the-jb/tofu_Llama-3.2-3B-Instruct_forget10_GradDiff
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: the-jb/tofu_Llama-3.2-3B-Instruct_forget10_GradDiff
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-12 20:44:39 +08:00
commit 7f97a0382d
23 changed files with 42747 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

211
trainer_state.json Normal file
View File

@@ -0,0 +1,211 @@
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 9.24,
"eval_steps": 500,
"global_step": 120,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.4,
"grad_norm": 18.6333627466956,
"learning_rate": 3.3333333333333333e-06,
"loss": -0.028,
"step": 5
},
{
"epoch": 0.8,
"grad_norm": 16.032357759228358,
"learning_rate": 7.500000000000001e-06,
"loss": -0.064,
"step": 10
},
{
"epoch": 1.16,
"grad_norm": 34.99217902500999,
"learning_rate": 9.814814814814815e-06,
"loss": -0.3462,
"step": 15
},
{
"epoch": 1.56,
"grad_norm": 247.49125320648662,
"learning_rate": 9.351851851851854e-06,
"loss": -1.9847,
"step": 20
},
{
"epoch": 1.96,
"grad_norm": 1038.7636425274402,
"learning_rate": 8.888888888888888e-06,
"loss": -7.2508,
"step": 25
},
{
"epoch": 2.32,
"grad_norm": 3263.564556779214,
"learning_rate": 8.425925925925926e-06,
"loss": -12.6012,
"step": 30
},
{
"epoch": 2.7199999999999998,
"grad_norm": 5369.050190792846,
"learning_rate": 7.962962962962963e-06,
"loss": -85.1783,
"step": 35
},
{
"epoch": 3.08,
"grad_norm": 3475.076920659395,
"learning_rate": 7.500000000000001e-06,
"loss": -229.2591,
"step": 40
},
{
"epoch": 3.48,
"grad_norm": 3651.1737873444736,
"learning_rate": 7.0370370370370375e-06,
"loss": -356.1301,
"step": 45
},
{
"epoch": 3.88,
"grad_norm": 439.7775722171736,
"learning_rate": 6.574074074074075e-06,
"loss": -378.3872,
"step": 50
},
{
"epoch": 4.24,
"grad_norm": 434.2588747466371,
"learning_rate": 6.111111111111112e-06,
"loss": -341.9133,
"step": 55
},
{
"epoch": 4.64,
"grad_norm": 444.98196659861003,
"learning_rate": 5.6481481481481485e-06,
"loss": -378.8201,
"step": 60
},
{
"epoch": 5.0,
"grad_norm": 408.67513933006654,
"learning_rate": 5.185185185185185e-06,
"loss": -345.2147,
"step": 65
},
{
"epoch": 5.4,
"grad_norm": 392.3829022170684,
"learning_rate": 4.722222222222222e-06,
"loss": -385.2472,
"step": 70
},
{
"epoch": 5.8,
"grad_norm": 372.4425529988918,
"learning_rate": 4.2592592592592596e-06,
"loss": -384.5583,
"step": 75
},
{
"epoch": 6.16,
"grad_norm": 376.5960722072173,
"learning_rate": 3.796296296296297e-06,
"loss": -356.0032,
"step": 80
},
{
"epoch": 6.5600000000000005,
"grad_norm": 395.12972202440875,
"learning_rate": 3.3333333333333333e-06,
"loss": -388.6089,
"step": 85
},
{
"epoch": 6.96,
"grad_norm": 367.0863836880796,
"learning_rate": 2.8703703703703706e-06,
"loss": -389.7163,
"step": 90
},
{
"epoch": 7.32,
"grad_norm": 519.4771966031532,
"learning_rate": 2.4074074074074075e-06,
"loss": -356.6014,
"step": 95
},
{
"epoch": 7.72,
"grad_norm": 378.1832167305145,
"learning_rate": 1.944444444444445e-06,
"loss": -392.9878,
"step": 100
},
{
"epoch": 8.08,
"grad_norm": 381.25866581784237,
"learning_rate": 1.4814814814814815e-06,
"loss": -356.5252,
"step": 105
},
{
"epoch": 8.48,
"grad_norm": 1027.7502523973324,
"learning_rate": 1.0185185185185185e-06,
"loss": -395.6457,
"step": 110
},
{
"epoch": 8.88,
"grad_norm": 502.8117462390182,
"learning_rate": 5.555555555555555e-07,
"loss": -396.9123,
"step": 115
},
{
"epoch": 9.24,
"grad_norm": 914.2328850512966,
"learning_rate": 9.259259259259259e-08,
"loss": -359.5569,
"step": 120
},
{
"epoch": 9.24,
"step": 120,
"total_flos": 0.0,
"train_loss": -262.48087577770156,
"train_runtime": 1884.8014,
"train_samples_per_second": 2.122,
"train_steps_per_second": 0.064
}
],
"logging_steps": 5,
"max_steps": 120,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}