Add final paper checkpoint: checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462
Browse files- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/config.json +43 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank0.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank1.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank2.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank3.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/model.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank0.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank1.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank2.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank3.pt +3 -0
- checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/training_state.pt +3 -0
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/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": 72,
|
| 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_seed72/checkpoint-50462/loader_gen_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b21b4160791b11c535b1414390199117f5857a94ac58d7b970468583a72bb0bc
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32ec129fe94d28aa0ddb1827422f7723c72b31ebc7a9306e98894b6131bb7fe9
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c5588d56d5c8519bb2c45d4b62d3684da8142cd70467b7c6b989cd944f4e533
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/loader_gen_state_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05faccfef6d5dfcc8dd8aca9b50fb4a4f1b5e5ee52a9a29e99e1e68bd3e62b77
|
| 3 |
+
size 6802
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccfbfd7f701f00d6a2d27a4bf9272079be226e27787059d35a8cc20761845c47
|
| 3 |
+
size 2404819483
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d62c2c3615d59753ddc8e91b74332bfa8becde24b23c1bcc1291dd9a5b776f9
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb5eac4bfd5eaf229864a141d4f6bbc2c86ecc7631fc7d00fd626174c3cd1ec9
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:465ea33e22d29b39a0a48ea456b10b40de1a5a2b1f3097c96b8dba1ccaaaef8d
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/rng_state_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ff37894556981d6a566954e289cc73e8bb495e372cf638ff3dd29bdb1e7d5ce
|
| 3 |
+
size 14821
|
checkpoints/scale_1p2b_current/l8k1p2b_mixed_seed72/checkpoint-50462/training_state.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:477ecd58203befa7088a988df0f35e6e692ce5a1ed64cb2b39903bf21be0baeb
|
| 3 |
+
size 4809681879
|