Upload picoLLM SFT inference checkpoints
Browse files
chatsft_checkpoints/d24/meta_000482.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 482,
|
| 3 |
+
"val_bpb": 0.273515701872867,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 2048,
|
| 6 |
+
"vocab_size": 32768,
|
| 7 |
+
"n_layer": 24,
|
| 8 |
+
"n_head": 12,
|
| 9 |
+
"n_kv_head": 12,
|
| 10 |
+
"n_embd": 1536,
|
| 11 |
+
"window_pattern": "SSSL"
|
| 12 |
+
},
|
| 13 |
+
"user_config": {
|
| 14 |
+
"run": "dummy",
|
| 15 |
+
"wandb_entity": "montekkundan",
|
| 16 |
+
"device_type": "",
|
| 17 |
+
"model_tag": null,
|
| 18 |
+
"model_step": null,
|
| 19 |
+
"load_optimizer": 1,
|
| 20 |
+
"num_iterations": -1,
|
| 21 |
+
"max_seq_len": null,
|
| 22 |
+
"device_batch_size": 16,
|
| 23 |
+
"total_batch_size": 1048576,
|
| 24 |
+
"embedding_lr": null,
|
| 25 |
+
"unembedding_lr": null,
|
| 26 |
+
"matrix_lr": null,
|
| 27 |
+
"init_lr_frac": 0.8,
|
| 28 |
+
"warmup_ratio": 0.0,
|
| 29 |
+
"warmdown_ratio": 0.5,
|
| 30 |
+
"final_lr_frac": 0.0,
|
| 31 |
+
"eval_every": 200,
|
| 32 |
+
"eval_tokens": 20971520,
|
| 33 |
+
"chatcore_every": 200,
|
| 34 |
+
"chatcore_max_cat": -1,
|
| 35 |
+
"chatcore_max_sample": 24,
|
| 36 |
+
"mmlu_epochs": 3,
|
| 37 |
+
"gsm8k_epochs": 4
|
| 38 |
+
}
|
| 39 |
+
}
|
chatsft_checkpoints/d24/model_000482.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3174e5fc5003e66a53ae1fd5d98f382ddc7a760173f82dae24d191a5fef89f04
|
| 3 |
+
size 4227935530
|