Upload folder using huggingface_hub
Browse files- meta_021400.json +50 -0
- model_021400.pt +3 -0
- optim_021400_rank0.pt +3 -0
- optim_021400_rank1.pt +3 -0
- optim_021400_rank2.pt +3 -0
- optim_021400_rank3.pt +3 -0
- optim_021400_rank4.pt +3 -0
- optim_021400_rank5.pt +3 -0
- optim_021400_rank6.pt +3 -0
- optim_021400_rank7.pt +3 -0
meta_021400.json
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 21400,
|
| 3 |
+
"val_bpb": 0.811933320366963,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 2048,
|
| 6 |
+
"vocab_size": 65536,
|
| 7 |
+
"n_layer": 20,
|
| 8 |
+
"n_head": 10,
|
| 9 |
+
"n_kv_head": 10,
|
| 10 |
+
"n_embd": 1280
|
| 11 |
+
},
|
| 12 |
+
"user_config": {
|
| 13 |
+
"run": "speedrun_fixed",
|
| 14 |
+
"device_type": "",
|
| 15 |
+
"depth": 20,
|
| 16 |
+
"max_seq_len": 2048,
|
| 17 |
+
"num_iterations": -1,
|
| 18 |
+
"target_flops": -1.0,
|
| 19 |
+
"target_param_data_ratio": 20,
|
| 20 |
+
"device_batch_size": 32,
|
| 21 |
+
"total_batch_size": 524288,
|
| 22 |
+
"embedding_lr": 0.2,
|
| 23 |
+
"unembedding_lr": 0.004,
|
| 24 |
+
"weight_decay": 0.0,
|
| 25 |
+
"matrix_lr": 0.02,
|
| 26 |
+
"grad_clip": 1.0,
|
| 27 |
+
"warmup_ratio": 0.0,
|
| 28 |
+
"warmdown_ratio": 0.2,
|
| 29 |
+
"final_lr_frac": 0.0,
|
| 30 |
+
"resume_from_step": -1,
|
| 31 |
+
"eval_every": 250,
|
| 32 |
+
"eval_tokens": 10485760,
|
| 33 |
+
"core_metric_every": -1,
|
| 34 |
+
"core_metric_max_per_task": 500,
|
| 35 |
+
"sample_every": 2000,
|
| 36 |
+
"save_every": 500,
|
| 37 |
+
"model_tag": ""
|
| 38 |
+
},
|
| 39 |
+
"device_batch_size": 32,
|
| 40 |
+
"max_seq_len": 2048,
|
| 41 |
+
"dataloader_state_dict": {
|
| 42 |
+
"pq_idx": 198,
|
| 43 |
+
"rg_idx": 48
|
| 44 |
+
},
|
| 45 |
+
"loop_state": {
|
| 46 |
+
"min_val_bpb": 0.811933320366963,
|
| 47 |
+
"smooth_train_loss": 2.715633973883271,
|
| 48 |
+
"total_training_time": 10385.098405361176
|
| 49 |
+
}
|
| 50 |
+
}
|
model_021400.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:702f6352c011848fa3d71d6cf05e888ea10137ed87780334007907808b6b1c96
|
| 3 |
+
size 2076230219
|
optim_021400_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b38bb5ed042e987c11d16ee4a53c3d522e6916bddf0764e8a15b5bf631d877d
|
| 3 |
+
size 407643053
|
optim_021400_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8e9d0bc89e5c201e87e5ffa94b00690e58384bf5caf1b7276c9dda273c1aac9
|
| 3 |
+
size 407643053
|
optim_021400_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9c107e657bc2e8c173cbdd1e984ce52da24460886826d060e7dce936f9a3017
|
| 3 |
+
size 407643053
|
optim_021400_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa6ec89738adc64fe5cebb31eba044d3db220f5fd8a5cc1325461bbcf6a72b25
|
| 3 |
+
size 407643053
|
optim_021400_rank4.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:427f33306c8aa25cb07c5c5d989649a406e4504a51fd8aaeb633cf092dfeeea6
|
| 3 |
+
size 355213661
|
optim_021400_rank5.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec803553e4e9ed258e70aa30a179fa9b37c80643ce67cfeacce2bd3a40cafde9
|
| 3 |
+
size 355213661
|
optim_021400_rank6.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56c58d2cae52f894b3fb077edcd6f46ddfdbf29cf6a7388fd0cd9f4fe2b53a59
|
| 3 |
+
size 355213661
|
optim_021400_rank7.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79e0ed3e2060654c2df8aacce99ad1809b285e62199d8fed09b9667367aa59c5
|
| 3 |
+
size 355213661
|