Upload folder using huggingface_hub
Browse files- models/30b/checklist.chk +6 -0
- models/30b/consolidate_params.json +8 -0
- models/30b/consolidated.00.pth +3 -0
- models/30b/consolidated.01.pth +3 -0
- models/30b/consolidated.02.pth +3 -0
- models/30b/consolidated.03.pth +3 -0
- models/30b/params.json +169 -0
models/30b/checklist.chk
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
d1c053d174845e350427d385850af0f1 consolidated.00.pth
|
2 |
+
09ad3734acf6082d2284f0e297f30e5f consolidated.01.pth
|
3 |
+
a184ca575a4e8cdc7229966509077980 consolidated.02.pth
|
4 |
+
e54adfa3e07d78af0c656d69fce8d6d8 consolidated.03.pth
|
5 |
+
3dd10b90ed4295f92155863d309cd0c7 consolidate_params.json
|
6 |
+
967b52c1f1b2d3f40fb9aafea7834bb9 params.json
|
models/30b/consolidate_params.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"dtype": "bf16",
|
3 |
+
"model_parallel_size": 4,
|
4 |
+
"on_gpu": true,
|
5 |
+
"src": "/fsx-onellm/rpasunuru/SFT/v2.1_textpp_30b_730k_sftv1.4_exp1/v2.1_textpp_30b_730k_sftv1.4_exp1_run000/checkpoints/checkpoint_0001200_noimggen",
|
6 |
+
"tgt": "/fsx-onellm/rpasunuru/SFT/v2.1_textpp_30b_730k_sftv1.4_exp1/v2.1_textpp_30b_730k_sftv1.4_exp1_run000/checkpoints/checkpoint_0001200_noimggen_consolidated/",
|
7 |
+
"tokenizer_path": null
|
8 |
+
}
|
models/30b/consolidated.00.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fca533c6c74bc8e933e8241040b5bcb007c7299cb1a8090791869b34c8f32a26
|
3 |
+
size 17148158769
|
models/30b/consolidated.01.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:42603cc253231fb9e15feb1754daacfbbd0d04aabda4caf1473ec8576581c8cd
|
3 |
+
size 17148158769
|
models/30b/consolidated.02.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b9f4970b72db7ebfbeb29745875996837588ac80315b76fa1050dc05a887fb28
|
3 |
+
size 17148158769
|
models/30b/consolidated.03.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:becf308502539eb8165f419ff1c675858537dd418ddad9b3adcc46e4404c376b
|
3 |
+
size 17148158769
|
models/30b/params.json
ADDED
@@ -0,0 +1,169 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"async_checkpointing": false,
|
3 |
+
"async_eval_ngpus": -1,
|
4 |
+
"batch_size": 4,
|
5 |
+
"data": "",
|
6 |
+
"disable_logging": false,
|
7 |
+
"disable_workers_print": false,
|
8 |
+
"dtype": "bf16",
|
9 |
+
"dump_after_steps": 0,
|
10 |
+
"dump_dir": "/fsx-onellm/rpasunuru/SFT/v2.1_textpp_30b_730k_sftv1.4_exp1/v2.1_textpp_30b_730k_sftv1.4_exp1_run000",
|
11 |
+
"dump_freq": 400,
|
12 |
+
"dump_profile_traces": false,
|
13 |
+
"enable_loss_tracker": false,
|
14 |
+
"epochs": -1,
|
15 |
+
"eval_freq": 100000,
|
16 |
+
"exp_id": "",
|
17 |
+
"exp_name": "",
|
18 |
+
"finetuning_dir": "/fsx-onellm/shared/from_rsc//v2.1_30b_qk_zloss_snorm_Nov_26_3_run000_checkpoint_0730000",
|
19 |
+
"fp32_reduce_scatter": "all",
|
20 |
+
"gpu_check_level": 3,
|
21 |
+
"image_loss_weight": 1.0,
|
22 |
+
"image_text_rotation_prob": 0.0,
|
23 |
+
"instruct": {
|
24 |
+
"no_loss_prompt": true,
|
25 |
+
"no_loss_truncated": false,
|
26 |
+
"use_eot": true
|
27 |
+
},
|
28 |
+
"instruct_data": "/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/long_caption:2.92,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/vqa:4.59,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/text2image:10.44,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/llama2_rjv6_helpful:43.27,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/code_llama:0.51,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/interleaved_batch1-17:27.45,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/image_dialogue:7.46,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/llama2_rjv6_harmless:0.97,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/cybersec_safety:0.33,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/onellm_multimodal_safety:0.86,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/autosafety:0.51,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/rainbow_safety:0.10,/fsx-onellm/shared/finetuning/sft_v1.4_data/splits/train/genai_safety:0.58",
|
29 |
+
"iter_gopher": {
|
30 |
+
"buffer_size": 16,
|
31 |
+
"max_precompute": 10,
|
32 |
+
"n_chars_by_tok": 15,
|
33 |
+
"n_seqs_to_concat": 10,
|
34 |
+
"num_processes": 1
|
35 |
+
},
|
36 |
+
"iter_jsonl": {
|
37 |
+
"buffer_size": 64,
|
38 |
+
"same_data": false
|
39 |
+
},
|
40 |
+
"iter_multi": {
|
41 |
+
"buffer_size": 512,
|
42 |
+
"ignore_extra_chunks": true,
|
43 |
+
"max_precompute": 20,
|
44 |
+
"multiprocess": true
|
45 |
+
},
|
46 |
+
"iter_type": "multi",
|
47 |
+
"keep_checkpoints_every_steps": 400,
|
48 |
+
"keep_eval_checkpoints": true,
|
49 |
+
"keep_n_last_checkpoints": 2,
|
50 |
+
"log_all_steps": false,
|
51 |
+
"log_freq": 10,
|
52 |
+
"log_updates": true,
|
53 |
+
"log_wandb": false,
|
54 |
+
"loss_rescaling": false,
|
55 |
+
"model": {
|
56 |
+
"add_extra_toks": "0",
|
57 |
+
"alpha_depth": "disabled",
|
58 |
+
"attn_dropout": 0,
|
59 |
+
"attn_to_keep": "all",
|
60 |
+
"custom_bwd": false,
|
61 |
+
"dim": 8192,
|
62 |
+
"dropout": 0.05,
|
63 |
+
"efficient_attn": "flash",
|
64 |
+
"emb_dropout": 0,
|
65 |
+
"ffn_dim_multiplier": 1.0,
|
66 |
+
"ffn_dropout": 0,
|
67 |
+
"full_logging_n_layers": 4,
|
68 |
+
"fuse_sequence_parallel": false,
|
69 |
+
"init": {
|
70 |
+
"coeff_std": null,
|
71 |
+
"depth_last": false,
|
72 |
+
"fixed_std": null,
|
73 |
+
"no_init": false,
|
74 |
+
"pos_init_scalar": null,
|
75 |
+
"use_depth": "current",
|
76 |
+
"use_gaussian": true
|
77 |
+
},
|
78 |
+
"layer_ckpt": "0::2",
|
79 |
+
"linear_residual_dropout": false,
|
80 |
+
"loss_parallel": true,
|
81 |
+
"max_length": 2048,
|
82 |
+
"multiple_of": 256,
|
83 |
+
"n_heads": 64,
|
84 |
+
"n_kv_heads": 8,
|
85 |
+
"n_layers": 48,
|
86 |
+
"non_linearity": "swiglu",
|
87 |
+
"norm_affine": true,
|
88 |
+
"norm_eps": 1e-05,
|
89 |
+
"norm_type": "rmsnorm",
|
90 |
+
"output_dropout": 0,
|
91 |
+
"output_size": -1,
|
92 |
+
"pre_norm": true,
|
93 |
+
"qk_normalization": true,
|
94 |
+
"recompute_attn": true,
|
95 |
+
"recompute_fc1_out": true,
|
96 |
+
"recompute_fc3_out": true,
|
97 |
+
"residual_dropout": 0.0,
|
98 |
+
"rope_theta": 10000.0,
|
99 |
+
"sequence_parallel": false,
|
100 |
+
"swin_norm": true,
|
101 |
+
"turn_eos_token": "<eos>",
|
102 |
+
"use_rope": true,
|
103 |
+
"vocab_size": 65536
|
104 |
+
},
|
105 |
+
"model_parallel_size": 4,
|
106 |
+
"no_final_ckpt": false,
|
107 |
+
"num_retrieved_docs": 0,
|
108 |
+
"old_mp": -1,
|
109 |
+
"old_world_size": -1,
|
110 |
+
"optim": {
|
111 |
+
"beta1": 0.9,
|
112 |
+
"beta2": 0.95,
|
113 |
+
"clip": 1.0,
|
114 |
+
"cosine_theta": 1.0,
|
115 |
+
"cycle_length": 1.0,
|
116 |
+
"epsilon": 1e-08,
|
117 |
+
"exp_factor": 0.5,
|
118 |
+
"lr": 1e-05,
|
119 |
+
"lr_min_ratio": 0.1,
|
120 |
+
"scheduler": "cosine",
|
121 |
+
"use_deprecated_optim": false,
|
122 |
+
"warmup": 100,
|
123 |
+
"weight_decay": 0.1
|
124 |
+
},
|
125 |
+
"periodic_gpu_check": true,
|
126 |
+
"profile_freq": -1,
|
127 |
+
"reshard_after_forward": true,
|
128 |
+
"restore_dataloader_position": false,
|
129 |
+
"retrieval_prob": 0.0,
|
130 |
+
"rlhf": null,
|
131 |
+
"root_dump_dir": "",
|
132 |
+
"save_optimizer_states": true,
|
133 |
+
"seq_len": 4096,
|
134 |
+
"slurm": {
|
135 |
+
"global_rank": 0,
|
136 |
+
"is_slurm_job": true,
|
137 |
+
"world_size": 128
|
138 |
+
},
|
139 |
+
"steps": 1200,
|
140 |
+
"tokenizer": "/fsx-onellm/rpasunuru/models/cm3z/cm3v2_7b_placeholder/gpt2-unified-image-sentinel.json",
|
141 |
+
"tokenizer_dir": "/fsx/guismay/data/large_experiments/fair_llm/datasets/tokenizers",
|
142 |
+
"torch_seed": -1,
|
143 |
+
"unlimited_steps": false,
|
144 |
+
"use_hf_tokenizer": true,
|
145 |
+
"valid": {
|
146 |
+
"batch_size": 1,
|
147 |
+
"debug": false,
|
148 |
+
"majority_voting": 0,
|
149 |
+
"n_batches": 100,
|
150 |
+
"onellm_eval": false,
|
151 |
+
"onellm_eval_media_storage": "",
|
152 |
+
"ppl_files_str": "",
|
153 |
+
"prompt_path": "",
|
154 |
+
"prompt_templates": "{}",
|
155 |
+
"random_fewshots": false,
|
156 |
+
"seq_len": 4096,
|
157 |
+
"tasks_root_dir": "",
|
158 |
+
"tasks_str": "",
|
159 |
+
"temperature": 1.0,
|
160 |
+
"top_k": 0,
|
161 |
+
"top_p": 0.0,
|
162 |
+
"use_sampling": false,
|
163 |
+
"write_eval": false
|
164 |
+
},
|
165 |
+
"wandb_entity": "violet-zct",
|
166 |
+
"wandb_project": "instruct_sft",
|
167 |
+
"water_marking_codes_str": null,
|
168 |
+
"z_loss_weight": 0.0001
|
169 |
+
}
|