Add final paper checkpoint: checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462
Browse files- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/config.json +43 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank0.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank1.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank2.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank3.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank4.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank5.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/model.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank0.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank1.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank2.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank3.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank4.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank5.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/training_state.pt +3 -0
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"vocab_size": 151936,
|
| 3 |
+
"hidden_size": 1536,
|
| 4 |
+
"intermediate_size": 6144,
|
| 5 |
+
"num_hidden_layers": 28,
|
| 6 |
+
"num_attention_heads": 12,
|
| 7 |
+
"num_key_value_heads": 4,
|
| 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": true,
|
| 23 |
+
"attention_type": "sparse_stochastic",
|
| 24 |
+
"sparse_block_size": 128,
|
| 25 |
+
"sparse_num_random_blocks": 35,
|
| 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": 70,
|
| 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/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3d3d87a844d4fd6548dfc4c1235a1ee3e262506bc6260da75e7cbf9d9da819c
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2c029e9127ff4c346ab99e6af8377a6903003909fec0b073b70bfafc01d4517
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:917e752ef93a10f1da3dc49f9c18adc43ae1bb568ee13d528c9824ee4513b511
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:389dee6373df6f406c9ae897b97bc5acc1b426bc8c9810e2754f7d2f892a256f
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank4.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:003f8a68f6377e121aa1e72c457466320b22898c8d61a840ded7da21790b0d73
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/loader_gen_state_rank5.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3e0c9bf4dad32ac42f0c9f053baa4af26df892924de882f06738359b36322b1
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e33991c18c04fb1a5c29d325bdd39b343c0299e581d93699c62e0d5c6e3b70b
|
| 3 |
+
size 2404819483
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca040b4e3213859fff31ec3fc67cfe849fdb82ba8b23b9a1071944559f0c7844
|
| 3 |
+
size 14757
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9713836b925fec9e3a9c4b4d13fe0241d9ef10ca0a9214076621dfa6e2832860
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1de0a928a6b70dd8f236cabbeb948257ed63806a9233f192c1740d8960f99620
|
| 3 |
+
size 14885
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d730d115312760eacbf364f605b05ff6146c3816bbb0de7f070a066bd4137dc
|
| 3 |
+
size 14885
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank4.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af3ae40373129ff4f062ba96852005157b422f017292d6d8f7738632c50d1abe
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/rng_state_rank5.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d063d8b4cd439658f32a4b786d0eb1833c14c733aafcd7538381afc1cd9dc38
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed70_repaired/checkpoint-50462/training_state.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:964941370917066c5770c6f0d63207dd5182122996e9d891f9f47a2d1e3ebac5
|
| 3 |
+
size 4809681879
|