End of training
Browse files- README.md +2 -1
- all_results.json +16 -0
- train_results.json +16 -0
- trainer_state.json +597 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen3-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: taskmaster2-3k-traces
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# taskmaster2-3k-traces
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen3-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: taskmaster2-3k-traces
|
|
|
|
| 16 |
|
| 17 |
# taskmaster2-3k-traces
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the DCAgent/taskmaster2-3k-traces dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.955181724833952,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 155.84360988986572,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.20247647166252136,
|
| 7 |
+
"mfu_percent": 0.35018952118967855,
|
| 8 |
+
"mfu_percent_theoretical": 11.013682677729026,
|
| 9 |
+
"total_flos": 4.268418036190413e+17,
|
| 10 |
+
"train_loss": 0.4559480676651001,
|
| 11 |
+
"train_runtime": 5383.7809,
|
| 12 |
+
"train_samples_per_second": 2.935,
|
| 13 |
+
"train_steps_per_second": 0.046,
|
| 14 |
+
"valid_targets_mean": 3850.3,
|
| 15 |
+
"valid_targets_min": 771
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.955181724833952,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 155.84360988986572,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.20247647166252136,
|
| 7 |
+
"mfu_percent": 0.35018952118967855,
|
| 8 |
+
"mfu_percent_theoretical": 11.013682677729026,
|
| 9 |
+
"total_flos": 4.268418036190413e+17,
|
| 10 |
+
"train_loss": 0.4559480676651001,
|
| 11 |
+
"train_runtime": 5383.7809,
|
| 12 |
+
"train_samples_per_second": 2.935,
|
| 13 |
+
"train_steps_per_second": 0.046,
|
| 14 |
+
"valid_targets_mean": 3850.3,
|
| 15 |
+
"valid_targets_min": 771
|
| 16 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,597 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 250,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.10101010101010101,
|
| 14 |
+
"grad_norm": 5.43838668916967,
|
| 15 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 16 |
+
"loss": 0.768,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.20029328763484955,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 3894.4,
|
| 21 |
+
"valid_targets_min": 1551
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.20202020202020202,
|
| 25 |
+
"grad_norm": 1.8173704060845222,
|
| 26 |
+
"learning_rate": 1.4400000000000001e-05,
|
| 27 |
+
"loss": 0.6809,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.12393468618392944,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 3313.9,
|
| 32 |
+
"valid_targets_min": 658
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.30303030303030304,
|
| 36 |
+
"grad_norm": 0.8994415488525478,
|
| 37 |
+
"learning_rate": 2.2400000000000002e-05,
|
| 38 |
+
"loss": 0.6367,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.15267133712768555,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 3214.2,
|
| 43 |
+
"valid_targets_min": 748
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.40404040404040403,
|
| 47 |
+
"grad_norm": 0.5231491285791463,
|
| 48 |
+
"learning_rate": 3.0400000000000004e-05,
|
| 49 |
+
"loss": 0.603,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.1451815664768219,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 4062.5,
|
| 54 |
+
"valid_targets_min": 752
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.5050505050505051,
|
| 58 |
+
"grad_norm": 0.48610136532802667,
|
| 59 |
+
"learning_rate": 3.8400000000000005e-05,
|
| 60 |
+
"loss": 0.5739,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.15977895259857178,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 4920.6,
|
| 65 |
+
"valid_targets_min": 689
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.6060606060606061,
|
| 69 |
+
"grad_norm": 0.3657562059313739,
|
| 70 |
+
"learning_rate": 3.9968815283639625e-05,
|
| 71 |
+
"loss": 0.531,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.12183522433042526,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 4170.1,
|
| 76 |
+
"valid_targets_min": 848
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.7070707070707071,
|
| 80 |
+
"grad_norm": 0.3117441303599555,
|
| 81 |
+
"learning_rate": 3.9842294026289565e-05,
|
| 82 |
+
"loss": 0.5137,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.11094912886619568,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 4246.0,
|
| 87 |
+
"valid_targets_min": 626
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8080808080808081,
|
| 91 |
+
"grad_norm": 0.31427213112004465,
|
| 92 |
+
"learning_rate": 3.9619103106983835e-05,
|
| 93 |
+
"loss": 0.5033,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.11524395644664764,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 3335.3,
|
| 98 |
+
"valid_targets_min": 589
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.9090909090909091,
|
| 102 |
+
"grad_norm": 0.27843711420334,
|
| 103 |
+
"learning_rate": 3.930032988944623e-05,
|
| 104 |
+
"loss": 0.4921,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.12719261646270752,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 4673.1,
|
| 109 |
+
"valid_targets_min": 842
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.0,
|
| 113 |
+
"grad_norm": 0.35336756908591643,
|
| 114 |
+
"learning_rate": 3.888752740474962e-05,
|
| 115 |
+
"loss": 0.4725,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.23522385954856873,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 4525.1,
|
| 120 |
+
"valid_targets_min": 618
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.101010101010101,
|
| 124 |
+
"grad_norm": 0.2618047514835116,
|
| 125 |
+
"learning_rate": 3.838270678510469e-05,
|
| 126 |
+
"loss": 0.4631,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.0997854694724083,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 3254.4,
|
| 131 |
+
"valid_targets_min": 620
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.202020202020202,
|
| 135 |
+
"grad_norm": 0.2596681731982267,
|
| 136 |
+
"learning_rate": 3.778832746582596e-05,
|
| 137 |
+
"loss": 0.4695,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.15060186386108398,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 5336.2,
|
| 142 |
+
"valid_targets_min": 994
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.303030303030303,
|
| 146 |
+
"grad_norm": 0.26134191299357834,
|
| 147 |
+
"learning_rate": 3.710728520321014e-05,
|
| 148 |
+
"loss": 0.4622,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.10998662561178207,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 3828.9,
|
| 153 |
+
"valid_targets_min": 813
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.404040404040404,
|
| 157 |
+
"grad_norm": 0.2505910129541225,
|
| 158 |
+
"learning_rate": 3.634289796670257e-05,
|
| 159 |
+
"loss": 0.4639,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.11412525177001953,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 4118.6,
|
| 164 |
+
"valid_targets_min": 589
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.5050505050505052,
|
| 168 |
+
"grad_norm": 0.28469306377718256,
|
| 169 |
+
"learning_rate": 3.549888977408359e-05,
|
| 170 |
+
"loss": 0.4547,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.10224676132202148,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 3621.8,
|
| 175 |
+
"valid_targets_min": 807
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.606060606060606,
|
| 179 |
+
"grad_norm": 0.26506094775038846,
|
| 180 |
+
"learning_rate": 3.457937254842823e-05,
|
| 181 |
+
"loss": 0.4478,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.1216074526309967,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 4121.2,
|
| 186 |
+
"valid_targets_min": 900
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.7070707070707072,
|
| 190 |
+
"grad_norm": 0.27422041538604486,
|
| 191 |
+
"learning_rate": 3.3588826085230336e-05,
|
| 192 |
+
"loss": 0.4405,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.09807954728603363,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 3189.9,
|
| 197 |
+
"valid_targets_min": 777
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.808080808080808,
|
| 201 |
+
"grad_norm": 0.25008604850493343,
|
| 202 |
+
"learning_rate": 3.253207622728921e-05,
|
| 203 |
+
"loss": 0.441,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.11396118998527527,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 4577.4,
|
| 208 |
+
"valid_targets_min": 875
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 1.9090909090909092,
|
| 212 |
+
"grad_norm": 0.26410650818418535,
|
| 213 |
+
"learning_rate": 3.141427135368864e-05,
|
| 214 |
+
"loss": 0.4447,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.10761921852827072,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 3803.5,
|
| 219 |
+
"valid_targets_min": 613
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.0,
|
| 223 |
+
"grad_norm": 0.37011104936880673,
|
| 224 |
+
"learning_rate": 3.024085729741143e-05,
|
| 225 |
+
"loss": 0.4481,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.254130095243454,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 4326.2,
|
| 230 |
+
"valid_targets_min": 550
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.101010101010101,
|
| 234 |
+
"grad_norm": 0.2633444829921788,
|
| 235 |
+
"learning_rate": 2.9017550813788616e-05,
|
| 236 |
+
"loss": 0.4286,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.10730044543743134,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 4681.4,
|
| 241 |
+
"valid_targets_min": 808
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.202020202020202,
|
| 245 |
+
"grad_norm": 0.27947047290967925,
|
| 246 |
+
"learning_rate": 2.7750311729042062e-05,
|
| 247 |
+
"loss": 0.4249,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.11274366080760956,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 4363.5,
|
| 252 |
+
"valid_targets_min": 721
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 2.303030303030303,
|
| 256 |
+
"grad_norm": 0.26878537952096837,
|
| 257 |
+
"learning_rate": 2.6445313904610227e-05,
|
| 258 |
+
"loss": 0.4319,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.10825060307979584,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 3570.9,
|
| 263 |
+
"valid_targets_min": 740
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.404040404040404,
|
| 267 |
+
"grad_norm": 0.27801838465637635,
|
| 268 |
+
"learning_rate": 2.510891515871581e-05,
|
| 269 |
+
"loss": 0.4335,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.09939642250537872,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 3389.2,
|
| 274 |
+
"valid_targets_min": 712
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.505050505050505,
|
| 278 |
+
"grad_norm": 0.26781305081846457,
|
| 279 |
+
"learning_rate": 2.37476262917145e-05,
|
| 280 |
+
"loss": 0.4247,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.109991654753685,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 4646.0,
|
| 285 |
+
"valid_targets_min": 748
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.606060606060606,
|
| 289 |
+
"grad_norm": 0.26237181084548494,
|
| 290 |
+
"learning_rate": 2.2368079366130028e-05,
|
| 291 |
+
"loss": 0.4187,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.11524119973182678,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 4932.1,
|
| 296 |
+
"valid_targets_min": 845
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.707070707070707,
|
| 300 |
+
"grad_norm": 0.2585522405239844,
|
| 301 |
+
"learning_rate": 2.097699539591227e-05,
|
| 302 |
+
"loss": 0.4274,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.08941829204559326,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 3169.6,
|
| 307 |
+
"valid_targets_min": 832
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.808080808080808,
|
| 311 |
+
"grad_norm": 0.26016350721342923,
|
| 312 |
+
"learning_rate": 1.9581151602332865e-05,
|
| 313 |
+
"loss": 0.4177,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.1133737787604332,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 5011.3,
|
| 318 |
+
"valid_targets_min": 1002
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.909090909090909,
|
| 322 |
+
"grad_norm": 0.27030558024007334,
|
| 323 |
+
"learning_rate": 1.8187348396044402e-05,
|
| 324 |
+
"loss": 0.4231,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.10732856392860413,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 4244.3,
|
| 329 |
+
"valid_targets_min": 849
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.0,
|
| 333 |
+
"grad_norm": 0.3695239483214702,
|
| 334 |
+
"learning_rate": 1.6802376246163307e-05,
|
| 335 |
+
"loss": 0.4307,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.2080117017030716,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 3466.7,
|
| 340 |
+
"valid_targets_min": 600
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 3.101010101010101,
|
| 344 |
+
"grad_norm": 0.2516175673157399,
|
| 345 |
+
"learning_rate": 1.5432982597786886e-05,
|
| 346 |
+
"loss": 0.4228,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.11081644892692566,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 4376.6,
|
| 351 |
+
"valid_targets_min": 1393
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.202020202020202,
|
| 355 |
+
"grad_norm": 0.25899041127457506,
|
| 356 |
+
"learning_rate": 1.4085838999119075e-05,
|
| 357 |
+
"loss": 0.4183,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.10279497504234314,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 3922.6,
|
| 362 |
+
"valid_targets_min": 1037
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.303030303030303,
|
| 366 |
+
"grad_norm": 0.2568302854985244,
|
| 367 |
+
"learning_rate": 1.2767508598358158e-05,
|
| 368 |
+
"loss": 0.4119,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.10400432348251343,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 4147.3,
|
| 373 |
+
"valid_targets_min": 840
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 3.404040404040404,
|
| 377 |
+
"grad_norm": 0.2575745984749214,
|
| 378 |
+
"learning_rate": 1.1484414168698547e-05,
|
| 379 |
+
"loss": 0.4109,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.1014273464679718,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 3870.5,
|
| 384 |
+
"valid_targets_min": 738
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 3.505050505050505,
|
| 388 |
+
"grad_norm": 0.2708052183398422,
|
| 389 |
+
"learning_rate": 1.0242806817225344e-05,
|
| 390 |
+
"loss": 0.4183,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.0942554920911789,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 4129.5,
|
| 395 |
+
"valid_targets_min": 691
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.606060606060606,
|
| 399 |
+
"grad_norm": 0.39589597035376406,
|
| 400 |
+
"learning_rate": 9.048735530148998e-06,
|
| 401 |
+
"loss": 0.4057,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.10414694994688034,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 3647.8,
|
| 406 |
+
"valid_targets_min": 1355
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.707070707070707,
|
| 410 |
+
"grad_norm": 0.2754352204699097,
|
| 411 |
+
"learning_rate": 7.908017702752504e-06,
|
| 412 |
+
"loss": 0.4167,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.11268728971481323,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 4025.8,
|
| 417 |
+
"valid_targets_min": 631
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 3.808080808080808,
|
| 421 |
+
"grad_norm": 0.26307739642578865,
|
| 422 |
+
"learning_rate": 6.826210797626389e-06,
|
| 423 |
+
"loss": 0.4158,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.08769197762012482,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 3541.8,
|
| 428 |
+
"valid_targets_min": 1132
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.909090909090909,
|
| 432 |
+
"grad_norm": 0.25965429680059315,
|
| 433 |
+
"learning_rate": 5.8085852692695864e-06,
|
| 434 |
+
"loss": 0.4088,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.10568307340145111,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 4072.8,
|
| 439 |
+
"valid_targets_min": 658
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 4.0,
|
| 443 |
+
"grad_norm": 0.37112181378749864,
|
| 444 |
+
"learning_rate": 4.8600988869648745e-06,
|
| 445 |
+
"loss": 0.4077,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.18167036771774292,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 3388.4,
|
| 450 |
+
"valid_targets_min": 757
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 4.101010101010101,
|
| 454 |
+
"grad_norm": 0.24190724490489265,
|
| 455 |
+
"learning_rate": 3.985372581025333e-06,
|
| 456 |
+
"loss": 0.3992,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.09219323098659515,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 3395.9,
|
| 461 |
+
"valid_targets_min": 1280
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 4.202020202020202,
|
| 465 |
+
"grad_norm": 0.26150260454768026,
|
| 466 |
+
"learning_rate": 3.1886679300863156e-06,
|
| 467 |
+
"loss": 0.407,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.09850673377513885,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 3216.6,
|
| 472 |
+
"valid_targets_min": 784
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.303030303030303,
|
| 476 |
+
"grad_norm": 0.26872823184659334,
|
| 477 |
+
"learning_rate": 2.473866399122733e-06,
|
| 478 |
+
"loss": 0.4133,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.10747270286083221,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 3487.2,
|
| 483 |
+
"valid_targets_min": 773
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 4.404040404040404,
|
| 487 |
+
"grad_norm": 0.2549088626047512,
|
| 488 |
+
"learning_rate": 1.8444504293418286e-06,
|
| 489 |
+
"loss": 0.4124,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.10171965509653091,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 3727.9,
|
| 494 |
+
"valid_targets_min": 928
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 4.505050505050505,
|
| 498 |
+
"grad_norm": 0.32301587125223463,
|
| 499 |
+
"learning_rate": 1.3034864720797112e-06,
|
| 500 |
+
"loss": 0.4125,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.11444351077079773,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 3867.6,
|
| 505 |
+
"valid_targets_min": 791
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 4.606060606060606,
|
| 509 |
+
"grad_norm": 0.24811664458418323,
|
| 510 |
+
"learning_rate": 8.536100493586552e-07,
|
| 511 |
+
"loss": 0.4023,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.0989425927400589,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 3649.8,
|
| 516 |
+
"valid_targets_min": 675
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.707070707070707,
|
| 520 |
+
"grad_norm": 0.24999205236156857,
|
| 521 |
+
"learning_rate": 4.970129138887347e-07,
|
| 522 |
+
"loss": 0.4078,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.09448540210723877,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 3779.3,
|
| 527 |
+
"valid_targets_min": 613
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 4.808080808080808,
|
| 531 |
+
"grad_norm": 0.2533525398749002,
|
| 532 |
+
"learning_rate": 2.3543237106894434e-07,
|
| 533 |
+
"loss": 0.4138,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.10717539489269257,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 3630.1,
|
| 538 |
+
"valid_targets_min": 1185
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.909090909090909,
|
| 542 |
+
"grad_norm": 0.2324383918083103,
|
| 543 |
+
"learning_rate": 7.01428150099126e-08,
|
| 544 |
+
"loss": 0.4099,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.09807487577199936,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 4431.1,
|
| 549 |
+
"valid_targets_min": 1588
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 5.0,
|
| 553 |
+
"grad_norm": 0.3564637724019629,
|
| 554 |
+
"learning_rate": 1.949519813915224e-09,
|
| 555 |
+
"loss": 0.4103,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.20247647166252136,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 3850.3,
|
| 560 |
+
"valid_targets_min": 771
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 5.0,
|
| 564 |
+
"loss_nan_ranks": 0,
|
| 565 |
+
"loss_rank_avg": 0.20247647166252136,
|
| 566 |
+
"step": 250,
|
| 567 |
+
"total_flos": 4.268418036190413e+17,
|
| 568 |
+
"train_loss": 0.4559480676651001,
|
| 569 |
+
"train_runtime": 5383.7809,
|
| 570 |
+
"train_samples_per_second": 2.935,
|
| 571 |
+
"train_steps_per_second": 0.046,
|
| 572 |
+
"valid_targets_mean": 3850.3,
|
| 573 |
+
"valid_targets_min": 771
|
| 574 |
+
}
|
| 575 |
+
],
|
| 576 |
+
"logging_steps": 5,
|
| 577 |
+
"max_steps": 250,
|
| 578 |
+
"num_input_tokens_seen": 0,
|
| 579 |
+
"num_train_epochs": 5,
|
| 580 |
+
"save_steps": 500,
|
| 581 |
+
"stateful_callbacks": {
|
| 582 |
+
"TrainerControl": {
|
| 583 |
+
"args": {
|
| 584 |
+
"should_epoch_stop": false,
|
| 585 |
+
"should_evaluate": false,
|
| 586 |
+
"should_log": false,
|
| 587 |
+
"should_save": false,
|
| 588 |
+
"should_training_stop": false
|
| 589 |
+
},
|
| 590 |
+
"attributes": {}
|
| 591 |
+
}
|
| 592 |
+
},
|
| 593 |
+
"total_flos": 4.268418036190413e+17,
|
| 594 |
+
"train_batch_size": 1,
|
| 595 |
+
"trial_name": null,
|
| 596 |
+
"trial_params": null
|
| 597 |
+
}
|
training_loss.png
ADDED
|