the-jb/tofu_Llama-3.2-3B-Instruct_forget10_RMU
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_RMU
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-12 20:02:35 +08:00
commit 16c69541d7
24 changed files with 42781 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": 0.6055574595462542,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0585,
"step": 5
},
{
"epoch": 0.8,
"grad_norm": 0.19756118382686896,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0494,
"step": 10
},
{
"epoch": 1.16,
"grad_norm": 0.1919675100597038,
"learning_rate": 9.814814814814815e-06,
"loss": 0.0365,
"step": 15
},
{
"epoch": 1.56,
"grad_norm": 0.16799293615119748,
"learning_rate": 9.351851851851854e-06,
"loss": 0.0356,
"step": 20
},
{
"epoch": 1.96,
"grad_norm": 0.19130951276513097,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0292,
"step": 25
},
{
"epoch": 2.32,
"grad_norm": 0.15553184009775037,
"learning_rate": 8.425925925925926e-06,
"loss": 0.0202,
"step": 30
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.1624446099025784,
"learning_rate": 7.962962962962963e-06,
"loss": 0.0174,
"step": 35
},
{
"epoch": 3.08,
"grad_norm": 0.1179964545542203,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0124,
"step": 40
},
{
"epoch": 3.48,
"grad_norm": 0.15209315018643382,
"learning_rate": 7.0370370370370375e-06,
"loss": 0.0111,
"step": 45
},
{
"epoch": 3.88,
"grad_norm": 0.14192327495594986,
"learning_rate": 6.574074074074075e-06,
"loss": 0.0107,
"step": 50
},
{
"epoch": 4.24,
"grad_norm": 0.06740151668181002,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0075,
"step": 55
},
{
"epoch": 4.64,
"grad_norm": 0.13210528909388133,
"learning_rate": 5.6481481481481485e-06,
"loss": 0.0083,
"step": 60
},
{
"epoch": 5.0,
"grad_norm": 0.07753865089908243,
"learning_rate": 5.185185185185185e-06,
"loss": 0.0058,
"step": 65
},
{
"epoch": 5.4,
"grad_norm": 0.05734572561784096,
"learning_rate": 4.722222222222222e-06,
"loss": 0.0063,
"step": 70
},
{
"epoch": 5.8,
"grad_norm": 0.08821602203064255,
"learning_rate": 4.2592592592592596e-06,
"loss": 0.0058,
"step": 75
},
{
"epoch": 6.16,
"grad_norm": 0.0778017341776081,
"learning_rate": 3.796296296296297e-06,
"loss": 0.0052,
"step": 80
},
{
"epoch": 6.5600000000000005,
"grad_norm": 0.1499250102955357,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0049,
"step": 85
},
{
"epoch": 6.96,
"grad_norm": 0.049505032136154786,
"learning_rate": 2.8703703703703706e-06,
"loss": 0.0042,
"step": 90
},
{
"epoch": 7.32,
"grad_norm": 0.04421733042808474,
"learning_rate": 2.4074074074074075e-06,
"loss": 0.0037,
"step": 95
},
{
"epoch": 7.72,
"grad_norm": 0.053788292070017965,
"learning_rate": 1.944444444444445e-06,
"loss": 0.0048,
"step": 100
},
{
"epoch": 8.08,
"grad_norm": 0.058760245803698326,
"learning_rate": 1.4814814814814815e-06,
"loss": 0.0037,
"step": 105
},
{
"epoch": 8.48,
"grad_norm": 0.09203517306494899,
"learning_rate": 1.0185185185185185e-06,
"loss": 0.0039,
"step": 110
},
{
"epoch": 8.88,
"grad_norm": 0.13995065222222594,
"learning_rate": 5.555555555555555e-07,
"loss": 0.0036,
"step": 115
},
{
"epoch": 9.24,
"grad_norm": 0.09337234175753063,
"learning_rate": 9.259259259259259e-08,
"loss": 0.0027,
"step": 120
},
{
"epoch": 9.24,
"step": 120,
"total_flos": 0.0,
"train_loss": 0.014640537897745769,
"train_runtime": 1252.5714,
"train_samples_per_second": 3.193,
"train_steps_per_second": 0.096
}
],
"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
}