Update Qwen2.5-3B-Instruct PersonaMem-128K reader from validation-selected optimization
Browse files
README.md
CHANGED
|
@@ -77,6 +77,18 @@ Default reader: **highest-lr-step-1000**. Only the final OPD + GRPO phase was re
|
|
| 77 |
|
| 78 |
273 validation questions and 233 test questions; greedy writer and reader, five fixed option permutations. Use the [matching 128K code and protocol](https://github.com/Johnny221B/memfold/tree/c242a424bc98f2cb11e9d456f60ccd1e2158e74a) for YaRN4.5, full inputs without truncation, and the historical answer-format normalization. [Full settings and results](personamem-128k/qwen3-4b/optimization.json). Previous weights remain at revision `7de39f1557b44a3cdac0dec768f31e62e0c2180f`. These post-release results do not replace historical paper results.
|
| 79 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
## Components and names
|
| 81 |
|
| 82 |
MemFold uses one workflow:
|
|
|
|
| 77 |
|
| 78 |
273 validation questions and 233 test questions; greedy writer and reader, five fixed option permutations. Use the [matching 128K code and protocol](https://github.com/Johnny221B/memfold/tree/c242a424bc98f2cb11e9d456f60ccd1e2158e74a) for YaRN4.5, full inputs without truncation, and the historical answer-format normalization. [Full settings and results](personamem-128k/qwen3-4b/optimization.json). Previous weights remain at revision `7de39f1557b44a3cdac0dec768f31e62e0c2180f`. These post-release results do not replace historical paper results.
|
| 79 |
|
| 80 |
+
## Updated PersonaMem-128K / Qwen2.5-3B-Instruct
|
| 81 |
+
|
| 82 |
+
Default reader: **current-ratio-step-600**. Only the final OPD + GRPO phase was retrained; its compressor is unchanged. Selected by validation mean, with test evaluated afterward.
|
| 83 |
+
|
| 84 |
+
| Split | Mean accuracy | Best trial |
|
| 85 |
+
|---|---:|---:|
|
| 86 |
+
| Previous reader validation | 89.01% | 91.58% |
|
| 87 |
+
| Updated reader validation | 89.23% | 90.84% |
|
| 88 |
+
| Updated reader test | 87.64% | 89.27% |
|
| 89 |
+
|
| 90 |
+
273 validation questions and 233 test questions; greedy writer and reader, five fixed option permutations. Use the [matching 128K code and protocol](https://github.com/Johnny221B/memfold/tree/c242a424bc98f2cb11e9d456f60ccd1e2158e74a) for YaRN4.5, full inputs without truncation, and the historical answer-format normalization. [Full settings and results](personamem-128k/qwen2.5-3b/optimization.json). Previous weights remain at revision `71a718d07b7c61b3fdf53509557e05571368ba8d`. These post-release results do not replace historical paper results.
|
| 91 |
+
|
| 92 |
## Components and names
|
| 93 |
|
| 94 |
MemFold uses one workflow:
|
VALIDATION.json
CHANGED
|
@@ -425,6 +425,88 @@
|
|
| 425 |
"finite_adapter": true,
|
| 426 |
"matching_frozen_compressor": true,
|
| 427 |
"details": "personamem-128k/qwen3-4b/optimization.json"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 428 |
}
|
| 429 |
}
|
| 430 |
}
|
|
|
|
| 425 |
"finite_adapter": true,
|
| 426 |
"matching_frozen_compressor": true,
|
| 427 |
"details": "personamem-128k/qwen3-4b/optimization.json"
|
| 428 |
+
},
|
| 429 |
+
"personamem-128k/qwen2.5-3b": {
|
| 430 |
+
"validation": {
|
| 431 |
+
"name": "current-ratio-step-600",
|
| 432 |
+
"trials": [
|
| 433 |
+
{
|
| 434 |
+
"trial": 0,
|
| 435 |
+
"correct": 245,
|
| 436 |
+
"total": 273,
|
| 437 |
+
"accuracy": 0.8974358974358975
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"trial": 1,
|
| 441 |
+
"correct": 244,
|
| 442 |
+
"total": 273,
|
| 443 |
+
"accuracy": 0.8937728937728938
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"trial": 2,
|
| 447 |
+
"correct": 240,
|
| 448 |
+
"total": 273,
|
| 449 |
+
"accuracy": 0.8791208791208791
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"trial": 3,
|
| 453 |
+
"correct": 248,
|
| 454 |
+
"total": 273,
|
| 455 |
+
"accuracy": 0.9084249084249084
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"trial": 4,
|
| 459 |
+
"correct": 241,
|
| 460 |
+
"total": 273,
|
| 461 |
+
"accuracy": 0.8827838827838828
|
| 462 |
+
}
|
| 463 |
+
],
|
| 464 |
+
"mean_accuracy": 0.8923076923076924,
|
| 465 |
+
"best_accuracy": 0.9084249084249084,
|
| 466 |
+
"pipeline_seconds": 517.7519376315176
|
| 467 |
+
},
|
| 468 |
+
"test": {
|
| 469 |
+
"name": "selected",
|
| 470 |
+
"trials": [
|
| 471 |
+
{
|
| 472 |
+
"trial": 0,
|
| 473 |
+
"correct": 203,
|
| 474 |
+
"total": 233,
|
| 475 |
+
"accuracy": 0.871244635193133
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"trial": 1,
|
| 479 |
+
"correct": 203,
|
| 480 |
+
"total": 233,
|
| 481 |
+
"accuracy": 0.871244635193133
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"trial": 2,
|
| 485 |
+
"correct": 208,
|
| 486 |
+
"total": 233,
|
| 487 |
+
"accuracy": 0.8927038626609443
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"trial": 3,
|
| 491 |
+
"correct": 205,
|
| 492 |
+
"total": 233,
|
| 493 |
+
"accuracy": 0.8798283261802575
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"trial": 4,
|
| 497 |
+
"correct": 202,
|
| 498 |
+
"total": 233,
|
| 499 |
+
"accuracy": 0.8669527896995708
|
| 500 |
+
}
|
| 501 |
+
],
|
| 502 |
+
"mean_accuracy": 0.8763948497854077,
|
| 503 |
+
"best_accuracy": 0.8927038626609443,
|
| 504 |
+
"pipeline_seconds": 370.2440517041832
|
| 505 |
+
},
|
| 506 |
+
"adapter_config_load": true,
|
| 507 |
+
"finite_adapter": true,
|
| 508 |
+
"matching_frozen_compressor": true,
|
| 509 |
+
"details": "personamem-128k/qwen2.5-3b/optimization.json"
|
| 510 |
}
|
| 511 |
}
|
| 512 |
}
|
manifest.json
CHANGED
|
@@ -250,16 +250,16 @@
|
|
| 250 |
"path": "personamem-128k/qwen2.5-3b/reader/adapter_model.safetensors",
|
| 251 |
"role": "reader_lora",
|
| 252 |
"size": 119801528,
|
| 253 |
-
"sha256": "
|
| 254 |
-
"original_sha256": "
|
| 255 |
"metadata_only_repack": false
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"path": "personamem-128k/qwen2.5-3b/reader/adapter_config.json",
|
| 259 |
"role": "reader_config",
|
| 260 |
"size": 975,
|
| 261 |
-
"sha256": "
|
| 262 |
-
"original_sha256": "
|
| 263 |
"metadata_only_repack": true
|
| 264 |
},
|
| 265 |
{
|
|
@@ -414,7 +414,8 @@
|
|
| 414 |
"2026-09-29: PersonaMem-32K Qwen3-4B reader updated to validation-selected more-grpo epoch 5 (615 steps). Original paper release remains at revision d39842eeca8132288c5e2814fdd66733fa92fc3d. See OPTIMIZATION.json.",
|
| 415 |
"Post-release update of personamem-32k/qwen2.5-3b: current-ratio-epoch-4, selected by greedy validation mean. Previous weights: 10e71a32171ccde6f3a53bc8ee5b407264321c3d. See personamem-32k/qwen2.5-3b/optimization.json.",
|
| 416 |
"Post-release update of personamem-32k/qwen2.5-7b: current-ratio-epoch-5, selected by greedy validation mean. Previous weights: c86ee86e01369478b300007d7500a1a63ea582b2. See personamem-32k/qwen2.5-7b/optimization.json.",
|
| 417 |
-
"personamem-128k/qwen3-4b updated with validation-selected 128K final-phase optimization. Protocol and matching code: personamem-128k/qwen3-4b/optimization.json."
|
|
|
|
| 418 |
],
|
| 419 |
"default_qwen3_4b_personamem32k": {
|
| 420 |
"configuration": "more-grpo",
|
|
@@ -444,6 +445,13 @@
|
|
| 444 |
"previous_revision": "7de39f1557b44a3cdac0dec768f31e62e0c2180f",
|
| 445 |
"validation_mean": 0.8915750915750916,
|
| 446 |
"test_mean": 0.8763948497854077
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 447 |
}
|
| 448 |
}
|
| 449 |
}
|
|
|
|
| 250 |
"path": "personamem-128k/qwen2.5-3b/reader/adapter_model.safetensors",
|
| 251 |
"role": "reader_lora",
|
| 252 |
"size": 119801528,
|
| 253 |
+
"sha256": "727d8e4de0ad4ffe224124e3d3a515ae0f1a8c919f3828c840a0d7e31b512006",
|
| 254 |
+
"original_sha256": "727d8e4de0ad4ffe224124e3d3a515ae0f1a8c919f3828c840a0d7e31b512006",
|
| 255 |
"metadata_only_repack": false
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"path": "personamem-128k/qwen2.5-3b/reader/adapter_config.json",
|
| 259 |
"role": "reader_config",
|
| 260 |
"size": 975,
|
| 261 |
+
"sha256": "390450b001ca8ec2440df5edbb55f5d8662624a1da9508d424fdee100d822a55",
|
| 262 |
+
"original_sha256": "43eb328a7aaa99b7859ba62a69dcf5fb33ed7a6dd92274166e06d38920fb4406",
|
| 263 |
"metadata_only_repack": true
|
| 264 |
},
|
| 265 |
{
|
|
|
|
| 414 |
"2026-09-29: PersonaMem-32K Qwen3-4B reader updated to validation-selected more-grpo epoch 5 (615 steps). Original paper release remains at revision d39842eeca8132288c5e2814fdd66733fa92fc3d. See OPTIMIZATION.json.",
|
| 415 |
"Post-release update of personamem-32k/qwen2.5-3b: current-ratio-epoch-4, selected by greedy validation mean. Previous weights: 10e71a32171ccde6f3a53bc8ee5b407264321c3d. See personamem-32k/qwen2.5-3b/optimization.json.",
|
| 416 |
"Post-release update of personamem-32k/qwen2.5-7b: current-ratio-epoch-5, selected by greedy validation mean. Previous weights: c86ee86e01369478b300007d7500a1a63ea582b2. See personamem-32k/qwen2.5-7b/optimization.json.",
|
| 417 |
+
"personamem-128k/qwen3-4b updated with validation-selected 128K final-phase optimization. Protocol and matching code: personamem-128k/qwen3-4b/optimization.json.",
|
| 418 |
+
"personamem-128k/qwen2.5-3b updated with validation-selected 128K final-phase optimization. Protocol and matching code: personamem-128k/qwen2.5-3b/optimization.json."
|
| 419 |
],
|
| 420 |
"default_qwen3_4b_personamem32k": {
|
| 421 |
"configuration": "more-grpo",
|
|
|
|
| 445 |
"previous_revision": "7de39f1557b44a3cdac0dec768f31e62e0c2180f",
|
| 446 |
"validation_mean": 0.8915750915750916,
|
| 447 |
"test_mean": 0.8763948497854077
|
| 448 |
+
},
|
| 449 |
+
"personamem-128k/qwen2.5-3b": {
|
| 450 |
+
"checkpoint": "current-ratio-step-600",
|
| 451 |
+
"code_revision": "c242a424bc98f2cb11e9d456f60ccd1e2158e74a",
|
| 452 |
+
"previous_revision": "71a718d07b7c61b3fdf53509557e05571368ba8d",
|
| 453 |
+
"validation_mean": 0.8923076923076924,
|
| 454 |
+
"test_mean": 0.8763948497854077
|
| 455 |
}
|
| 456 |
}
|
| 457 |
}
|
personamem-128k/qwen2.5-3b/optimization.json
ADDED
|
@@ -0,0 +1,902 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dataset": "PersonaMem-128K",
|
| 3 |
+
"model": "Qwen2.5-3B-Instruct",
|
| 4 |
+
"code_revision": "c242a424bc98f2cb11e9d456f60ccd1e2158e74a",
|
| 5 |
+
"phase": "on_policy_optimization",
|
| 6 |
+
"selected_checkpoint": "current-ratio-step-600",
|
| 7 |
+
"selection": "Highest greedy validation mean across five fixed option permutations; ties prefer earlier step and configuration order. Test only after selection.",
|
| 8 |
+
"validation_delta_vs_incumbent": 0.00219780219780219,
|
| 9 |
+
"configs": [
|
| 10 |
+
{
|
| 11 |
+
"name": "current-ratio",
|
| 12 |
+
"learning_rate": 3e-07,
|
| 13 |
+
"opd_weight": 0.002,
|
| 14 |
+
"grpo_weight": 0.3
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"name": "lower-lr",
|
| 18 |
+
"learning_rate": 1.5e-07,
|
| 19 |
+
"opd_weight": 0.002,
|
| 20 |
+
"grpo_weight": 0.3
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"name": "more-grpo",
|
| 24 |
+
"learning_rate": 1.5e-07,
|
| 25 |
+
"opd_weight": 0.0010033222591362126,
|
| 26 |
+
"grpo_weight": 0.30099667774086375
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"name": "more-opd",
|
| 30 |
+
"learning_rate": 1.5e-07,
|
| 31 |
+
"opd_weight": 0.003973684210526316,
|
| 32 |
+
"grpo_weight": 0.29802631578947364
|
| 33 |
+
}
|
| 34 |
+
],
|
| 35 |
+
"steps_per_configuration": 1000,
|
| 36 |
+
"save_steps": [
|
| 37 |
+
200,
|
| 38 |
+
400,
|
| 39 |
+
600,
|
| 40 |
+
800,
|
| 41 |
+
1000
|
| 42 |
+
],
|
| 43 |
+
"world_size": 2,
|
| 44 |
+
"questions_per_update": 2,
|
| 45 |
+
"rollouts_per_question": 8,
|
| 46 |
+
"rollouts_per_update": 16,
|
| 47 |
+
"reference_kl_weight": 0,
|
| 48 |
+
"teacher_update_mode": "frozen_initial",
|
| 49 |
+
"sampling_temperature": 1.0,
|
| 50 |
+
"top_p": 0.98,
|
| 51 |
+
"seed": 42,
|
| 52 |
+
"train_examples": 2221,
|
| 53 |
+
"validation_examples": 273,
|
| 54 |
+
"test_examples": 233,
|
| 55 |
+
"evaluation": "Greedy writer and reader; five deterministic option permutations; historical explicit-choice normalization. Full writer input, YaRN4.5, max_model_len140000.",
|
| 56 |
+
"writer_memory_budget": 4096,
|
| 57 |
+
"validation_results": [
|
| 58 |
+
{
|
| 59 |
+
"name": "current-ratio-step-200",
|
| 60 |
+
"trials": [
|
| 61 |
+
{
|
| 62 |
+
"trial": 0,
|
| 63 |
+
"correct": 240,
|
| 64 |
+
"total": 273,
|
| 65 |
+
"accuracy": 0.8791208791208791
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"trial": 1,
|
| 69 |
+
"correct": 241,
|
| 70 |
+
"total": 273,
|
| 71 |
+
"accuracy": 0.8827838827838828
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"trial": 2,
|
| 75 |
+
"correct": 236,
|
| 76 |
+
"total": 273,
|
| 77 |
+
"accuracy": 0.8644688644688645
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"trial": 3,
|
| 81 |
+
"correct": 248,
|
| 82 |
+
"total": 273,
|
| 83 |
+
"accuracy": 0.9084249084249084
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"trial": 4,
|
| 87 |
+
"correct": 236,
|
| 88 |
+
"total": 273,
|
| 89 |
+
"accuracy": 0.8644688644688645
|
| 90 |
+
}
|
| 91 |
+
],
|
| 92 |
+
"mean_accuracy": 0.87985347985348,
|
| 93 |
+
"best_accuracy": 0.9084249084249084,
|
| 94 |
+
"pipeline_seconds": 524.9968059062958
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"name": "current-ratio-step-400",
|
| 98 |
+
"trials": [
|
| 99 |
+
{
|
| 100 |
+
"trial": 0,
|
| 101 |
+
"correct": 242,
|
| 102 |
+
"total": 273,
|
| 103 |
+
"accuracy": 0.8864468864468864
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"trial": 1,
|
| 107 |
+
"correct": 244,
|
| 108 |
+
"total": 273,
|
| 109 |
+
"accuracy": 0.8937728937728938
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"trial": 2,
|
| 113 |
+
"correct": 236,
|
| 114 |
+
"total": 273,
|
| 115 |
+
"accuracy": 0.8644688644688645
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"trial": 3,
|
| 119 |
+
"correct": 248,
|
| 120 |
+
"total": 273,
|
| 121 |
+
"accuracy": 0.9084249084249084
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"trial": 4,
|
| 125 |
+
"correct": 239,
|
| 126 |
+
"total": 273,
|
| 127 |
+
"accuracy": 0.8754578754578755
|
| 128 |
+
}
|
| 129 |
+
],
|
| 130 |
+
"mean_accuracy": 0.8857142857142858,
|
| 131 |
+
"best_accuracy": 0.9084249084249084,
|
| 132 |
+
"pipeline_seconds": 485.4800155526027
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"name": "current-ratio-step-600",
|
| 136 |
+
"trials": [
|
| 137 |
+
{
|
| 138 |
+
"trial": 0,
|
| 139 |
+
"correct": 245,
|
| 140 |
+
"total": 273,
|
| 141 |
+
"accuracy": 0.8974358974358975
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"trial": 1,
|
| 145 |
+
"correct": 244,
|
| 146 |
+
"total": 273,
|
| 147 |
+
"accuracy": 0.8937728937728938
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"trial": 2,
|
| 151 |
+
"correct": 240,
|
| 152 |
+
"total": 273,
|
| 153 |
+
"accuracy": 0.8791208791208791
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"trial": 3,
|
| 157 |
+
"correct": 248,
|
| 158 |
+
"total": 273,
|
| 159 |
+
"accuracy": 0.9084249084249084
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"trial": 4,
|
| 163 |
+
"correct": 241,
|
| 164 |
+
"total": 273,
|
| 165 |
+
"accuracy": 0.8827838827838828
|
| 166 |
+
}
|
| 167 |
+
],
|
| 168 |
+
"mean_accuracy": 0.8923076923076924,
|
| 169 |
+
"best_accuracy": 0.9084249084249084,
|
| 170 |
+
"pipeline_seconds": 517.7519376315176
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"name": "current-ratio-step-800",
|
| 174 |
+
"trials": [
|
| 175 |
+
{
|
| 176 |
+
"trial": 0,
|
| 177 |
+
"correct": 243,
|
| 178 |
+
"total": 273,
|
| 179 |
+
"accuracy": 0.8901098901098901
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"trial": 1,
|
| 183 |
+
"correct": 244,
|
| 184 |
+
"total": 273,
|
| 185 |
+
"accuracy": 0.8937728937728938
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"trial": 2,
|
| 189 |
+
"correct": 238,
|
| 190 |
+
"total": 273,
|
| 191 |
+
"accuracy": 0.8717948717948718
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"trial": 3,
|
| 195 |
+
"correct": 247,
|
| 196 |
+
"total": 273,
|
| 197 |
+
"accuracy": 0.9047619047619048
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"trial": 4,
|
| 201 |
+
"correct": 239,
|
| 202 |
+
"total": 273,
|
| 203 |
+
"accuracy": 0.8754578754578755
|
| 204 |
+
}
|
| 205 |
+
],
|
| 206 |
+
"mean_accuracy": 0.8871794871794872,
|
| 207 |
+
"best_accuracy": 0.9047619047619048,
|
| 208 |
+
"pipeline_seconds": 487.93090889789164
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"name": "current-ratio-step-1000",
|
| 212 |
+
"trials": [
|
| 213 |
+
{
|
| 214 |
+
"trial": 0,
|
| 215 |
+
"correct": 244,
|
| 216 |
+
"total": 273,
|
| 217 |
+
"accuracy": 0.8937728937728938
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"trial": 1,
|
| 221 |
+
"correct": 244,
|
| 222 |
+
"total": 273,
|
| 223 |
+
"accuracy": 0.8937728937728938
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"trial": 2,
|
| 227 |
+
"correct": 240,
|
| 228 |
+
"total": 273,
|
| 229 |
+
"accuracy": 0.8791208791208791
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"trial": 3,
|
| 233 |
+
"correct": 249,
|
| 234 |
+
"total": 273,
|
| 235 |
+
"accuracy": 0.9120879120879121
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"trial": 4,
|
| 239 |
+
"correct": 241,
|
| 240 |
+
"total": 273,
|
| 241 |
+
"accuracy": 0.8827838827838828
|
| 242 |
+
}
|
| 243 |
+
],
|
| 244 |
+
"mean_accuracy": 0.8923076923076924,
|
| 245 |
+
"best_accuracy": 0.9120879120879121,
|
| 246 |
+
"pipeline_seconds": 443.16614191699773
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"name": "incumbent",
|
| 250 |
+
"trials": [
|
| 251 |
+
{
|
| 252 |
+
"trial": 0,
|
| 253 |
+
"correct": 243,
|
| 254 |
+
"total": 273,
|
| 255 |
+
"accuracy": 0.8901098901098901
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"trial": 1,
|
| 259 |
+
"correct": 243,
|
| 260 |
+
"total": 273,
|
| 261 |
+
"accuracy": 0.8901098901098901
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"trial": 2,
|
| 265 |
+
"correct": 239,
|
| 266 |
+
"total": 273,
|
| 267 |
+
"accuracy": 0.8754578754578755
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"trial": 3,
|
| 271 |
+
"correct": 250,
|
| 272 |
+
"total": 273,
|
| 273 |
+
"accuracy": 0.9157509157509157
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"trial": 4,
|
| 277 |
+
"correct": 240,
|
| 278 |
+
"total": 273,
|
| 279 |
+
"accuracy": 0.8791208791208791
|
| 280 |
+
}
|
| 281 |
+
],
|
| 282 |
+
"mean_accuracy": 0.8901098901098902,
|
| 283 |
+
"best_accuracy": 0.9157509157509157,
|
| 284 |
+
"pipeline_seconds": 463.15512090176344
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"name": "lower-lr-step-200",
|
| 288 |
+
"trials": [
|
| 289 |
+
{
|
| 290 |
+
"trial": 0,
|
| 291 |
+
"correct": 237,
|
| 292 |
+
"total": 273,
|
| 293 |
+
"accuracy": 0.8681318681318682
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"trial": 1,
|
| 297 |
+
"correct": 237,
|
| 298 |
+
"total": 273,
|
| 299 |
+
"accuracy": 0.8681318681318682
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"trial": 2,
|
| 303 |
+
"correct": 228,
|
| 304 |
+
"total": 273,
|
| 305 |
+
"accuracy": 0.8351648351648352
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"trial": 3,
|
| 309 |
+
"correct": 246,
|
| 310 |
+
"total": 273,
|
| 311 |
+
"accuracy": 0.9010989010989011
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"trial": 4,
|
| 315 |
+
"correct": 238,
|
| 316 |
+
"total": 273,
|
| 317 |
+
"accuracy": 0.8717948717948718
|
| 318 |
+
}
|
| 319 |
+
],
|
| 320 |
+
"mean_accuracy": 0.868864468864469,
|
| 321 |
+
"best_accuracy": 0.9010989010989011,
|
| 322 |
+
"pipeline_seconds": 455.8428437123075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"name": "lower-lr-step-400",
|
| 326 |
+
"trials": [
|
| 327 |
+
{
|
| 328 |
+
"trial": 0,
|
| 329 |
+
"correct": 243,
|
| 330 |
+
"total": 273,
|
| 331 |
+
"accuracy": 0.8901098901098901
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"trial": 1,
|
| 335 |
+
"correct": 239,
|
| 336 |
+
"total": 273,
|
| 337 |
+
"accuracy": 0.8754578754578755
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"trial": 2,
|
| 341 |
+
"correct": 234,
|
| 342 |
+
"total": 273,
|
| 343 |
+
"accuracy": 0.8571428571428571
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"trial": 3,
|
| 347 |
+
"correct": 249,
|
| 348 |
+
"total": 273,
|
| 349 |
+
"accuracy": 0.9120879120879121
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"trial": 4,
|
| 353 |
+
"correct": 236,
|
| 354 |
+
"total": 273,
|
| 355 |
+
"accuracy": 0.8644688644688645
|
| 356 |
+
}
|
| 357 |
+
],
|
| 358 |
+
"mean_accuracy": 0.87985347985348,
|
| 359 |
+
"best_accuracy": 0.9120879120879121,
|
| 360 |
+
"pipeline_seconds": 443.79029164928943
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"name": "lower-lr-step-600",
|
| 364 |
+
"trials": [
|
| 365 |
+
{
|
| 366 |
+
"trial": 0,
|
| 367 |
+
"correct": 243,
|
| 368 |
+
"total": 273,
|
| 369 |
+
"accuracy": 0.8901098901098901
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"trial": 1,
|
| 373 |
+
"correct": 238,
|
| 374 |
+
"total": 273,
|
| 375 |
+
"accuracy": 0.8717948717948718
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"trial": 2,
|
| 379 |
+
"correct": 234,
|
| 380 |
+
"total": 273,
|
| 381 |
+
"accuracy": 0.8571428571428571
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"trial": 3,
|
| 385 |
+
"correct": 248,
|
| 386 |
+
"total": 273,
|
| 387 |
+
"accuracy": 0.9084249084249084
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"trial": 4,
|
| 391 |
+
"correct": 239,
|
| 392 |
+
"total": 273,
|
| 393 |
+
"accuracy": 0.8754578754578755
|
| 394 |
+
}
|
| 395 |
+
],
|
| 396 |
+
"mean_accuracy": 0.8805860805860807,
|
| 397 |
+
"best_accuracy": 0.9084249084249084,
|
| 398 |
+
"pipeline_seconds": 449.5708834640682
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"name": "lower-lr-step-800",
|
| 402 |
+
"trials": [
|
| 403 |
+
{
|
| 404 |
+
"trial": 0,
|
| 405 |
+
"correct": 243,
|
| 406 |
+
"total": 273,
|
| 407 |
+
"accuracy": 0.8901098901098901
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"trial": 1,
|
| 411 |
+
"correct": 245,
|
| 412 |
+
"total": 273,
|
| 413 |
+
"accuracy": 0.8974358974358975
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"trial": 2,
|
| 417 |
+
"correct": 235,
|
| 418 |
+
"total": 273,
|
| 419 |
+
"accuracy": 0.8608058608058609
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"trial": 3,
|
| 423 |
+
"correct": 250,
|
| 424 |
+
"total": 273,
|
| 425 |
+
"accuracy": 0.9157509157509157
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"trial": 4,
|
| 429 |
+
"correct": 237,
|
| 430 |
+
"total": 273,
|
| 431 |
+
"accuracy": 0.8681318681318682
|
| 432 |
+
}
|
| 433 |
+
],
|
| 434 |
+
"mean_accuracy": 0.8864468864468865,
|
| 435 |
+
"best_accuracy": 0.9157509157509157,
|
| 436 |
+
"pipeline_seconds": 440.65121043566614
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"name": "lower-lr-step-1000",
|
| 440 |
+
"trials": [
|
| 441 |
+
{
|
| 442 |
+
"trial": 0,
|
| 443 |
+
"correct": 244,
|
| 444 |
+
"total": 273,
|
| 445 |
+
"accuracy": 0.8937728937728938
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"trial": 1,
|
| 449 |
+
"correct": 242,
|
| 450 |
+
"total": 273,
|
| 451 |
+
"accuracy": 0.8864468864468864
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"trial": 2,
|
| 455 |
+
"correct": 235,
|
| 456 |
+
"total": 273,
|
| 457 |
+
"accuracy": 0.8608058608058609
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"trial": 3,
|
| 461 |
+
"correct": 249,
|
| 462 |
+
"total": 273,
|
| 463 |
+
"accuracy": 0.9120879120879121
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"trial": 4,
|
| 467 |
+
"correct": 237,
|
| 468 |
+
"total": 273,
|
| 469 |
+
"accuracy": 0.8681318681318682
|
| 470 |
+
}
|
| 471 |
+
],
|
| 472 |
+
"mean_accuracy": 0.8842490842490843,
|
| 473 |
+
"best_accuracy": 0.9120879120879121,
|
| 474 |
+
"pipeline_seconds": 487.40875728148967
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"name": "more-grpo-step-200",
|
| 478 |
+
"trials": [
|
| 479 |
+
{
|
| 480 |
+
"trial": 0,
|
| 481 |
+
"correct": 238,
|
| 482 |
+
"total": 273,
|
| 483 |
+
"accuracy": 0.8717948717948718
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"trial": 1,
|
| 487 |
+
"correct": 236,
|
| 488 |
+
"total": 273,
|
| 489 |
+
"accuracy": 0.8644688644688645
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"trial": 2,
|
| 493 |
+
"correct": 230,
|
| 494 |
+
"total": 273,
|
| 495 |
+
"accuracy": 0.8424908424908425
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"trial": 3,
|
| 499 |
+
"correct": 247,
|
| 500 |
+
"total": 273,
|
| 501 |
+
"accuracy": 0.9047619047619048
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"trial": 4,
|
| 505 |
+
"correct": 232,
|
| 506 |
+
"total": 273,
|
| 507 |
+
"accuracy": 0.8498168498168498
|
| 508 |
+
}
|
| 509 |
+
],
|
| 510 |
+
"mean_accuracy": 0.8666666666666666,
|
| 511 |
+
"best_accuracy": 0.9047619047619048,
|
| 512 |
+
"pipeline_seconds": 501.1349828578532
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"name": "more-grpo-step-400",
|
| 516 |
+
"trials": [
|
| 517 |
+
{
|
| 518 |
+
"trial": 0,
|
| 519 |
+
"correct": 238,
|
| 520 |
+
"total": 273,
|
| 521 |
+
"accuracy": 0.8717948717948718
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"trial": 1,
|
| 525 |
+
"correct": 236,
|
| 526 |
+
"total": 273,
|
| 527 |
+
"accuracy": 0.8644688644688645
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"trial": 2,
|
| 531 |
+
"correct": 232,
|
| 532 |
+
"total": 273,
|
| 533 |
+
"accuracy": 0.8498168498168498
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"trial": 3,
|
| 537 |
+
"correct": 248,
|
| 538 |
+
"total": 273,
|
| 539 |
+
"accuracy": 0.9084249084249084
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"trial": 4,
|
| 543 |
+
"correct": 236,
|
| 544 |
+
"total": 273,
|
| 545 |
+
"accuracy": 0.8644688644688645
|
| 546 |
+
}
|
| 547 |
+
],
|
| 548 |
+
"mean_accuracy": 0.8717948717948719,
|
| 549 |
+
"best_accuracy": 0.9084249084249084,
|
| 550 |
+
"pipeline_seconds": 504.87802879791707
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"name": "more-grpo-step-600",
|
| 554 |
+
"trials": [
|
| 555 |
+
{
|
| 556 |
+
"trial": 0,
|
| 557 |
+
"correct": 242,
|
| 558 |
+
"total": 273,
|
| 559 |
+
"accuracy": 0.8864468864468864
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"trial": 1,
|
| 563 |
+
"correct": 242,
|
| 564 |
+
"total": 273,
|
| 565 |
+
"accuracy": 0.8864468864468864
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"trial": 2,
|
| 569 |
+
"correct": 236,
|
| 570 |
+
"total": 273,
|
| 571 |
+
"accuracy": 0.8644688644688645
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"trial": 3,
|
| 575 |
+
"correct": 248,
|
| 576 |
+
"total": 273,
|
| 577 |
+
"accuracy": 0.9084249084249084
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"trial": 4,
|
| 581 |
+
"correct": 239,
|
| 582 |
+
"total": 273,
|
| 583 |
+
"accuracy": 0.8754578754578755
|
| 584 |
+
}
|
| 585 |
+
],
|
| 586 |
+
"mean_accuracy": 0.8842490842490843,
|
| 587 |
+
"best_accuracy": 0.9084249084249084,
|
| 588 |
+
"pipeline_seconds": 482.52498442307115
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"name": "more-grpo-step-800",
|
| 592 |
+
"trials": [
|
| 593 |
+
{
|
| 594 |
+
"trial": 0,
|
| 595 |
+
"correct": 241,
|
| 596 |
+
"total": 273,
|
| 597 |
+
"accuracy": 0.8827838827838828
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"trial": 1,
|
| 601 |
+
"correct": 244,
|
| 602 |
+
"total": 273,
|
| 603 |
+
"accuracy": 0.8937728937728938
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"trial": 2,
|
| 607 |
+
"correct": 236,
|
| 608 |
+
"total": 273,
|
| 609 |
+
"accuracy": 0.8644688644688645
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"trial": 3,
|
| 613 |
+
"correct": 249,
|
| 614 |
+
"total": 273,
|
| 615 |
+
"accuracy": 0.9120879120879121
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"trial": 4,
|
| 619 |
+
"correct": 239,
|
| 620 |
+
"total": 273,
|
| 621 |
+
"accuracy": 0.8754578754578755
|
| 622 |
+
}
|
| 623 |
+
],
|
| 624 |
+
"mean_accuracy": 0.8857142857142858,
|
| 625 |
+
"best_accuracy": 0.9120879120879121,
|
| 626 |
+
"pipeline_seconds": 462.4395873360336
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"name": "more-grpo-step-1000",
|
| 630 |
+
"trials": [
|
| 631 |
+
{
|
| 632 |
+
"trial": 0,
|
| 633 |
+
"correct": 243,
|
| 634 |
+
"total": 273,
|
| 635 |
+
"accuracy": 0.8901098901098901
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"trial": 1,
|
| 639 |
+
"correct": 245,
|
| 640 |
+
"total": 273,
|
| 641 |
+
"accuracy": 0.8974358974358975
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"trial": 2,
|
| 645 |
+
"correct": 236,
|
| 646 |
+
"total": 273,
|
| 647 |
+
"accuracy": 0.8644688644688645
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"trial": 3,
|
| 651 |
+
"correct": 250,
|
| 652 |
+
"total": 273,
|
| 653 |
+
"accuracy": 0.9157509157509157
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"trial": 4,
|
| 657 |
+
"correct": 239,
|
| 658 |
+
"total": 273,
|
| 659 |
+
"accuracy": 0.8754578754578755
|
| 660 |
+
}
|
| 661 |
+
],
|
| 662 |
+
"mean_accuracy": 0.8886446886446887,
|
| 663 |
+
"best_accuracy": 0.9157509157509157,
|
| 664 |
+
"pipeline_seconds": 489.50726920738816
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"name": "more-opd-step-200",
|
| 668 |
+
"trials": [
|
| 669 |
+
{
|
| 670 |
+
"trial": 0,
|
| 671 |
+
"correct": 234,
|
| 672 |
+
"total": 273,
|
| 673 |
+
"accuracy": 0.8571428571428571
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"trial": 1,
|
| 677 |
+
"correct": 235,
|
| 678 |
+
"total": 273,
|
| 679 |
+
"accuracy": 0.8608058608058609
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"trial": 2,
|
| 683 |
+
"correct": 227,
|
| 684 |
+
"total": 273,
|
| 685 |
+
"accuracy": 0.8315018315018315
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"trial": 3,
|
| 689 |
+
"correct": 247,
|
| 690 |
+
"total": 273,
|
| 691 |
+
"accuracy": 0.9047619047619048
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"trial": 4,
|
| 695 |
+
"correct": 233,
|
| 696 |
+
"total": 273,
|
| 697 |
+
"accuracy": 0.8534798534798534
|
| 698 |
+
}
|
| 699 |
+
],
|
| 700 |
+
"mean_accuracy": 0.8615384615384615,
|
| 701 |
+
"best_accuracy": 0.9047619047619048,
|
| 702 |
+
"pipeline_seconds": 458.6599548086524
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"name": "more-opd-step-400",
|
| 706 |
+
"trials": [
|
| 707 |
+
{
|
| 708 |
+
"trial": 0,
|
| 709 |
+
"correct": 240,
|
| 710 |
+
"total": 273,
|
| 711 |
+
"accuracy": 0.8791208791208791
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"trial": 1,
|
| 715 |
+
"correct": 242,
|
| 716 |
+
"total": 273,
|
| 717 |
+
"accuracy": 0.8864468864468864
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"trial": 2,
|
| 721 |
+
"correct": 232,
|
| 722 |
+
"total": 273,
|
| 723 |
+
"accuracy": 0.8498168498168498
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"trial": 3,
|
| 727 |
+
"correct": 248,
|
| 728 |
+
"total": 273,
|
| 729 |
+
"accuracy": 0.9084249084249084
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"trial": 4,
|
| 733 |
+
"correct": 235,
|
| 734 |
+
"total": 273,
|
| 735 |
+
"accuracy": 0.8608058608058609
|
| 736 |
+
}
|
| 737 |
+
],
|
| 738 |
+
"mean_accuracy": 0.876923076923077,
|
| 739 |
+
"best_accuracy": 0.9084249084249084,
|
| 740 |
+
"pipeline_seconds": 442.6371235605329
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"name": "more-opd-step-600",
|
| 744 |
+
"trials": [
|
| 745 |
+
{
|
| 746 |
+
"trial": 0,
|
| 747 |
+
"correct": 242,
|
| 748 |
+
"total": 273,
|
| 749 |
+
"accuracy": 0.8864468864468864
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"trial": 1,
|
| 753 |
+
"correct": 241,
|
| 754 |
+
"total": 273,
|
| 755 |
+
"accuracy": 0.8827838827838828
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"trial": 2,
|
| 759 |
+
"correct": 235,
|
| 760 |
+
"total": 273,
|
| 761 |
+
"accuracy": 0.8608058608058609
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"trial": 3,
|
| 765 |
+
"correct": 249,
|
| 766 |
+
"total": 273,
|
| 767 |
+
"accuracy": 0.9120879120879121
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"trial": 4,
|
| 771 |
+
"correct": 237,
|
| 772 |
+
"total": 273,
|
| 773 |
+
"accuracy": 0.8681318681318682
|
| 774 |
+
}
|
| 775 |
+
],
|
| 776 |
+
"mean_accuracy": 0.8820512820512821,
|
| 777 |
+
"best_accuracy": 0.9120879120879121,
|
| 778 |
+
"pipeline_seconds": 474.46449717786163
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"name": "more-opd-step-800",
|
| 782 |
+
"trials": [
|
| 783 |
+
{
|
| 784 |
+
"trial": 0,
|
| 785 |
+
"correct": 243,
|
| 786 |
+
"total": 273,
|
| 787 |
+
"accuracy": 0.8901098901098901
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"trial": 1,
|
| 791 |
+
"correct": 244,
|
| 792 |
+
"total": 273,
|
| 793 |
+
"accuracy": 0.8937728937728938
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"trial": 2,
|
| 797 |
+
"correct": 235,
|
| 798 |
+
"total": 273,
|
| 799 |
+
"accuracy": 0.8608058608058609
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"trial": 3,
|
| 803 |
+
"correct": 249,
|
| 804 |
+
"total": 273,
|
| 805 |
+
"accuracy": 0.9120879120879121
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"trial": 4,
|
| 809 |
+
"correct": 239,
|
| 810 |
+
"total": 273,
|
| 811 |
+
"accuracy": 0.8754578754578755
|
| 812 |
+
}
|
| 813 |
+
],
|
| 814 |
+
"mean_accuracy": 0.8864468864468865,
|
| 815 |
+
"best_accuracy": 0.9120879120879121,
|
| 816 |
+
"pipeline_seconds": 446.2496670978144
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"name": "more-opd-step-1000",
|
| 820 |
+
"trials": [
|
| 821 |
+
{
|
| 822 |
+
"trial": 0,
|
| 823 |
+
"correct": 241,
|
| 824 |
+
"total": 273,
|
| 825 |
+
"accuracy": 0.8827838827838828
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"trial": 1,
|
| 829 |
+
"correct": 244,
|
| 830 |
+
"total": 273,
|
| 831 |
+
"accuracy": 0.8937728937728938
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"trial": 2,
|
| 835 |
+
"correct": 237,
|
| 836 |
+
"total": 273,
|
| 837 |
+
"accuracy": 0.8681318681318682
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"trial": 3,
|
| 841 |
+
"correct": 250,
|
| 842 |
+
"total": 273,
|
| 843 |
+
"accuracy": 0.9157509157509157
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"trial": 4,
|
| 847 |
+
"correct": 241,
|
| 848 |
+
"total": 273,
|
| 849 |
+
"accuracy": 0.8827838827838828
|
| 850 |
+
}
|
| 851 |
+
],
|
| 852 |
+
"mean_accuracy": 0.8886446886446887,
|
| 853 |
+
"best_accuracy": 0.9157509157509157,
|
| 854 |
+
"pipeline_seconds": 462.7830137303099
|
| 855 |
+
}
|
| 856 |
+
],
|
| 857 |
+
"selected_test": {
|
| 858 |
+
"name": "selected",
|
| 859 |
+
"trials": [
|
| 860 |
+
{
|
| 861 |
+
"trial": 0,
|
| 862 |
+
"correct": 203,
|
| 863 |
+
"total": 233,
|
| 864 |
+
"accuracy": 0.871244635193133
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"trial": 1,
|
| 868 |
+
"correct": 203,
|
| 869 |
+
"total": 233,
|
| 870 |
+
"accuracy": 0.871244635193133
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"trial": 2,
|
| 874 |
+
"correct": 208,
|
| 875 |
+
"total": 233,
|
| 876 |
+
"accuracy": 0.8927038626609443
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"trial": 3,
|
| 880 |
+
"correct": 205,
|
| 881 |
+
"total": 233,
|
| 882 |
+
"accuracy": 0.8798283261802575
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"trial": 4,
|
| 886 |
+
"correct": 202,
|
| 887 |
+
"total": 233,
|
| 888 |
+
"accuracy": 0.8669527896995708
|
| 889 |
+
}
|
| 890 |
+
],
|
| 891 |
+
"mean_accuracy": 0.8763948497854077,
|
| 892 |
+
"best_accuracy": 0.8927038626609443,
|
| 893 |
+
"pipeline_seconds": 370.2440517041832
|
| 894 |
+
},
|
| 895 |
+
"train_seconds": {
|
| 896 |
+
"current-ratio": 1312.184554793872,
|
| 897 |
+
"lower-lr": 1311.2459088759497,
|
| 898 |
+
"more-grpo": 1308.0780487433076,
|
| 899 |
+
"more-opd": 1309.390459395945
|
| 900 |
+
},
|
| 901 |
+
"previous_revision": "71a718d07b7c61b3fdf53509557e05571368ba8d"
|
| 902 |
+
}
|
personamem-128k/qwen2.5-3b/reader/adapter_config.json
CHANGED
|
@@ -25,13 +25,13 @@
|
|
| 25 |
"rank_pattern": {},
|
| 26 |
"revision": "aa8e72537993ba99e69dfaafa59ed015b17504d1",
|
| 27 |
"target_modules": [
|
| 28 |
-
"gate_proj",
|
| 29 |
"down_proj",
|
| 30 |
-
"
|
| 31 |
-
"o_proj",
|
| 32 |
-
"v_proj",
|
| 33 |
"q_proj",
|
| 34 |
-
"
|
|
|
|
|
|
|
|
|
|
| 35 |
],
|
| 36 |
"target_parameters": null,
|
| 37 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 25 |
"rank_pattern": {},
|
| 26 |
"revision": "aa8e72537993ba99e69dfaafa59ed015b17504d1",
|
| 27 |
"target_modules": [
|
|
|
|
| 28 |
"down_proj",
|
| 29 |
+
"k_proj",
|
|
|
|
|
|
|
| 30 |
"q_proj",
|
| 31 |
+
"o_proj",
|
| 32 |
+
"gate_proj",
|
| 33 |
+
"up_proj",
|
| 34 |
+
"v_proj"
|
| 35 |
],
|
| 36 |
"target_parameters": null,
|
| 37 |
"task_type": "CAUSAL_LM",
|
personamem-128k/qwen2.5-3b/reader/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 119801528
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:727d8e4de0ad4ffe224124e3d3a515ae0f1a8c919f3828c840a0d7e31b512006
|
| 3 |
size 119801528
|