Add FINAL5 seed-72 Dense resumable checkpoint at step 2000
Browse files- FINAL5_20261014/seed72_dense/checkpoint-2000/config.json +38 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank0.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank1.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank2.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank3.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank4.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/model.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank0.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank1.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank2.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank3.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank4.pt +3 -0
- FINAL5_20261014/seed72_dense/checkpoint-2000/training_state.pt +3 -0
FINAL5_20261014/seed72_dense/checkpoint-2000/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"qk_gain_fp32": true,
|
| 18 |
+
"rope_train_position_scale_set": null,
|
| 19 |
+
"attention_bias": false,
|
| 20 |
+
"attention_dropout": 0.0,
|
| 21 |
+
"bos_token_id": 151643,
|
| 22 |
+
"eos_token_id": 151643,
|
| 23 |
+
"checkpoint_activations": false,
|
| 24 |
+
"attention_type": "sparse_stochastic",
|
| 25 |
+
"sparse_block_size": 128,
|
| 26 |
+
"sparse_num_random_blocks": 31,
|
| 27 |
+
"sparse_mask_pool_size": 1,
|
| 28 |
+
"sparse_max_seq_length": 8192,
|
| 29 |
+
"sparse_layer_indices": null,
|
| 30 |
+
"sparse_mask_cache_dir": null,
|
| 31 |
+
"sparse_mask_seed": 0,
|
| 32 |
+
"sparse_mixed_nrb_set": null,
|
| 33 |
+
"moba_block_size": 128,
|
| 34 |
+
"moba_top_k": 32,
|
| 35 |
+
"asentmax_alpha": 1.5,
|
| 36 |
+
"asentmax_learn_temp": true,
|
| 37 |
+
"asentmax_init_temp": 1.0
|
| 38 |
+
}
|
FINAL5_20261014/seed72_dense/checkpoint-2000/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
|
FINAL5_20261014/seed72_dense/checkpoint-2000/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
|
FINAL5_20261014/seed72_dense/checkpoint-2000/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
|
FINAL5_20261014/seed72_dense/checkpoint-2000/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
|
FINAL5_20261014/seed72_dense/checkpoint-2000/loader_gen_state_rank4.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e0a234bd3fe283f29b7ed9ee4980692f92fb206de53b64fe8758792b41c82e6
|
| 3 |
+
size 6802
|
FINAL5_20261014/seed72_dense/checkpoint-2000/model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2694e7b70058c4a97275d2b870720d0756aaf0ddfe57243c0e981cef660a78ab
|
| 3 |
+
size 2404833883
|
FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank0.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf3f8461a21545d5bfeb0ed4aff0464bf5f8cb8e60b48198288d617ace67ddcc
|
| 3 |
+
size 14821
|
FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank1.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48bed20d6e249d8a32b5e17bcd02c6be69e4d360570b3ddc274278ce1ca5f9af
|
| 3 |
+
size 14885
|
FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank2.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d328ed111f5cf761888e6e562e88a94d1457c15f9ecf5fa4e50a28902161bbc
|
| 3 |
+
size 14821
|
FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank3.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e0050bab53bb96b17649dc0bc1558202e57a72c5dbfeb3d372a763bf74889ec
|
| 3 |
+
size 14885
|
FINAL5_20261014/seed72_dense/checkpoint-2000/rng_state_rank4.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d0475f089464589ac57151ac13fe9fa04d01e2fcb10adc09c9e5e4cb73ba837
|
| 3 |
+
size 14757
|
FINAL5_20261014/seed72_dense/checkpoint-2000/training_state.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b842c8fd3e2c8d671c17f9d631944dd59ae9d6737b2cb1678585d62ce6371507
|
| 3 |
+
size 4809712479
|