686186847553187
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.861652739090065,
|
| 987 |
-
"grad_norm": 2.640625,
|
| 988 |
-
"learning_rate": 5.796935854200764e-07,
|
| 989 |
-
"loss": 0.420436954498291,
|
| 990 |
-
"step": 580,
|
| 991 |
-
"token_acc": 0.8676813085630721
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 0.8690807799442897,
|
| 995 |
-
"grad_norm": 2.703125,
|
| 996 |
-
"learning_rate": 5.241835432246888e-07,
|
| 997 |
-
"loss": 0.4056081771850586,
|
| 998 |
-
"step": 585,
|
| 999 |
-
"token_acc": 0.8762209344704425
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.8765088207985144,
|
| 1003 |
-
"grad_norm": 2.96875,
|
| 1004 |
-
"learning_rate": 4.713197188420027e-07,
|
| 1005 |
-
"loss": 0.4248391628265381,
|
| 1006 |
-
"step": 590,
|
| 1007 |
-
"token_acc": 0.891635065804149
|
| 1008 |
-
},
|
| 1009 |
-
{
|
| 1010 |
-
"epoch": 0.8839368616527391,
|
| 1011 |
-
"grad_norm": 1.421875,
|
| 1012 |
-
"learning_rate": 4.211333667247125e-07,
|
| 1013 |
-
"loss": 0.34991230964660647,
|
| 1014 |
-
"step": 595,
|
| 1015 |
-
"token_acc": 0.8906507536191592
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 0.8913649025069638,
|
| 1019 |
-
"grad_norm": 2.140625,
|
| 1020 |
-
"learning_rate": 3.736541583350473e-07,
|
| 1021 |
-
"loss": 0.4032041072845459,
|
| 1022 |
-
"step": 600,
|
| 1023 |
-
"token_acc": 0.8810850357363949
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 0.8913649025069638,
|
| 1027 |
-
"eval_loss": 0.37688374519348145,
|
| 1028 |
-
"eval_runtime": 36.2808,
|
| 1029 |
-
"eval_samples_per_second": 0.882,
|
| 1030 |
-
"eval_steps_per_second": 0.221,
|
| 1031 |
-
"eval_token_acc": 0.872217539258843,
|
| 1032 |
-
"step": 600
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.8987929433611885,
|
| 1036 |
-
"grad_norm": 2.8125,
|
| 1037 |
-
"learning_rate": 3.289101646022297e-07,
|
| 1038 |
-
"loss": 0.45639562606811523,
|
| 1039 |
-
"step": 605,
|
| 1040 |
-
"token_acc": 0.8531121445179753
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 0.9062209842154132,
|
| 1044 |
-
"grad_norm": 2.375,
|
| 1045 |
-
"learning_rate": 2.869278393262226e-07,
|
| 1046 |
-
"loss": 0.4369607448577881,
|
| 1047 |
-
"step": 610,
|
| 1048 |
-
"token_acc": 0.8665694654222579
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 0.9136490250696379,
|
| 1052 |
-
"grad_norm": 1.6640625,
|
| 1053 |
-
"learning_rate": 2.47732003537568e-07,
|
| 1054 |
-
"loss": 0.37709152698516846,
|
| 1055 |
-
"step": 615,
|
| 1056 |
-
"token_acc": 0.8860294117647058
|
| 1057 |
-
},
|
| 1058 |
-
{
|
| 1059 |
-
"epoch": 0.9210770659238626,
|
| 1060 |
-
"grad_norm": 3.84375,
|
| 1061 |
-
"learning_rate": 2.113458308225458e-07,
|
| 1062 |
-
"loss": 0.4648869514465332,
|
| 1063 |
-
"step": 620,
|
| 1064 |
-
"token_acc": 0.8534752590483293
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 0.9285051067780873,
|
| 1068 |
-
"grad_norm": 5.125,
|
| 1069 |
-
"learning_rate": 1.777908336223655e-07,
|
| 1070 |
-
"loss": 0.4395432949066162,
|
| 1071 |
-
"step": 625,
|
| 1072 |
-
"token_acc": 0.8737245182753924
|
| 1073 |
-
},
|
| 1074 |
-
{
|
| 1075 |
-
"epoch": 0.935933147632312,
|
| 1076 |
-
"grad_norm": 1.8203125,
|
| 1077 |
-
"learning_rate": 1.4708685051444515e-07,
|
| 1078 |
-
"loss": 0.3888163805007935,
|
| 1079 |
-
"step": 630,
|
| 1080 |
-
"token_acc": 0.880680708909087
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 0.9433611884865367,
|
| 1084 |
-
"grad_norm": 2.578125,
|
| 1085 |
-
"learning_rate": 1.19252034483342e-07,
|
| 1086 |
-
"loss": 0.462689208984375,
|
| 1087 |
-
"step": 635,
|
| 1088 |
-
"token_acc": 0.8559235668789809
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.9507892293407614,
|
| 1092 |
-
"grad_norm": 2.078125,
|
| 1093 |
-
"learning_rate": 9.430284218824026e-08,
|
| 1094 |
-
"loss": 0.42983808517456057,
|
| 1095 |
-
"step": 640,
|
| 1096 |
-
"token_acc": 0.862285726545236
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 0.958217270194986,
|
| 1100 |
-
"grad_norm": 1.2578125,
|
| 1101 |
-
"learning_rate": 7.225402423334694e-08,
|
| 1102 |
-
"loss": 0.40550642013549804,
|
| 1103 |
-
"step": 645,
|
| 1104 |
-
"token_acc": 0.8780291711611515
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 0.9656453110492108,
|
| 1108 |
-
"grad_norm": 3.609375,
|
| 1109 |
-
"learning_rate": 5.3118616446960484e-08,
|
| 1110 |
-
"loss": 0.4838839054107666,
|
| 1111 |
-
"step": 650,
|
| 1112 |
-
"token_acc": 0.8556079835832016
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 0.9730733519034355,
|
| 1116 |
-
"grad_norm": 1.9296875,
|
| 1117 |
-
"learning_rate": 3.690793217434985e-08,
|
| 1118 |
-
"loss": 0.40026402473449707,
|
| 1119 |
-
"step": 655,
|
| 1120 |
-
"token_acc": 0.87586234477794
|
| 1121 |
-
},
|
| 1122 |
-
{
|
| 1123 |
-
"epoch": 0.9805013927576601,
|
| 1124 |
-
"grad_norm": 2.765625,
|
| 1125 |
-
"learning_rate": 2.363155558901542e-08,
|
| 1126 |
-
"loss": 0.42069473266601565,
|
| 1127 |
-
"step": 660,
|
| 1128 |
-
"token_acc": 0.8785162827933342
|
| 1129 |
-
},
|
| 1130 |
-
{
|
| 1131 |
-
"epoch": 0.9879294336118849,
|
| 1132 |
-
"grad_norm": 2.34375,
|
| 1133 |
-
"learning_rate": 1.3297336026280027e-08,
|
| 1134 |
-
"loss": 0.43651704788208007,
|
| 1135 |
-
"step": 665,
|
| 1136 |
-
"token_acc": 0.8615758615758615
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.9953574744661096,
|
| 1140 |
-
"grad_norm": 1.9296875,
|
| 1141 |
-
"learning_rate": 5.911383342556143e-09,
|
| 1142 |
-
"loss": 0.43561439514160155,
|
| 1143 |
-
"step": 670,
|
| 1144 |
-
"token_acc": 0.8611667439497233
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 1.001485608170845,
|
| 1148 |
-
"grad_norm": 2.546875,
|
| 1149 |
-
"learning_rate": 1.4780643030476439e-09,
|
| 1150 |
-
"loss": 0.41878299713134765,
|
| 1151 |
-
"step": 675,
|
| 1152 |
-
"token_acc": 0.8758858169477992
|
| 1153 |
-
},
|
| 1154 |
-
{
|
| 1155 |
-
"epoch": 1.0089136490250696,
|
| 1156 |
-
"grad_norm": 2.75,
|
| 1157 |
-
"learning_rate": 0.0,
|
| 1158 |
-
"loss": 0.3725011348724365,
|
| 1159 |
-
"step": 680,
|
| 1160 |
-
"token_acc": 0.8969813128893148
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 1.0089136490250696,
|
| 1164 |
-
"eval_loss": 0.37545016407966614,
|
| 1165 |
-
"eval_runtime": 36.3443,
|
| 1166 |
-
"eval_samples_per_second": 0.88,
|
| 1167 |
-
"eval_steps_per_second": 0.22,
|
| 1168 |
-
"eval_token_acc": 0.8720614404795355,
|
| 1169 |
-
"step": 680
|
| 1170 |
-
}
|
| 1171 |
-
],
|
| 1172 |
-
"logging_steps": 5,
|
| 1173 |
-
"max_steps": 680,
|
| 1174 |
-
"num_input_tokens_seen": 0,
|
| 1175 |
-
"num_train_epochs": 2,
|
| 1176 |
-
"save_steps": 100,
|
| 1177 |
-
"stateful_callbacks": {
|
| 1178 |
-
"TrainerControl": {
|
| 1179 |
-
"args": {
|
| 1180 |
-
"should_epoch_stop": false,
|
| 1181 |
-
"should_evaluate": false,
|
| 1182 |
-
"should_log": false,
|
| 1183 |
-
"should_save": true,
|
| 1184 |
-
"should_training_stop": true
|
| 1185 |
-
},
|
| 1186 |
-
"attributes": {}
|
| 1187 |
-
}
|
| 1188 |
-
},
|
| 1189 |
-
"total_flos": 8.09570838451218e+19,
|
| 1190 |
-
"train_batch_size": 1,
|
| 1191 |
-
"trial_name": null,
|
| 1192 |
-
"trial_params": null
|
| 1193 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
base-longtrace-step680/README.md
CHANGED
|
@@ -45,7 +45,7 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 45 |
|
| 46 |
### Train loss vs eval loss
|
| 47 |
|
| 48 |
-
|
| 49 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 50 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 51 |
interchangeable, so both are stated here for Base-LongTrace:
|
|
@@ -56,4 +56,5 @@ interchangeable, so both are stated here for Base-LongTrace:
|
|
| 56 |
| Train loss, final step 680 | 0.0040 |
|
| 57 |
| Eval loss, step 680 | 0.17741 |
|
| 58 |
|
| 59 |
-
This repository's table reports eval loss only; the train-loss figures above are provided for completeness and are not the reported metric.
|
|
|
|
|
|
| 45 |
|
| 46 |
### Train loss vs eval loss
|
| 47 |
|
| 48 |
+
The preserved numeric training metrics record per-step train loss
|
| 49 |
(`log_history[].loss`, 137 entries) and periodic eval loss
|
| 50 |
(`log_history[].eval_loss`, 7 entries: steps 100-680). The two scales are not
|
| 51 |
interchangeable, so both are stated here for Base-LongTrace:
|
|
|
|
| 56 |
| Train loss, final step 680 | 0.0040 |
|
| 57 |
| Eval loss, step 680 | 0.17741 |
|
| 58 |
|
| 59 |
+
This repository's table reports eval loss only; the train-loss figures above are provided for completeness and are not the reported metric.
|
| 60 |
+
The numeric metric history is preserved in [private research records](https://github.com/Xnhyacinth/longworld/blob/worlds/docs/research/baseline-training-metrics-20261030.json). The current model folder does not include the trainer-state file.
|
base-longtrace-step680/trainer_state.json
DELETED
|
@@ -1,1193 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": 200,
|
| 3 |
-
"best_metric": 0.12950581,
|
| 4 |
-
"best_model_checkpoint": "/workspace/<user>/longworld/data/sft/swift_ext_longtrace_base/v0-20260904-062722/checkpoint-200",
|
| 5 |
-
"epoch": 3.9083722601509163,
|
| 6 |
-
"eval_steps": 100,
|
| 7 |
-
"global_step": 680,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.005749191519942508,
|
| 14 |
-
"grad_norm": 246.0,
|
| 15 |
-
"learning_rate": 2.9411764705882356e-07,
|
| 16 |
-
"loss": 1.265736699104309,
|
| 17 |
-
"step": 1,
|
| 18 |
-
"token_acc": 0.7311827956989247
|
| 19 |
-
},
|
| 20 |
-
{
|
| 21 |
-
"epoch": 0.02874595759971254,
|
| 22 |
-
"grad_norm": 221.0,
|
| 23 |
-
"learning_rate": 1.4705882352941177e-06,
|
| 24 |
-
"loss": 1.229498028755188,
|
| 25 |
-
"step": 5,
|
| 26 |
-
"token_acc": 0.7360406091370558
|
| 27 |
-
},
|
| 28 |
-
{
|
| 29 |
-
"epoch": 0.05749191519942508,
|
| 30 |
-
"grad_norm": 34.75,
|
| 31 |
-
"learning_rate": 2.9411764705882355e-06,
|
| 32 |
-
"loss": 0.7008364200592041,
|
| 33 |
-
"step": 10,
|
| 34 |
-
"token_acc": 0.8400770712909441
|
| 35 |
-
},
|
| 36 |
-
{
|
| 37 |
-
"epoch": 0.08623787279913762,
|
| 38 |
-
"grad_norm": 38.25,
|
| 39 |
-
"learning_rate": 4.411764705882353e-06,
|
| 40 |
-
"loss": 0.3976879596710205,
|
| 41 |
-
"step": 15,
|
| 42 |
-
"token_acc": 0.9001883239171374
|
| 43 |
-
},
|
| 44 |
-
{
|
| 45 |
-
"epoch": 0.11498383039885016,
|
| 46 |
-
"grad_norm": 26.125,
|
| 47 |
-
"learning_rate": 5.882352941176471e-06,
|
| 48 |
-
"loss": 0.2368929147720337,
|
| 49 |
-
"step": 20,
|
| 50 |
-
"token_acc": 0.9281314168377823
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"epoch": 0.1437297879985627,
|
| 54 |
-
"grad_norm": 42.25,
|
| 55 |
-
"learning_rate": 7.352941176470589e-06,
|
| 56 |
-
"loss": 0.18809716701507567,
|
| 57 |
-
"step": 25,
|
| 58 |
-
"token_acc": 0.946
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.17247574559827525,
|
| 62 |
-
"grad_norm": 17.0,
|
| 63 |
-
"learning_rate": 8.823529411764707e-06,
|
| 64 |
-
"loss": 0.16786643266677856,
|
| 65 |
-
"step": 30,
|
| 66 |
-
"token_acc": 0.9397810218978102
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.2012217031979878,
|
| 70 |
-
"grad_norm": 19.25,
|
| 71 |
-
"learning_rate": 9.999940874631278e-06,
|
| 72 |
-
"loss": 0.12734075784683227,
|
| 73 |
-
"step": 35,
|
| 74 |
-
"token_acc": 0.9614678899082569
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 0.2299676607977003,
|
| 78 |
-
"grad_norm": 23.625,
|
| 79 |
-
"learning_rate": 9.997871633546257e-06,
|
| 80 |
-
"loss": 0.1446619749069214,
|
| 81 |
-
"step": 40,
|
| 82 |
-
"token_acc": 0.9587073608617595
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 0.25871361839741286,
|
| 86 |
-
"grad_norm": 14.875,
|
| 87 |
-
"learning_rate": 9.992847522200132e-06,
|
| 88 |
-
"loss": 0.19110403060913086,
|
| 89 |
-
"step": 45,
|
| 90 |
-
"token_acc": 0.952212389380531
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.2874595759971254,
|
| 94 |
-
"grad_norm": 22.25,
|
| 95 |
-
"learning_rate": 9.98487151097676e-06,
|
| 96 |
-
"loss": 0.22810838222503663,
|
| 97 |
-
"step": 50,
|
| 98 |
-
"token_acc": 0.934
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 0.31620553359683795,
|
| 102 |
-
"grad_norm": 10.375,
|
| 103 |
-
"learning_rate": 9.973948315499126e-06,
|
| 104 |
-
"loss": 0.08706344366073608,
|
| 105 |
-
"step": 55,
|
| 106 |
-
"token_acc": 0.9744058500914077
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 0.3449514911965505,
|
| 110 |
-
"grad_norm": 26.25,
|
| 111 |
-
"learning_rate": 9.960084393841355e-06,
|
| 112 |
-
"loss": 0.08517480492591858,
|
| 113 |
-
"step": 60,
|
| 114 |
-
"token_acc": 0.9685185185185186
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.37369744879626304,
|
| 118 |
-
"grad_norm": 23.75,
|
| 119 |
-
"learning_rate": 9.943287942710527e-06,
|
| 120 |
-
"loss": 0.055770975351333615,
|
| 121 |
-
"step": 65,
|
| 122 |
-
"token_acc": 0.9758364312267658
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.4024434063959756,
|
| 126 |
-
"grad_norm": 53.0,
|
| 127 |
-
"learning_rate": 9.923568892600579e-06,
|
| 128 |
-
"loss": 0.1589362144470215,
|
| 129 |
-
"step": 70,
|
| 130 |
-
"token_acc": 0.941398865784499
|
| 131 |
-
},
|
| 132 |
-
{
|
| 133 |
-
"epoch": 0.4311893639956881,
|
| 134 |
-
"grad_norm": 16.75,
|
| 135 |
-
"learning_rate": 9.90093890192113e-06,
|
| 136 |
-
"loss": 0.11662691831588745,
|
| 137 |
-
"step": 75,
|
| 138 |
-
"token_acc": 0.9691119691119691
|
| 139 |
-
},
|
| 140 |
-
{
|
| 141 |
-
"epoch": 0.4599353215954006,
|
| 142 |
-
"grad_norm": 16.625,
|
| 143 |
-
"learning_rate": 9.875411350104745e-06,
|
| 144 |
-
"loss": 0.06587117314338684,
|
| 145 |
-
"step": 80,
|
| 146 |
-
"token_acc": 0.9788867562380038
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.48868127919511317,
|
| 150 |
-
"grad_norm": 15.25,
|
| 151 |
-
"learning_rate": 9.847001329696653e-06,
|
| 152 |
-
"loss": 0.1521129846572876,
|
| 153 |
-
"step": 85,
|
| 154 |
-
"token_acc": 0.9492481203007519
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 0.5174272367948257,
|
| 158 |
-
"grad_norm": 20.875,
|
| 159 |
-
"learning_rate": 9.815725637431663e-06,
|
| 160 |
-
"loss": 0.19719940423965454,
|
| 161 |
-
"step": 90,
|
| 162 |
-
"token_acc": 0.9478584729981379
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"epoch": 0.5461731943945383,
|
| 166 |
-
"grad_norm": 18.0,
|
| 167 |
-
"learning_rate": 9.781602764303488e-06,
|
| 168 |
-
"loss": 0.11014801263809204,
|
| 169 |
-
"step": 95,
|
| 170 |
-
"token_acc": 0.9713261648745519
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.5749191519942508,
|
| 174 |
-
"grad_norm": 7.84375,
|
| 175 |
-
"learning_rate": 9.744652884632406e-06,
|
| 176 |
-
"loss": 0.12801656723022461,
|
| 177 |
-
"step": 100,
|
| 178 |
-
"token_acc": 0.9639278557114228
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.5749191519942508,
|
| 182 |
-
"eval_loss": 0.12975458800792694,
|
| 183 |
-
"eval_runtime": 52.8234,
|
| 184 |
-
"eval_samples_per_second": 0.606,
|
| 185 |
-
"eval_steps_per_second": 0.151,
|
| 186 |
-
"eval_token_acc": 0.973568281938326,
|
| 187 |
-
"step": 100
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 0.6036651095939634,
|
| 191 |
-
"grad_norm": 14.9375,
|
| 192 |
-
"learning_rate": 9.704897844137674e-06,
|
| 193 |
-
"loss": 0.15032684803009033,
|
| 194 |
-
"step": 105,
|
| 195 |
-
"token_acc": 0.9570093457943926
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.6324110671936759,
|
| 199 |
-
"grad_norm": 15.75,
|
| 200 |
-
"learning_rate": 9.66236114702178e-06,
|
| 201 |
-
"loss": 0.11219625473022461,
|
| 202 |
-
"step": 110,
|
| 203 |
-
"token_acc": 0.9590316573556797
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.6611570247933884,
|
| 207 |
-
"grad_norm": 8.125,
|
| 208 |
-
"learning_rate": 9.617067942074155e-06,
|
| 209 |
-
"loss": 0.11560609340667724,
|
| 210 |
-
"step": 115,
|
| 211 |
-
"token_acc": 0.9640831758034026
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.689902982393101,
|
| 215 |
-
"grad_norm": 14.625,
|
| 216 |
-
"learning_rate": 9.569045007802558e-06,
|
| 217 |
-
"loss": 0.07864299416542053,
|
| 218 |
-
"step": 120,
|
| 219 |
-
"token_acc": 0.9737302977232924
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.7186489399928135,
|
| 223 |
-
"grad_norm": 16.25,
|
| 224 |
-
"learning_rate": 9.518320736600943e-06,
|
| 225 |
-
"loss": 0.09496331810951233,
|
| 226 |
-
"step": 125,
|
| 227 |
-
"token_acc": 0.9672131147540983
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.7473948975925261,
|
| 231 |
-
"grad_norm": 51.75,
|
| 232 |
-
"learning_rate": 9.464925117963133e-06,
|
| 233 |
-
"loss": 0.09346672892570496,
|
| 234 |
-
"step": 130,
|
| 235 |
-
"token_acc": 0.974559686888454
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.7761408551922386,
|
| 239 |
-
"grad_norm": 8.4375,
|
| 240 |
-
"learning_rate": 9.408889720752265e-06,
|
| 241 |
-
"loss": 0.10687379837036133,
|
| 242 |
-
"step": 135,
|
| 243 |
-
"token_acc": 0.96484375
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.8048868127919512,
|
| 247 |
-
"grad_norm": 22.0,
|
| 248 |
-
"learning_rate": 9.35024767453647e-06,
|
| 249 |
-
"loss": 0.1338568687438965,
|
| 250 |
-
"step": 140,
|
| 251 |
-
"token_acc": 0.9604743083003953
|
| 252 |
-
},
|
| 253 |
-
{
|
| 254 |
-
"epoch": 0.8336327703916637,
|
| 255 |
-
"grad_norm": 36.75,
|
| 256 |
-
"learning_rate": 9.289033650001817e-06,
|
| 257 |
-
"loss": 0.1174044132232666,
|
| 258 |
-
"step": 145,
|
| 259 |
-
"token_acc": 0.9683168316831683
|
| 260 |
-
},
|
| 261 |
-
{
|
| 262 |
-
"epoch": 0.8623787279913762,
|
| 263 |
-
"grad_norm": 15.25,
|
| 264 |
-
"learning_rate": 9.225283838454111e-06,
|
| 265 |
-
"loss": 0.09538493156433106,
|
| 266 |
-
"step": 150,
|
| 267 |
-
"token_acc": 0.9715447154471545
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.8911246855910887,
|
| 271 |
-
"grad_norm": 15.5,
|
| 272 |
-
"learning_rate": 9.159035930421658e-06,
|
| 273 |
-
"loss": 0.057299858331680296,
|
| 274 |
-
"step": 155,
|
| 275 |
-
"token_acc": 0.9814814814814815
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.9198706431908013,
|
| 279 |
-
"grad_norm": 22.375,
|
| 280 |
-
"learning_rate": 9.090329093371667e-06,
|
| 281 |
-
"loss": 0.1181220293045044,
|
| 282 |
-
"step": 160,
|
| 283 |
-
"token_acc": 0.9641509433962264
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.9486166007905138,
|
| 287 |
-
"grad_norm": 6.09375,
|
| 288 |
-
"learning_rate": 9.019203948553422e-06,
|
| 289 |
-
"loss": 0.11957811117172241,
|
| 290 |
-
"step": 165,
|
| 291 |
-
"token_acc": 0.9625246548323472
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.9773625583902263,
|
| 295 |
-
"grad_norm": 2.3125,
|
| 296 |
-
"learning_rate": 8.94570254698197e-06,
|
| 297 |
-
"loss": 0.10146663188934327,
|
| 298 |
-
"step": 170,
|
| 299 |
-
"token_acc": 0.9759036144578314
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 1.0057491915199426,
|
| 303 |
-
"grad_norm": 33.75,
|
| 304 |
-
"learning_rate": 8.86986834457646e-06,
|
| 305 |
-
"loss": 0.11313971281051635,
|
| 306 |
-
"step": 175,
|
| 307 |
-
"token_acc": 0.9730769230769231
|
| 308 |
-
},
|
| 309 |
-
{
|
| 310 |
-
"epoch": 1.034495149119655,
|
| 311 |
-
"grad_norm": 2.765625,
|
| 312 |
-
"learning_rate": 8.791746176467908e-06,
|
| 313 |
-
"loss": 0.05893188714981079,
|
| 314 |
-
"step": 180,
|
| 315 |
-
"token_acc": 0.9839679358717435
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 1.0632411067193677,
|
| 319 |
-
"grad_norm": 18.5,
|
| 320 |
-
"learning_rate": 8.711382230491494e-06,
|
| 321 |
-
"loss": 0.09759573936462403,
|
| 322 |
-
"step": 185,
|
| 323 |
-
"token_acc": 0.9724409448818898
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 1.09198706431908,
|
| 327 |
-
"grad_norm": 12.0,
|
| 328 |
-
"learning_rate": 8.628824019879137e-06,
|
| 329 |
-
"loss": 0.050233447551727296,
|
| 330 |
-
"step": 190,
|
| 331 |
-
"token_acc": 0.9849906191369606
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 1.1207330219187928,
|
| 335 |
-
"grad_norm": 3.953125,
|
| 336 |
-
"learning_rate": 8.544120355168451e-06,
|
| 337 |
-
"loss": 0.028193503618240356,
|
| 338 |
-
"step": 195,
|
| 339 |
-
"token_acc": 0.9899193548387096
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.1494789795185052,
|
| 343 |
-
"grad_norm": 34.0,
|
| 344 |
-
"learning_rate": 8.457321315344695e-06,
|
| 345 |
-
"loss": 0.017485323548316955,
|
| 346 |
-
"step": 200,
|
| 347 |
-
"token_acc": 0.9959514170040485
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 1.1494789795185052,
|
| 351 |
-
"eval_loss": 0.12950581312179565,
|
| 352 |
-
"eval_runtime": 53.2119,
|
| 353 |
-
"eval_samples_per_second": 0.601,
|
| 354 |
-
"eval_steps_per_second": 0.15,
|
| 355 |
-
"eval_token_acc": 0.973568281938326,
|
| 356 |
-
"step": 200
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 1.1782249371182179,
|
| 360 |
-
"grad_norm": 4.09375,
|
| 361 |
-
"learning_rate": 8.368478218232787e-06,
|
| 362 |
-
"loss": 0.01435326635837555,
|
| 363 |
-
"step": 205,
|
| 364 |
-
"token_acc": 0.9910554561717353
|
| 365 |
-
},
|
| 366 |
-
{
|
| 367 |
-
"epoch": 1.2069708947179303,
|
| 368 |
-
"grad_norm": 3.390625,
|
| 369 |
-
"learning_rate": 8.277643590156893e-06,
|
| 370 |
-
"loss": 0.009520187228918075,
|
| 371 |
-
"step": 210,
|
| 372 |
-
"token_acc": 0.9963702359346642
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 1.2357168523176427,
|
| 376 |
-
"grad_norm": 6.3125,
|
| 377 |
-
"learning_rate": 8.184871134885512e-06,
|
| 378 |
-
"loss": 0.03157503604888916,
|
| 379 |
-
"step": 215,
|
| 380 |
-
"token_acc": 0.9908424908424909
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 1.2644628099173554,
|
| 384 |
-
"grad_norm": 9.75,
|
| 385 |
-
"learning_rate": 8.090215701880418e-06,
|
| 386 |
-
"loss": 0.042767885327339175,
|
| 387 |
-
"step": 220,
|
| 388 |
-
"token_acc": 0.9895104895104895
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 1.293208767517068,
|
| 392 |
-
"grad_norm": 0.34375,
|
| 393 |
-
"learning_rate": 7.993733253868256e-06,
|
| 394 |
-
"loss": 0.029948288202285768,
|
| 395 |
-
"step": 225,
|
| 396 |
-
"token_acc": 0.9899598393574297
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 1.3219547251167805,
|
| 400 |
-
"grad_norm": 0.0986328125,
|
| 401 |
-
"learning_rate": 7.895480833753942e-06,
|
| 402 |
-
"loss": 0.00912516936659813,
|
| 403 |
-
"step": 230,
|
| 404 |
-
"token_acc": 0.9963302752293578
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 1.350700682716493,
|
| 408 |
-
"grad_norm": 3.59375,
|
| 409 |
-
"learning_rate": 7.795516530895414e-06,
|
| 410 |
-
"loss": 0.003934210538864136,
|
| 411 |
-
"step": 235,
|
| 412 |
-
"token_acc": 1.0
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 1.3794466403162056,
|
| 416 |
-
"grad_norm": 25.875,
|
| 417 |
-
"learning_rate": 7.693899446759727e-06,
|
| 418 |
-
"loss": 0.006348265707492829,
|
| 419 |
-
"step": 240,
|
| 420 |
-
"token_acc": 0.9962546816479401
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 1.408192597915918,
|
| 424 |
-
"grad_norm": 24.75,
|
| 425 |
-
"learning_rate": 7.59068965998074e-06,
|
| 426 |
-
"loss": 0.03542393147945404,
|
| 427 |
-
"step": 245,
|
| 428 |
-
"token_acc": 0.9904397705544933
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 1.4369385555156307,
|
| 432 |
-
"grad_norm": 35.75,
|
| 433 |
-
"learning_rate": 7.485948190839076e-06,
|
| 434 |
-
"loss": 0.024394223093986513,
|
| 435 |
-
"step": 250,
|
| 436 |
-
"token_acc": 0.9961315280464217
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 1.465684513115343,
|
| 440 |
-
"grad_norm": 5.375,
|
| 441 |
-
"learning_rate": 7.379736965185369e-06,
|
| 442 |
-
"loss": 0.006641557812690735,
|
| 443 |
-
"step": 255,
|
| 444 |
-
"token_acc": 0.9962335216572504
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.4944304707150557,
|
| 448 |
-
"grad_norm": 0.73828125,
|
| 449 |
-
"learning_rate": 7.272118777828109e-06,
|
| 450 |
-
"loss": 0.018291299045085908,
|
| 451 |
-
"step": 260,
|
| 452 |
-
"token_acc": 0.994413407821229
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 1.5231764283147682,
|
| 456 |
-
"grad_norm": 0.171875,
|
| 457 |
-
"learning_rate": 7.163157255407732e-06,
|
| 458 |
-
"loss": 0.00904131457209587,
|
| 459 |
-
"step": 265,
|
| 460 |
-
"token_acc": 0.9962121212121212
|
| 461 |
-
},
|
| 462 |
-
{
|
| 463 |
-
"epoch": 1.5519223859144806,
|
| 464 |
-
"grad_norm": 9.3125,
|
| 465 |
-
"learning_rate": 7.052916818778918e-06,
|
| 466 |
-
"loss": 0.005757877603173256,
|
| 467 |
-
"step": 270,
|
| 468 |
-
"token_acc": 0.9964093357271095
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 1.5806683435141933,
|
| 472 |
-
"grad_norm": 15.5625,
|
| 473 |
-
"learning_rate": 6.941462644923318e-06,
|
| 474 |
-
"loss": 0.02215876877307892,
|
| 475 |
-
"step": 275,
|
| 476 |
-
"token_acc": 0.993963782696177
|
| 477 |
-
},
|
| 478 |
-
{
|
| 479 |
-
"epoch": 1.609414301113906,
|
| 480 |
-
"grad_norm": 0.76171875,
|
| 481 |
-
"learning_rate": 6.8288606284152535e-06,
|
| 482 |
-
"loss": 0.025442394614219665,
|
| 483 |
-
"step": 280,
|
| 484 |
-
"token_acc": 0.9944547134935305
|
| 485 |
-
},
|
| 486 |
-
{
|
| 487 |
-
"epoch": 1.6381602587136184,
|
| 488 |
-
"grad_norm": 6.96875,
|
| 489 |
-
"learning_rate": 6.715177342463145e-06,
|
| 490 |
-
"loss": 0.02161753922700882,
|
| 491 |
-
"step": 285,
|
| 492 |
-
"token_acc": 0.9906890130353817
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 1.6669062163133308,
|
| 496 |
-
"grad_norm": 0.76171875,
|
| 497 |
-
"learning_rate": 6.600479999549721e-06,
|
| 498 |
-
"loss": 0.010664404183626176,
|
| 499 |
-
"step": 290,
|
| 500 |
-
"token_acc": 0.9962825278810409
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.6956521739130435,
|
| 504 |
-
"grad_norm": 0.05859375,
|
| 505 |
-
"learning_rate": 6.4848364116942665e-06,
|
| 506 |
-
"loss": 0.011162338405847549,
|
| 507 |
-
"step": 295,
|
| 508 |
-
"token_acc": 0.9963570127504554
|
| 509 |
-
},
|
| 510 |
-
{
|
| 511 |
-
"epoch": 1.7243981315127561,
|
| 512 |
-
"grad_norm": 5.21875,
|
| 513 |
-
"learning_rate": 6.368314950360416e-06,
|
| 514 |
-
"loss": 0.006973447650671006,
|
| 515 |
-
"step": 300,
|
| 516 |
-
"token_acc": 0.9959100204498977
|
| 517 |
-
},
|
| 518 |
-
{
|
| 519 |
-
"epoch": 1.7243981315127561,
|
| 520 |
-
"eval_loss": 0.17224322259426117,
|
| 521 |
-
"eval_runtime": 53.0705,
|
| 522 |
-
"eval_samples_per_second": 0.603,
|
| 523 |
-
"eval_steps_per_second": 0.151,
|
| 524 |
-
"eval_token_acc": 0.9691629955947136,
|
| 525 |
-
"step": 300
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 1.7531440891124686,
|
| 529 |
-
"grad_norm": 12.5625,
|
| 530 |
-
"learning_rate": 6.250984506033183e-06,
|
| 531 |
-
"loss": 0.01635773330926895,
|
| 532 |
-
"step": 305,
|
| 533 |
-
"token_acc": 0.9962049335863378
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 1.781890046712181,
|
| 537 |
-
"grad_norm": 0.06298828125,
|
| 538 |
-
"learning_rate": 6.132914447489137e-06,
|
| 539 |
-
"loss": 0.013150686025619506,
|
| 540 |
-
"step": 310,
|
| 541 |
-
"token_acc": 0.9941860465116279
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 1.8106360043118936,
|
| 545 |
-
"grad_norm": 0.33984375,
|
| 546 |
-
"learning_rate": 6.014174580783794e-06,
|
| 547 |
-
"loss": 0.02508157193660736,
|
| 548 |
-
"step": 315,
|
| 549 |
-
"token_acc": 0.9938900203665988
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.8393819619116063,
|
| 553 |
-
"grad_norm": 2.625,
|
| 554 |
-
"learning_rate": 5.8948351079804875e-06,
|
| 555 |
-
"loss": 0.02004099786281586,
|
| 556 |
-
"step": 320,
|
| 557 |
-
"token_acc": 0.9939759036144579
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 1.8681279195113187,
|
| 561 |
-
"grad_norm": 0.08642578125,
|
| 562 |
-
"learning_rate": 5.774966585645092e-06,
|
| 563 |
-
"loss": 0.007430817931890488,
|
| 564 |
-
"step": 325,
|
| 565 |
-
"token_acc": 0.99800796812749
|
| 566 |
-
},
|
| 567 |
-
{
|
| 568 |
-
"epoch": 1.8968738771110312,
|
| 569 |
-
"grad_norm": 0.0888671875,
|
| 570 |
-
"learning_rate": 5.6546398831311774e-06,
|
| 571 |
-
"loss": 0.011116303503513336,
|
| 572 |
-
"step": 330,
|
| 573 |
-
"token_acc": 0.9981617647058824
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 1.9256198347107438,
|
| 577 |
-
"grad_norm": 7.15625,
|
| 578 |
-
"learning_rate": 5.533926140680222e-06,
|
| 579 |
-
"loss": 0.012766455113887788,
|
| 580 |
-
"step": 335,
|
| 581 |
-
"token_acc": 0.9943714821763602
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 1.9543657923104565,
|
| 585 |
-
"grad_norm": 7.375,
|
| 586 |
-
"learning_rate": 5.412896727361663e-06,
|
| 587 |
-
"loss": 0.01869845986366272,
|
| 588 |
-
"step": 340,
|
| 589 |
-
"token_acc": 0.9924385633270322
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 1.983111749910169,
|
| 593 |
-
"grad_norm": 0.012451171875,
|
| 594 |
-
"learning_rate": 5.29162319887768e-06,
|
| 595 |
-
"loss": 0.00366860032081604,
|
| 596 |
-
"step": 345,
|
| 597 |
-
"token_acc": 0.9982142857142857
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 2.011498383039885,
|
| 601 |
-
"grad_norm": 1.4765625,
|
| 602 |
-
"learning_rate": 5.170177255257618e-06,
|
| 603 |
-
"loss": 0.005081301927566529,
|
| 604 |
-
"step": 350,
|
| 605 |
-
"token_acc": 1.0
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 2.0402443406395974,
|
| 609 |
-
"grad_norm": 12.75,
|
| 610 |
-
"learning_rate": 5.048630698467081e-06,
|
| 611 |
-
"loss": 0.003173941746354103,
|
| 612 |
-
"step": 355,
|
| 613 |
-
"token_acc": 0.9979879275653923
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 2.06899029823931,
|
| 617 |
-
"grad_norm": 0.083984375,
|
| 618 |
-
"learning_rate": 4.927055389956768e-06,
|
| 619 |
-
"loss": 0.002004711329936981,
|
| 620 |
-
"step": 360,
|
| 621 |
-
"token_acc": 1.0
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 2.0977362558390227,
|
| 625 |
-
"grad_norm": 0.0208740234375,
|
| 626 |
-
"learning_rate": 4.805523208176139e-06,
|
| 627 |
-
"loss": 0.010144311189651489,
|
| 628 |
-
"step": 365,
|
| 629 |
-
"token_acc": 0.9961832061068703
|
| 630 |
-
},
|
| 631 |
-
{
|
| 632 |
-
"epoch": 2.1264822134387353,
|
| 633 |
-
"grad_norm": 0.00457763671875,
|
| 634 |
-
"learning_rate": 4.684106006077015e-06,
|
| 635 |
-
"loss": 0.0042584285140037535,
|
| 636 |
-
"step": 370,
|
| 637 |
-
"token_acc": 0.99800796812749
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 2.1552281710384475,
|
| 641 |
-
"grad_norm": 0.0517578125,
|
| 642 |
-
"learning_rate": 4.562875568632278e-06,
|
| 643 |
-
"loss": 0.020306271314620972,
|
| 644 |
-
"step": 375,
|
| 645 |
-
"token_acc": 0.998019801980198
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 2.18397412863816,
|
| 649 |
-
"grad_norm": 0.0123291015625,
|
| 650 |
-
"learning_rate": 4.441903570394739e-06,
|
| 651 |
-
"loss": 0.0010607664473354817,
|
| 652 |
-
"step": 380,
|
| 653 |
-
"token_acc": 1.0
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 2.212720086237873,
|
| 657 |
-
"grad_norm": 0.93359375,
|
| 658 |
-
"learning_rate": 4.321261533121303e-06,
|
| 659 |
-
"loss": 0.0018964696675539017,
|
| 660 |
-
"step": 385,
|
| 661 |
-
"token_acc": 1.0
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 2.2414660438375855,
|
| 665 |
-
"grad_norm": 6.8125,
|
| 666 |
-
"learning_rate": 4.201020783487465e-06,
|
| 667 |
-
"loss": 0.007777389883995056,
|
| 668 |
-
"step": 390,
|
| 669 |
-
"token_acc": 0.9981617647058824
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.2702120014372977,
|
| 673 |
-
"grad_norm": 0.0213623046875,
|
| 674 |
-
"learning_rate": 4.081252410917148e-06,
|
| 675 |
-
"loss": 0.01082015410065651,
|
| 676 |
-
"step": 395,
|
| 677 |
-
"token_acc": 0.9981818181818182
|
| 678 |
-
},
|
| 679 |
-
{
|
| 680 |
-
"epoch": 2.2989579590370104,
|
| 681 |
-
"grad_norm": 1.65625,
|
| 682 |
-
"learning_rate": 3.962027225552807e-06,
|
| 683 |
-
"loss": 0.003003990836441517,
|
| 684 |
-
"step": 400,
|
| 685 |
-
"token_acc": 1.0
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 2.2989579590370104,
|
| 689 |
-
"eval_loss": 0.159748375415802,
|
| 690 |
-
"eval_runtime": 52.6857,
|
| 691 |
-
"eval_samples_per_second": 0.607,
|
| 692 |
-
"eval_steps_per_second": 0.152,
|
| 693 |
-
"eval_token_acc": 0.9779735682819384,
|
| 694 |
-
"step": 400
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 2.327703916636723,
|
| 698 |
-
"grad_norm": 0.06298828125,
|
| 699 |
-
"learning_rate": 3.843415716390644e-06,
|
| 700 |
-
"loss": 0.001179580856114626,
|
| 701 |
-
"step": 405,
|
| 702 |
-
"token_acc": 1.0
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 2.3564498742364357,
|
| 706 |
-
"grad_norm": 0.98828125,
|
| 707 |
-
"learning_rate": 3.725488009605708e-06,
|
| 708 |
-
"loss": 0.00594978854060173,
|
| 709 |
-
"step": 410,
|
| 710 |
-
"token_acc": 0.9981617647058824
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 2.385195831836148,
|
| 714 |
-
"grad_norm": 0.0213623046875,
|
| 715 |
-
"learning_rate": 3.6083138270914932e-06,
|
| 716 |
-
"loss": 0.008244612067937852,
|
| 717 |
-
"step": 415,
|
| 718 |
-
"token_acc": 0.9980769230769231
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 2.4139417894358606,
|
| 722 |
-
"grad_norm": 0.2353515625,
|
| 723 |
-
"learning_rate": 3.491962445238569e-06,
|
| 724 |
-
"loss": 0.007743784040212631,
|
| 725 |
-
"step": 420,
|
| 726 |
-
"token_acc": 0.9961759082217974
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 2.4426877470355732,
|
| 730 |
-
"grad_norm": 0.01953125,
|
| 731 |
-
"learning_rate": 3.3765026539765832e-06,
|
| 732 |
-
"loss": 0.020112380385398865,
|
| 733 |
-
"step": 425,
|
| 734 |
-
"token_acc": 0.9941634241245136
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 2.4714337046352854,
|
| 738 |
-
"grad_norm": 4.9375,
|
| 739 |
-
"learning_rate": 3.2620027161038975e-06,
|
| 740 |
-
"loss": 0.00260693933814764,
|
| 741 |
-
"step": 430,
|
| 742 |
-
"token_acc": 0.9981481481481481
|
| 743 |
-
},
|
| 744 |
-
{
|
| 745 |
-
"epoch": 2.500179662234998,
|
| 746 |
-
"grad_norm": 2.375,
|
| 747 |
-
"learning_rate": 3.1485303269288603e-06,
|
| 748 |
-
"loss": 0.0008619253523647786,
|
| 749 |
-
"step": 435,
|
| 750 |
-
"token_acc": 1.0
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 2.5289256198347108,
|
| 754 |
-
"grad_norm": 0.109375,
|
| 755 |
-
"learning_rate": 3.0361525742465975e-06,
|
| 756 |
-
"loss": 0.0006391198374330997,
|
| 757 |
-
"step": 440,
|
| 758 |
-
"token_acc": 1.0
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 2.5576715774344234,
|
| 762 |
-
"grad_norm": 0.50390625,
|
| 763 |
-
"learning_rate": 2.9249358986749922e-06,
|
| 764 |
-
"loss": 0.00029405634850263593,
|
| 765 |
-
"step": 445,
|
| 766 |
-
"token_acc": 1.0
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 2.586417535034136,
|
| 770 |
-
"grad_norm": 0.00830078125,
|
| 771 |
-
"learning_rate": 2.8149460543732666e-06,
|
| 772 |
-
"loss": 0.001843920163810253,
|
| 773 |
-
"step": 450,
|
| 774 |
-
"token_acc": 0.998015873015873
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 2.6151634926338483,
|
| 778 |
-
"grad_norm": 0.1572265625,
|
| 779 |
-
"learning_rate": 2.706248070166449e-06,
|
| 780 |
-
"loss": 0.0003178727813065052,
|
| 781 |
-
"step": 455,
|
| 782 |
-
"token_acc": 1.0
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 2.643909450233561,
|
| 786 |
-
"grad_norm": 0.018798828125,
|
| 787 |
-
"learning_rate": 2.598906211098643e-06,
|
| 788 |
-
"loss": 0.0005950949620455503,
|
| 789 |
-
"step": 460,
|
| 790 |
-
"token_acc": 1.0
|
| 791 |
-
},
|
| 792 |
-
{
|
| 793 |
-
"epoch": 2.6726554078332736,
|
| 794 |
-
"grad_norm": 0.76953125,
|
| 795 |
-
"learning_rate": 2.4929839404378935e-06,
|
| 796 |
-
"loss": 0.0009388749487698079,
|
| 797 |
-
"step": 465,
|
| 798 |
-
"token_acc": 1.0
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 2.701401365432986,
|
| 802 |
-
"grad_norm": 0.032470703125,
|
| 803 |
-
"learning_rate": 2.388543882155067e-06,
|
| 804 |
-
"loss": 0.00020109661854803563,
|
| 805 |
-
"step": 470,
|
| 806 |
-
"token_acc": 1.0
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 2.7301473230326985,
|
| 810 |
-
"grad_norm": 0.0045166015625,
|
| 811 |
-
"learning_rate": 2.2856477838989455e-06,
|
| 812 |
-
"loss": 0.0013651649467647075,
|
| 813 |
-
"step": 475,
|
| 814 |
-
"token_acc": 1.0
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 2.758893280632411,
|
| 818 |
-
"grad_norm": 0.158203125,
|
| 819 |
-
"learning_rate": 2.1843564804894316e-06,
|
| 820 |
-
"loss": 0.0003602589713409543,
|
| 821 |
-
"step": 480,
|
| 822 |
-
"token_acc": 1.0
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 2.7876392382321233,
|
| 826 |
-
"grad_norm": 0.453125,
|
| 827 |
-
"learning_rate": 2.0847298579504344e-06,
|
| 828 |
-
"loss": 0.00022451074328273536,
|
| 829 |
-
"step": 485,
|
| 830 |
-
"token_acc": 1.0
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 2.816385195831836,
|
| 834 |
-
"grad_norm": 0.330078125,
|
| 835 |
-
"learning_rate": 1.9868268181037186e-06,
|
| 836 |
-
"loss": 0.00015194707084447144,
|
| 837 |
-
"step": 490,
|
| 838 |
-
"token_acc": 1.0
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 2.8451311534315487,
|
| 842 |
-
"grad_norm": 0.0205078125,
|
| 843 |
-
"learning_rate": 1.8907052437446271e-06,
|
| 844 |
-
"loss": 0.00033508068881928923,
|
| 845 |
-
"step": 495,
|
| 846 |
-
"token_acc": 1.0
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"epoch": 2.8738771110312613,
|
| 850 |
-
"grad_norm": 0.005645751953125,
|
| 851 |
-
"learning_rate": 1.7964219644202852e-06,
|
| 852 |
-
"loss": 0.0003983211703598499,
|
| 853 |
-
"step": 500,
|
| 854 |
-
"token_acc": 1.0
|
| 855 |
-
},
|
| 856 |
-
{
|
| 857 |
-
"epoch": 2.8738771110312613,
|
| 858 |
-
"eval_loss": 0.18025675415992737,
|
| 859 |
-
"eval_runtime": 53.2379,
|
| 860 |
-
"eval_samples_per_second": 0.601,
|
| 861 |
-
"eval_steps_per_second": 0.15,
|
| 862 |
-
"eval_token_acc": 0.973568281938326,
|
| 863 |
-
"step": 500
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 2.902623068630974,
|
| 867 |
-
"grad_norm": 5.71875,
|
| 868 |
-
"learning_rate": 1.704032722830512e-06,
|
| 869 |
-
"loss": 0.0010727177374064921,
|
| 870 |
-
"step": 505,
|
| 871 |
-
"token_acc": 1.0
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 2.931369026230686,
|
| 875 |
-
"grad_norm": 3.140625,
|
| 876 |
-
"learning_rate": 1.6135921418712959e-06,
|
| 877 |
-
"loss": 0.011727692186832428,
|
| 878 |
-
"step": 510,
|
| 879 |
-
"token_acc": 0.9961832061068703
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 2.960114983830399,
|
| 883 |
-
"grad_norm": 0.162109375,
|
| 884 |
-
"learning_rate": 1.5251536923403427e-06,
|
| 885 |
-
"loss": 0.014498487114906311,
|
| 886 |
-
"step": 515,
|
| 887 |
-
"token_acc": 0.9963235294117647
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 2.9888609414301115,
|
| 891 |
-
"grad_norm": 0.059326171875,
|
| 892 |
-
"learning_rate": 1.4387696613237612e-06,
|
| 893 |
-
"loss": 3.579054027795792e-05,
|
| 894 |
-
"step": 520,
|
| 895 |
-
"token_acc": 1.0
|
| 896 |
-
},
|
| 897 |
-
{
|
| 898 |
-
"epoch": 3.0172475745598275,
|
| 899 |
-
"grad_norm": 0.0167236328125,
|
| 900 |
-
"learning_rate": 1.3544911212825907e-06,
|
| 901 |
-
"loss": 0.00019136755727231503,
|
| 902 |
-
"step": 525,
|
| 903 |
-
"token_acc": 1.0
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 3.04599353215954,
|
| 907 |
-
"grad_norm": 0.0107421875,
|
| 908 |
-
"learning_rate": 1.2723678998574512e-06,
|
| 909 |
-
"loss": 9.603083017282188e-05,
|
| 910 |
-
"step": 530,
|
| 911 |
-
"token_acc": 1.0
|
| 912 |
-
},
|
| 913 |
-
{
|
| 914 |
-
"epoch": 3.074739489759253,
|
| 915 |
-
"grad_norm": 0.04345703125,
|
| 916 |
-
"learning_rate": 1.1924485504091565e-06,
|
| 917 |
-
"loss": 7.079667993821203e-05,
|
| 918 |
-
"step": 535,
|
| 919 |
-
"token_acc": 1.0
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 3.103485447358965,
|
| 923 |
-
"grad_norm": 0.1123046875,
|
| 924 |
-
"learning_rate": 1.1147803233127241e-06,
|
| 925 |
-
"loss": 0.004302733391523361,
|
| 926 |
-
"step": 540,
|
| 927 |
-
"token_acc": 0.9961013645224172
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 3.1322314049586777,
|
| 931 |
-
"grad_norm": 0.0048828125,
|
| 932 |
-
"learning_rate": 1.0394091380217354e-06,
|
| 933 |
-
"loss": 5.177758866921067e-05,
|
| 934 |
-
"step": 545,
|
| 935 |
-
"token_acc": 1.0
|
| 936 |
-
},
|
| 937 |
-
{
|
| 938 |
-
"epoch": 3.1609773625583903,
|
| 939 |
-
"grad_norm": 0.00408935546875,
|
| 940 |
-
"learning_rate": 9.663795559195733e-07,
|
| 941 |
-
"loss": 5.5760378018021584e-05,
|
| 942 |
-
"step": 550,
|
| 943 |
-
"token_acc": 1.0
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 3.1897233201581026,
|
| 947 |
-
"grad_norm": 0.06103515625,
|
| 948 |
-
"learning_rate": 8.957347539735872e-07,
|
| 949 |
-
"loss": 6.924450281076133e-05,
|
| 950 |
-
"step": 555,
|
| 951 |
-
"token_acc": 1.0
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 3.218469277757815,
|
| 955 |
-
"grad_norm": 0.03955078125,
|
| 956 |
-
"learning_rate": 8.275164992077555e-07,
|
| 957 |
-
"loss": 5.618180148303509e-05,
|
| 958 |
-
"step": 560,
|
| 959 |
-
"token_acc": 1.0
|
| 960 |
-
},
|
| 961 |
-
{
|
| 962 |
-
"epoch": 3.247215235357528,
|
| 963 |
-
"grad_norm": 0.10400390625,
|
| 964 |
-
"learning_rate": 7.617651240089546e-07,
|
| 965 |
-
"loss": 0.002662208303809166,
|
| 966 |
-
"step": 565,
|
| 967 |
-
"token_acc": 0.9981785063752276
|
| 968 |
-
},
|
| 969 |
-
{
|
| 970 |
-
"epoch": 3.2759611929572405,
|
| 971 |
-
"grad_norm": 0.01519775390625,
|
| 972 |
-
"learning_rate": 6.985195022814068e-07,
|
| 973 |
-
"loss": 0.00019990296568721532,
|
| 974 |
-
"step": 570,
|
| 975 |
-
"token_acc": 1.0
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 3.3047071505569527,
|
| 979 |
-
"grad_norm": 0.00927734375,
|
| 980 |
-
"learning_rate": 6.37817026463432e-07,
|
| 981 |
-
"loss": 0.0002466671634465456,
|
| 982 |
-
"step": 575,
|
| 983 |
-
"token_acc": 1.0
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 3.3334531081566654,
|
| 987 |
-
"grad_norm": 0.103515625,
|
| 988 |
-
"learning_rate": 5.796935854200764e-07,
|
| 989 |
-
"loss": 6.880384171381593e-05,
|
| 990 |
-
"step": 580,
|
| 991 |
-
"token_acc": 1.0
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 3.362199065756378,
|
| 995 |
-
"grad_norm": 0.07763671875,
|
| 996 |
-
"learning_rate": 5.241835432246888e-07,
|
| 997 |
-
"loss": 4.42019896581769e-05,
|
| 998 |
-
"step": 585,
|
| 999 |
-
"token_acc": 1.0
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 3.3909450233560907,
|
| 1003 |
-
"grad_norm": 0.1806640625,
|
| 1004 |
-
"learning_rate": 4.713197188420027e-07,
|
| 1005 |
-
"loss": 0.00010410865070298314,
|
| 1006 |
-
"step": 590,
|
| 1007 |
-
"token_acc": 1.0
|
| 1008 |
-
},
|
| 1009 |
-
{
|
| 1010 |
-
"epoch": 3.419690980955803,
|
| 1011 |
-
"grad_norm": 0.388671875,
|
| 1012 |
-
"learning_rate": 4.211333667247125e-07,
|
| 1013 |
-
"loss": 0.0003400356974452734,
|
| 1014 |
-
"step": 595,
|
| 1015 |
-
"token_acc": 1.0
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 3.4484369385555156,
|
| 1019 |
-
"grad_norm": 0.0164794921875,
|
| 1020 |
-
"learning_rate": 3.736541583350473e-07,
|
| 1021 |
-
"loss": 0.00011566770263016224,
|
| 1022 |
-
"step": 600,
|
| 1023 |
-
"token_acc": 1.0
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 3.4484369385555156,
|
| 1027 |
-
"eval_loss": 0.18067888915538788,
|
| 1028 |
-
"eval_runtime": 53.1476,
|
| 1029 |
-
"eval_samples_per_second": 0.602,
|
| 1030 |
-
"eval_steps_per_second": 0.151,
|
| 1031 |
-
"eval_token_acc": 0.973568281938326,
|
| 1032 |
-
"step": 600
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 3.4771828961552282,
|
| 1036 |
-
"grad_norm": 0.0361328125,
|
| 1037 |
-
"learning_rate": 3.289101646022297e-07,
|
| 1038 |
-
"loss": 0.0003246673382818699,
|
| 1039 |
-
"step": 605,
|
| 1040 |
-
"token_acc": 1.0
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 3.5059288537549405,
|
| 1044 |
-
"grad_norm": 0.038330078125,
|
| 1045 |
-
"learning_rate": 2.869278393262226e-07,
|
| 1046 |
-
"loss": 4.617357917595655e-05,
|
| 1047 |
-
"step": 610,
|
| 1048 |
-
"token_acc": 1.0
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 3.534674811354653,
|
| 1052 |
-
"grad_norm": 0.01263427734375,
|
| 1053 |
-
"learning_rate": 2.47732003537568e-07,
|
| 1054 |
-
"loss": 4.190525796730071e-05,
|
| 1055 |
-
"step": 615,
|
| 1056 |
-
"token_acc": 1.0
|
| 1057 |
-
},
|
| 1058 |
-
{
|
| 1059 |
-
"epoch": 3.5634207689543658,
|
| 1060 |
-
"grad_norm": 0.05224609375,
|
| 1061 |
-
"learning_rate": 2.113458308225458e-07,
|
| 1062 |
-
"loss": 8.594701066613197e-05,
|
| 1063 |
-
"step": 620,
|
| 1064 |
-
"token_acc": 1.0
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 3.5921667265540784,
|
| 1068 |
-
"grad_norm": 0.099609375,
|
| 1069 |
-
"learning_rate": 1.777908336223655e-07,
|
| 1070 |
-
"loss": 0.00013726772740483285,
|
| 1071 |
-
"step": 625,
|
| 1072 |
-
"token_acc": 1.0
|
| 1073 |
-
},
|
| 1074 |
-
{
|
| 1075 |
-
"epoch": 3.620912684153791,
|
| 1076 |
-
"grad_norm": 0.11572265625,
|
| 1077 |
-
"learning_rate": 1.4708685051444515e-07,
|
| 1078 |
-
"loss": 0.0001238467637449503,
|
| 1079 |
-
"step": 630,
|
| 1080 |
-
"token_acc": 1.0
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 3.6496586417535033,
|
| 1084 |
-
"grad_norm": 0.19140625,
|
| 1085 |
-
"learning_rate": 1.19252034483342e-07,
|
| 1086 |
-
"loss": 0.0001841937890276313,
|
| 1087 |
-
"step": 635,
|
| 1088 |
-
"token_acc": 1.0
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 3.678404599353216,
|
| 1092 |
-
"grad_norm": 0.017578125,
|
| 1093 |
-
"learning_rate": 9.430284218824026e-08,
|
| 1094 |
-
"loss": 4.8907153541222215e-05,
|
| 1095 |
-
"step": 640,
|
| 1096 |
-
"token_acc": 1.0
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 3.7071505569529286,
|
| 1100 |
-
"grad_norm": 0.028564453125,
|
| 1101 |
-
"learning_rate": 7.225402423334694e-08,
|
| 1102 |
-
"loss": 5.174498655833304e-05,
|
| 1103 |
-
"step": 645,
|
| 1104 |
-
"token_acc": 1.0
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 3.735896514552641,
|
| 1108 |
-
"grad_norm": 0.0045166015625,
|
| 1109 |
-
"learning_rate": 5.3118616446960484e-08,
|
| 1110 |
-
"loss": 0.00013846734073013066,
|
| 1111 |
-
"step": 650,
|
| 1112 |
-
"token_acc": 1.0
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 3.7646424721523535,
|
| 1116 |
-
"grad_norm": 0.019775390625,
|
| 1117 |
-
"learning_rate": 3.690793217434985e-08,
|
| 1118 |
-
"loss": 0.00016624389681965112,
|
| 1119 |
-
"step": 655,
|
| 1120 |
-
"token_acc": 1.0
|
| 1121 |
-
},
|
| 1122 |
-
{
|
| 1123 |
-
"epoch": 3.793388429752066,
|
| 1124 |
-
"grad_norm": 0.06005859375,
|
| 1125 |
-
"learning_rate": 2.363155558901542e-08,
|
| 1126 |
-
"loss": 0.00012955744750797749,
|
| 1127 |
-
"step": 660,
|
| 1128 |
-
"token_acc": 1.0
|
| 1129 |
-
},
|
| 1130 |
-
{
|
| 1131 |
-
"epoch": 3.822134387351779,
|
| 1132 |
-
"grad_norm": 0.03662109375,
|
| 1133 |
-
"learning_rate": 1.3297336026280027e-08,
|
| 1134 |
-
"loss": 7.39644980058074e-05,
|
| 1135 |
-
"step": 665,
|
| 1136 |
-
"token_acc": 1.0
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 3.8508803449514915,
|
| 1140 |
-
"grad_norm": 0.004058837890625,
|
| 1141 |
-
"learning_rate": 5.911383342556143e-09,
|
| 1142 |
-
"loss": 9.104540222324431e-05,
|
| 1143 |
-
"step": 670,
|
| 1144 |
-
"token_acc": 1.0
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 3.8796263025512037,
|
| 1148 |
-
"grad_norm": 0.10302734375,
|
| 1149 |
-
"learning_rate": 1.4780643030476439e-09,
|
| 1150 |
-
"loss": 0.00016065937234088778,
|
| 1151 |
-
"step": 675,
|
| 1152 |
-
"token_acc": 1.0
|
| 1153 |
-
},
|
| 1154 |
-
{
|
| 1155 |
-
"epoch": 3.9083722601509163,
|
| 1156 |
-
"grad_norm": 8.5,
|
| 1157 |
-
"learning_rate": 0.0,
|
| 1158 |
-
"loss": 0.004035552591085434,
|
| 1159 |
-
"step": 680,
|
| 1160 |
-
"token_acc": 0.998158379373849
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 3.9083722601509163,
|
| 1164 |
-
"eval_loss": 0.17740729451179504,
|
| 1165 |
-
"eval_runtime": 52.9969,
|
| 1166 |
-
"eval_samples_per_second": 0.604,
|
| 1167 |
-
"eval_steps_per_second": 0.151,
|
| 1168 |
-
"eval_token_acc": 0.973568281938326,
|
| 1169 |
-
"step": 680
|
| 1170 |
-
}
|
| 1171 |
-
],
|
| 1172 |
-
"logging_steps": 5,
|
| 1173 |
-
"max_steps": 680,
|
| 1174 |
-
"num_input_tokens_seen": 0,
|
| 1175 |
-
"num_train_epochs": 4,
|
| 1176 |
-
"save_steps": 100,
|
| 1177 |
-
"stateful_callbacks": {
|
| 1178 |
-
"TrainerControl": {
|
| 1179 |
-
"args": {
|
| 1180 |
-
"should_epoch_stop": false,
|
| 1181 |
-
"should_evaluate": false,
|
| 1182 |
-
"should_log": false,
|
| 1183 |
-
"should_save": true,
|
| 1184 |
-
"should_training_stop": true
|
| 1185 |
-
},
|
| 1186 |
-
"attributes": {}
|
| 1187 |
-
}
|
| 1188 |
-
},
|
| 1189 |
-
"total_flos": 1.1056749194178567e+20,
|
| 1190 |
-
"train_batch_size": 1,
|
| 1191 |
-
"trial_name": null,
|
| 1192 |
-
"trial_params": null
|
| 1193 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|