Upload folder using huggingface_hub
Browse files- meta_000866.json +59 -0
- model_000866.pt +3 -0
meta_000866.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 866,
|
| 3 |
+
"val_bpb": 0.9346889608214854,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 2048,
|
| 6 |
+
"vocab_size": 32768,
|
| 7 |
+
"n_layer": 12,
|
| 8 |
+
"n_head": 6,
|
| 9 |
+
"n_kv_head": 6,
|
| 10 |
+
"n_embd": 768,
|
| 11 |
+
"window_pattern": "SSSL"
|
| 12 |
+
},
|
| 13 |
+
"user_config": {
|
| 14 |
+
"run": "d12_feb8_bf16_proj_fp32_state_3",
|
| 15 |
+
"device_type": "",
|
| 16 |
+
"fp8": true,
|
| 17 |
+
"fp8_recipe": "tensorwise",
|
| 18 |
+
"depth": 12,
|
| 19 |
+
"aspect_ratio": 64,
|
| 20 |
+
"head_dim": 128,
|
| 21 |
+
"max_seq_len": 2048,
|
| 22 |
+
"window_pattern": "SSSL",
|
| 23 |
+
"num_iterations": -1,
|
| 24 |
+
"target_flops": -1.0,
|
| 25 |
+
"target_param_data_ratio": 8.25,
|
| 26 |
+
"device_batch_size": 32,
|
| 27 |
+
"total_batch_size": 1048576,
|
| 28 |
+
"embedding_lr": 0.3,
|
| 29 |
+
"unembedding_lr": 0.004,
|
| 30 |
+
"weight_decay": 0.2,
|
| 31 |
+
"matrix_lr": 0.02,
|
| 32 |
+
"scalar_lr": 0.5,
|
| 33 |
+
"adam_beta1": 0.8,
|
| 34 |
+
"adam_beta2": 0.95,
|
| 35 |
+
"warmup_ratio": 0.0,
|
| 36 |
+
"warmdown_ratio": 0.5,
|
| 37 |
+
"final_lr_frac": 0.0,
|
| 38 |
+
"resume_from_step": -1,
|
| 39 |
+
"eval_every": 250,
|
| 40 |
+
"eval_tokens": 20971520,
|
| 41 |
+
"core_metric_every": 999999,
|
| 42 |
+
"core_metric_max_per_task": -1,
|
| 43 |
+
"sample_every": -1,
|
| 44 |
+
"save_every": -1,
|
| 45 |
+
"model_tag": "d12_feb8_bf16_proj_fp32_state_3"
|
| 46 |
+
},
|
| 47 |
+
"device_batch_size": 32,
|
| 48 |
+
"max_seq_len": 2048,
|
| 49 |
+
"dataloader_state_dict": {
|
| 50 |
+
"pq_idx": 21,
|
| 51 |
+
"rg_idx": 40,
|
| 52 |
+
"epoch": 1
|
| 53 |
+
},
|
| 54 |
+
"loop_state": {
|
| 55 |
+
"min_val_bpb": 0.9346889608214854,
|
| 56 |
+
"smooth_train_loss": 3.057110628503737,
|
| 57 |
+
"total_training_time": 218.9185266494751
|
| 58 |
+
}
|
| 59 |
+
}
|
model_000866.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f53286637620ba2c34bea44e5e4a4f7f482c456f5bab8ef9d4955843560fbf87
|
| 3 |
+
size 622891889
|