Upload paper checkpoint l8k365m_mixed_seed47 checkpoint-44000
Browse files- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/config.json +43 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/loader_gen_state_rank0.pt +3 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/loader_gen_state_rank1.pt +3 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/model.pt +3 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/rng_state_rank0.pt +3 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/rng_state_rank1.pt +3 -0
- checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/training_state.pt +3 -0
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"vocab_size": 151936,
|
| 3 |
+
"hidden_size": 768,
|
| 4 |
+
"intermediate_size": 3072,
|
| 5 |
+
"num_hidden_layers": 28,
|
| 6 |
+
"num_attention_heads": 6,
|
| 7 |
+
"num_key_value_heads": 2,
|
| 8 |
+
"head_dim": 128,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"max_position_embeddings": 32768,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"rms_norm_eps": 1e-06,
|
| 13 |
+
"use_cache": true,
|
| 14 |
+
"tie_word_embeddings": true,
|
| 15 |
+
"rope_theta": 1000000,
|
| 16 |
+
"rope_scaling": null,
|
| 17 |
+
"rope_train_position_scale_set": null,
|
| 18 |
+
"attention_bias": false,
|
| 19 |
+
"attention_dropout": 0.0,
|
| 20 |
+
"bos_token_id": 151643,
|
| 21 |
+
"eos_token_id": 151643,
|
| 22 |
+
"checkpoint_activations": false,
|
| 23 |
+
"attention_type": "sparse_stochastic",
|
| 24 |
+
"sparse_block_size": 128,
|
| 25 |
+
"sparse_num_random_blocks": 31,
|
| 26 |
+
"sparse_mask_pool_size": 10000,
|
| 27 |
+
"sparse_max_seq_length": 8192,
|
| 28 |
+
"sparse_layer_indices": null,
|
| 29 |
+
"sparse_mask_cache_dir": "/mloscratch/homes/navasard/sparse/mask_cache_8k_mixed",
|
| 30 |
+
"sparse_mask_seed": 0,
|
| 31 |
+
"sparse_mixed_nrb_set": [
|
| 32 |
+
63,
|
| 33 |
+
47,
|
| 34 |
+
31,
|
| 35 |
+
21,
|
| 36 |
+
11
|
| 37 |
+
],
|
| 38 |
+
"moba_block_size": 128,
|
| 39 |
+
"moba_top_k": 32,
|
| 40 |
+
"asentmax_alpha": 1.5,
|
| 41 |
+
"asentmax_learn_temp": true,
|
| 42 |
+
"asentmax_init_temp": 1.0
|
| 43 |
+
}
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/loader_gen_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d29ebcba1740400d02a9731a1b2ce2d06da520e9a5f365770ae3366476cfa6bb
|
| 3 |
+
size 6802
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/loader_gen_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6cbcc77de4742713528531e0ef8cdd89a1973c9e431acb1771b42ca22d7bd1f
|
| 3 |
+
size 6802
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81a31f54ec1978ee742549bfbfc5d0295d334cb7bf922243f6cdd95a0c1599e8
|
| 3 |
+
size 718031899
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/rng_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fbd1fa5b004499ce36b181c723a919d43d48f68b0162c3504b0e08e1a3ef1494
|
| 3 |
+
size 14757
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/rng_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f461b4609ca69bb0d278393931e91c3edac0ae14e00d9bd994a7987af0b5aba3
|
| 3 |
+
size 14821
|
checkpoints/primary_365m/l8k365m_mixed_seed47/checkpoint-44000/training_state.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d75522add2b30298b83a48616b7112d83b73788c0a8e14ac48d5fb93528285b
|
| 3 |
+
size 1436105299
|