Hugging Face's logo

rubentium
/
sparse-checkpoints

Model card Files
xet
 Community
rubentium commited on
Commit
d851b02
·
verified ·
1 Parent(s): 50d251b

Upload paper checkpoint l30k1_2b_fixed checkpoint-37465

Browse files
Files changed (11) hide show
  1. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/config.json +30 -0
  2. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank0.pt +3 -0
  3. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank1.pt +3 -0
  4. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank2.pt +3 -0
  5. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank3.pt +3 -0
  6. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/model.pt +3 -0
  7. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank0.pt +3 -0
  8. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank1.pt +3 -0
  9. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank2.pt +3 -0
  10. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank3.pt +3 -0
  11. checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/training_state.pt +3 -0
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "bos_token_id": 151643,
20
+ "eos_token_id": 151643,
21
+ "checkpoint_activations": true,
22
+ "attention_type": "sparse_stochastic",
23
+ "sparse_block_size": 128,
24
+ "sparse_num_random_blocks": 120,
25
+ "sparse_mask_pool_size": 1,
26
+ "sparse_max_seq_length": 30720,
27
+ "sparse_layer_indices": null,
28
+ "sparse_mask_cache_dir": "/mloscratch/homes/navasard/sparse/mask_cache_30k_b128_nrb120",
29
+ "sparse_mask_seed": 1
30
+ }
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f59a1416dd062038a7e2dbac8c330163f45326660faeb92a199f62460447539
3
+ size 6802
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9907050ee7aa7455a32aae231218aac0666dd0bc7b78294eeb2cb3adc77e63c8
3
+ size 6802
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3adc40f4c25919716a49dee06d626601e0abca351145ddf1ed45e0bbfca3ea0c
3
+ size 6802
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/loader_gen_state_rank3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e32dfbaa9aad434214e9087e9315fae33fa445ebde26c320765a117f68a6465
3
+ size 6802
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2afacf84d813157997c6cb4590b8e7868b2b3b5cd691039348bdf471780eb747
3
+ size 2404819483
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac6562f8fadc642c93e13bfb836c3143e5234ca65f6c372ea2e786a3b2986a80
3
+ size 14821
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:316919bdaccc1dfd5f4564c2ac6228b935df19820fb6023cfd9cf3f80f3977e1
3
+ size 14821
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c393554b63c76a39db9407c79a86f376125986db6095eadb4922bfd915c3ff0
3
+ size 14885
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/rng_state_rank3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b12b9242df6a822f5070efd703b173cf566fd7d32ca4687726c916531eed4b3
3
+ size 14821
checkpoints/scale_1p2b_factorial/l30k1_2b_fixed/checkpoint-37465/training_state.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e052fe5cbd3aa34c44ef2374821545f104b66b3b1bd890b273b9a913c4e6522
3
+ size 4809681879