Hugging Face's logo

mikudev
/
WizardLM-2-8x22B-GTPQ

Text Generation
Transformers
mixtral
4-bit precision
gptq
Model card Files
xet
 Community
mikudev commited on
Commit
7ce71b6
·
verified ·
1 Parent(s): ffa911d

Upload folder using huggingface_hub

Browse files
Files changed (39) hide show
  1. a.py +43 -0
  2. chunk_paths.json +1 -0
  3. config.json +44 -0
  4. gptq_model-4bit-128g-01-of-35.safetensors +3 -0
  5. gptq_model-4bit-128g-02-of-35.safetensors +3 -0
  6. gptq_model-4bit-128g-03-of-35.safetensors +3 -0
  7. gptq_model-4bit-128g-04-of-35.safetensors +3 -0
  8. gptq_model-4bit-128g-05-of-35.safetensors +3 -0
  9. gptq_model-4bit-128g-06-of-35.safetensors +3 -0
  10. gptq_model-4bit-128g-07-of-35.safetensors +3 -0
  11. gptq_model-4bit-128g-08-of-35.safetensors +3 -0
  12. gptq_model-4bit-128g-09-of-35.safetensors +3 -0
  13. gptq_model-4bit-128g-10-of-35.safetensors +3 -0
  14. gptq_model-4bit-128g-11-of-35.safetensors +3 -0
  15. gptq_model-4bit-128g-12-of-35.safetensors +3 -0
  16. gptq_model-4bit-128g-13-of-35.safetensors +3 -0
  17. gptq_model-4bit-128g-14-of-35.safetensors +3 -0
  18. gptq_model-4bit-128g-15-of-35.safetensors +3 -0
  19. gptq_model-4bit-128g-16-of-35.safetensors +3 -0
  20. gptq_model-4bit-128g-17-of-35.safetensors +3 -0
  21. gptq_model-4bit-128g-18-of-35.safetensors +3 -0
  22. gptq_model-4bit-128g-19-of-35.safetensors +3 -0
  23. gptq_model-4bit-128g-20-of-35.safetensors +3 -0
  24. gptq_model-4bit-128g-21-of-35.safetensors +3 -0
  25. gptq_model-4bit-128g-22-of-35.safetensors +3 -0
  26. gptq_model-4bit-128g-23-of-35.safetensors +3 -0
  27. gptq_model-4bit-128g-24-of-35.safetensors +3 -0
  28. gptq_model-4bit-128g-25-of-35.safetensors +3 -0
  29. gptq_model-4bit-128g-26-of-35.safetensors +3 -0
  30. gptq_model-4bit-128g-27-of-35.safetensors +3 -0
  31. gptq_model-4bit-128g-28-of-35.safetensors +3 -0
  32. gptq_model-4bit-128g-29-of-35.safetensors +3 -0
  33. gptq_model-4bit-128g-30-of-35.safetensors +3 -0
  34. gptq_model-4bit-128g-31-of-35.safetensors +3 -0
  35. gptq_model-4bit-128g-32-of-35.safetensors +3 -0
  36. gptq_model-4bit-128g-33-of-35.safetensors +3 -0
  37. gptq_model-4bit-128g-34-of-35.safetensors +3 -0
  38. gptq_model-4bit-128g-35-of-35.safetensors +3 -0
  39. quantize_config.json +13 -0
a.py ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import math
3
+ import json
4
+
5
+ CHUNK_SIZE = 2 * 1024**3 # 40GB
6
+
7
+ CHUNK_PATHS_FILE = "chunk_paths.json"
8
+
9
+ def split(filepath, chunk_size=CHUNK_SIZE):
10
+ basename = os.path.basename(filepath)
11
+ dirname = os.path.dirname(filepath)
12
+ extension = basename.split(".")[-1]
13
+
14
+ filename_no_ext = basename.split(".")[-2]
15
+ file_size = os.path.getsize(filepath)
16
+
17
+ num_chunks = math.ceil(file_size / chunk_size)
18
+ digit_count = len(str(num_chunks))
19
+
20
+ chunk_paths = []
21
+
22
+ for i in range(1, num_chunks+1):
23
+ start = (i-1) * chunk_size
24
+
25
+ chunk_filename = f"{filename_no_ext}-{str(i).zfill(digit_count)}-of-{str(num_chunks).zfill(digit_count)}.{extension}"
26
+ split_path = os.path.join(dirname, chunk_filename)
27
+
28
+ with open(filepath, "rb") as f_in:
29
+ f_in.seek(start)
30
+ chunk = f_in.read(chunk_size)
31
+
32
+ with open(split_path, "wb") as f_out:
33
+ f_out.write(chunk)
34
+
35
+ chunk_paths.append(split_path)
36
+
37
+ with open(CHUNK_PATHS_FILE, 'w') as f:
38
+ json.dump(chunk_paths, f)
39
+
40
+ return chunk_paths
41
+
42
+ main_filepath = "gptq_model-4bit-128g.safetensors" # File to be split
43
+ chunk_paths = split(main_filepath)
chunk_paths.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["gptq_model-4bit-128g-01-of-35.safetensors", "gptq_model-4bit-128g-02-of-35.safetensors", "gptq_model-4bit-128g-03-of-35.safetensors", "gptq_model-4bit-128g-04-of-35.safetensors", "gptq_model-4bit-128g-05-of-35.safetensors", "gptq_model-4bit-128g-06-of-35.safetensors", "gptq_model-4bit-128g-07-of-35.safetensors", "gptq_model-4bit-128g-08-of-35.safetensors", "gptq_model-4bit-128g-09-of-35.safetensors", "gptq_model-4bit-128g-10-of-35.safetensors", "gptq_model-4bit-128g-11-of-35.safetensors", "gptq_model-4bit-128g-12-of-35.safetensors", "gptq_model-4bit-128g-13-of-35.safetensors", "gptq_model-4bit-128g-14-of-35.safetensors", "gptq_model-4bit-128g-15-of-35.safetensors", "gptq_model-4bit-128g-16-of-35.safetensors", "gptq_model-4bit-128g-17-of-35.safetensors", "gptq_model-4bit-128g-18-of-35.safetensors", "gptq_model-4bit-128g-19-of-35.safetensors", "gptq_model-4bit-128g-20-of-35.safetensors", "gptq_model-4bit-128g-21-of-35.safetensors", "gptq_model-4bit-128g-22-of-35.safetensors", "gptq_model-4bit-128g-23-of-35.safetensors", "gptq_model-4bit-128g-24-of-35.safetensors", "gptq_model-4bit-128g-25-of-35.safetensors", "gptq_model-4bit-128g-26-of-35.safetensors", "gptq_model-4bit-128g-27-of-35.safetensors", "gptq_model-4bit-128g-28-of-35.safetensors", "gptq_model-4bit-128g-29-of-35.safetensors", "gptq_model-4bit-128g-30-of-35.safetensors", "gptq_model-4bit-128g-31-of-35.safetensors", "gptq_model-4bit-128g-32-of-35.safetensors", "gptq_model-4bit-128g-33-of-35.safetensors", "gptq_model-4bit-128g-34-of-35.safetensors", "gptq_model-4bit-128g-35-of-35.safetensors"]
config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./WizardLM-2-8x22B",
3
+ "architectures": [
4
+ "MixtralForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 6144,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16384,
13
+ "max_position_embeddings": 65536,
14
+ "model_type": "mixtral",
15
+ "num_attention_heads": 48,
16
+ "num_experts_per_tok": 2,
17
+ "num_hidden_layers": 56,
18
+ "num_key_value_heads": 8,
19
+ "num_local_experts": 8,
20
+ "output_router_logits": false,
21
+ "quantization_config": {
22
+ "bits": 4,
23
+ "damp_percent": 0.01,
24
+ "desc_act": false,
25
+ "group_size": 128,
26
+ "is_marlin_format": false,
27
+ "model_file_base_name": null,
28
+ "model_name_or_path": null,
29
+ "quant_method": "gptq",
30
+ "static_groups": false,
31
+ "sym": true,
32
+ "true_sequential": true
33
+ },
34
+ "rms_norm_eps": 1e-05,
35
+ "rope_theta": 1000000,
36
+ "router_aux_loss_coef": 0.001,
37
+ "router_jitter_noise": 0.0,
38
+ "sliding_window": null,
39
+ "tie_word_embeddings": false,
40
+ "torch_dtype": "float16",
41
+ "transformers_version": "4.39.0.dev0",
42
+ "use_cache": false,
43
+ "vocab_size": 32000
44
+ }
gptq_model-4bit-128g-01-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f55921feaa8635166ea261219c1124a63cfacc0266be253fef3f427997e916e
3
+ size 2147483648
gptq_model-4bit-128g-02-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c1fd43d673aed98c5ca45e1324d8aa669106b120e72c534bd7f2edbf7db82e6
3
+ size 2147483648
gptq_model-4bit-128g-03-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:606f869b2cb04fad5a1f82f3853927eb0bf633f3d8a7f95894c2218b99b16e33
3
+ size 2147483648
gptq_model-4bit-128g-04-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8b0705f533701d5ddbcc1dc8bc2bb3bc3d7d41a343ec00794f17104b5850346
3
+ size 2147483648
gptq_model-4bit-128g-05-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06eb1c7559a2efe5fbf87817e20ed8fa479541ec90a3206b09857f75ba881535
3
+ size 2147483648
gptq_model-4bit-128g-06-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1290e1fa49b978f46fc57d791e3c83f789c27933aa524ca55dfed405cf0e0fac
3
+ size 2147483648
gptq_model-4bit-128g-07-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9718022c8bbb31edb63b30ea7c8c99c16e5dd967edbbbd73e4743b7d5db732f3
3
+ size 2147483648
gptq_model-4bit-128g-08-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:101b2b8fafcb9aa5cc77058d7a26c33fefdeaa7c5adb36bae3cecf2b6321a9ff
3
+ size 2147483648
gptq_model-4bit-128g-09-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6369bcd961882a9a05e641c97b297e63eb84bd650e44bf83faf98c1de91cc27
3
+ size 2147483648
gptq_model-4bit-128g-10-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6530530faa15ce11caf89cdd41c90f52279d0a1b035e43b4ddcd28bf385dc6d7
3
+ size 2147483648
gptq_model-4bit-128g-11-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00a401a36680ae4ee83a315485ca2be4d3031186d444e3804142ab3bda392c23
3
+ size 2147483648
gptq_model-4bit-128g-12-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15cc7f311902cb052132098ed97793785cc83962f7fcd8bf486d4d4cede6518d
3
+ size 2147483648
gptq_model-4bit-128g-13-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:701d62948b1c21742c0f5801a5e559bb65b770745d3f877b18b48bb8d9e56d84
3
+ size 2147483648
gptq_model-4bit-128g-14-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fcd47b04db679b9bfe4e9c46d6fd9d93ca4da129e5675186e5004d8644dd9234
3
+ size 2147483648
gptq_model-4bit-128g-15-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5e166df3bdde912ea69528844833493343a4d04942f8d7ccda331988f82ba43
3
+ size 2147483648
gptq_model-4bit-128g-16-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3676b58447bf891f34e768edcb27118734c6fefc8ddca7c0667f54ac92b967c
3
+ size 2147483648
gptq_model-4bit-128g-17-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f513cf35e96535db2cc08e327561bca9e24542e8751c0709524aa1b1469fd42
3
+ size 2147483648
gptq_model-4bit-128g-18-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:429296cda16a716346da5d7a0e4152828b678ebf7a80c8e517268b2e41b2f7e8
3
+ size 2147483648
gptq_model-4bit-128g-19-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b8846a8f6d591b7de0114b3f196ac2d9972d2594c06a5724b58ce28451a6e4d
3
+ size 2147483648
gptq_model-4bit-128g-20-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fee6c7b17ac6fe6f277911476548d526b329d6a12960985af0f8171f14aa6802
3
+ size 2147483648
gptq_model-4bit-128g-21-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:617e93a4903b05ca26c33a6c5fbde89620e6119bcda34cfb42217f3adc08ad9f
3
+ size 2147483648
gptq_model-4bit-128g-22-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b65490779c94850c126e6e59aeb1142d74e2e68cfc59ff1a0b2fcb29edefd91a
3
+ size 2147483648
gptq_model-4bit-128g-23-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:848766f3f324511fba260bfa8596b00c0fbacfcd1912e6e1ea5c93d41f54d6ca
3
+ size 2147483648
gptq_model-4bit-128g-24-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:971da4a699fc86f164a4dd1219e2b0171f50a5f40a951c2d4db63b3e98a6057b
3
+ size 2147483648
gptq_model-4bit-128g-25-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a03d7028544cedd575379a9e0ce381e12b6b11da8bd0832cc8a24c22955d6088
3
+ size 2147483648
gptq_model-4bit-128g-26-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96e36271b45e5f66eaeb55515229159c57df4a847451757a0ed2d3bfd53b5650
3
+ size 2147483648
gptq_model-4bit-128g-27-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e5fd5135f7e2322affdfcc072ef071c307d7a8c7648e13c5832c96ab85616d3
3
+ size 2147483648
gptq_model-4bit-128g-28-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6a0bc2de6269f7cc3c7706a557ae4f1a55553b7d54cdfd1881608705ce496ac
3
+ size 2147483648
gptq_model-4bit-128g-29-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4120b66a667b669d2f12d273ac24a00c5ca4b1509cd40873ee7beab26ba9101a
3
+ size 2147483648
gptq_model-4bit-128g-30-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ca4f976260c053e6493d02b74e08dbd9fa7b59dc2d2424f4bd560b20f8cc2f0
3
+ size 2147483648
gptq_model-4bit-128g-31-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1749d1c7c3f580e46ab4d679d6eea8b9c269e70c22d274dc0ea6a58c6b1c6680
3
+ size 2147483648
gptq_model-4bit-128g-32-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e29fb4100ea5a1eae2b87b239dd46be5e56911838580315f730d89f18981442a
3
+ size 2147483648
gptq_model-4bit-128g-33-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59cb90f2eec5df5bc0d169988ac1a1882e16057197e7b6806e0a93b05bbc3b2d
3
+ size 2147483648
gptq_model-4bit-128g-34-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a68bd7e066c026c89c11adec619c390f975f79a30dbbd4f5d2cdc47b9f4a763b
3
+ size 2147483648
gptq_model-4bit-128g-35-of-35.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:792e7d6f7e61ee45910caa7fc59bba1d2fd7b02f99b5a8892170866c131e19dc
3
+ size 723987688
quantize_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "group_size": 128,
4
+ "damp_percent": 0.01,
5
+ "desc_act": false,
6
+ "static_groups": false,
7
+ "sym": true,
8
+ "true_sequential": true,
9
+ "model_name_or_path": null,
10
+ "model_file_base_name": null,
11
+ "is_marlin_format": false,
12
+ "quant_method": "gptq"
13
+ }