mlinmg commited on
Commit
61ad7ac
·
verified ·
1 Parent(s): 645c947

Upload Qwen2ForCausalLM

Browse files
config.json CHANGED
@@ -38,7 +38,7 @@
38
  }
39
  }
40
  },
41
- "format": "pack-quantized",
42
  "global_compression_ratio": 1.9011197903965849,
43
  "ignore": [
44
  "lm_head"
@@ -48,461 +48,14 @@
48
  "quantization_status": "compressed",
49
  "sparsity_config": {
50
  "format": "dense",
51
- "global_sparsity": 0.18732520751668438,
52
  "ignore": [
53
- "model.layers.0.self_attn.q_proj",
54
- "model.layers.0.self_attn.k_proj",
55
- "model.layers.0.self_attn.v_proj",
56
- "model.layers.0.self_attn.o_proj",
57
- "model.layers.0.mlp.gate_proj",
58
- "model.layers.0.mlp.up_proj",
59
- "model.layers.0.mlp.down_proj",
60
- "model.layers.1.self_attn.o_proj",
61
- "model.layers.2.self_attn.o_proj",
62
- "model.layers.3.self_attn.o_proj",
63
- "model.layers.3.mlp.up_proj",
64
- "model.layers.4.self_attn.v_proj",
65
- "model.layers.4.self_attn.o_proj",
66
- "model.layers.4.mlp.gate_proj",
67
- "model.layers.4.mlp.up_proj",
68
- "model.layers.5.self_attn.q_proj",
69
- "model.layers.5.self_attn.k_proj",
70
- "model.layers.5.self_attn.v_proj",
71
- "model.layers.5.self_attn.o_proj",
72
- "model.layers.5.mlp.gate_proj",
73
- "model.layers.5.mlp.up_proj",
74
- "model.layers.5.mlp.down_proj",
75
- "model.layers.6.self_attn.q_proj",
76
- "model.layers.6.self_attn.k_proj",
77
- "model.layers.6.self_attn.v_proj",
78
- "model.layers.6.self_attn.o_proj",
79
- "model.layers.6.mlp.gate_proj",
80
- "model.layers.6.mlp.up_proj",
81
- "model.layers.6.mlp.down_proj",
82
- "model.layers.7.self_attn.q_proj",
83
- "model.layers.7.self_attn.k_proj",
84
- "model.layers.7.self_attn.v_proj",
85
- "model.layers.7.self_attn.o_proj",
86
- "model.layers.7.mlp.gate_proj",
87
- "model.layers.7.mlp.up_proj",
88
- "model.layers.7.mlp.down_proj",
89
- "model.layers.8.self_attn.q_proj",
90
- "model.layers.8.self_attn.k_proj",
91
- "model.layers.8.self_attn.v_proj",
92
- "model.layers.8.self_attn.o_proj",
93
- "model.layers.8.mlp.gate_proj",
94
- "model.layers.8.mlp.up_proj",
95
- "model.layers.8.mlp.down_proj",
96
- "model.layers.9.self_attn.q_proj",
97
- "model.layers.9.self_attn.k_proj",
98
- "model.layers.9.self_attn.v_proj",
99
- "model.layers.9.self_attn.o_proj",
100
- "model.layers.9.mlp.gate_proj",
101
- "model.layers.9.mlp.up_proj",
102
- "model.layers.9.mlp.down_proj",
103
- "model.layers.10.self_attn.q_proj",
104
- "model.layers.10.self_attn.k_proj",
105
- "model.layers.10.self_attn.v_proj",
106
- "model.layers.10.self_attn.o_proj",
107
- "model.layers.10.mlp.gate_proj",
108
- "model.layers.10.mlp.up_proj",
109
- "model.layers.10.mlp.down_proj",
110
- "model.layers.11.self_attn.q_proj",
111
- "model.layers.11.self_attn.k_proj",
112
- "model.layers.11.self_attn.v_proj",
113
- "model.layers.11.self_attn.o_proj",
114
- "model.layers.11.mlp.gate_proj",
115
- "model.layers.11.mlp.up_proj",
116
- "model.layers.11.mlp.down_proj",
117
- "model.layers.12.self_attn.q_proj",
118
- "model.layers.12.self_attn.k_proj",
119
- "model.layers.12.self_attn.v_proj",
120
- "model.layers.12.self_attn.o_proj",
121
- "model.layers.12.mlp.gate_proj",
122
- "model.layers.12.mlp.up_proj",
123
- "model.layers.12.mlp.down_proj",
124
- "model.layers.13.self_attn.q_proj",
125
- "model.layers.13.self_attn.k_proj",
126
- "model.layers.13.self_attn.v_proj",
127
- "model.layers.13.self_attn.o_proj",
128
- "model.layers.13.mlp.gate_proj",
129
- "model.layers.13.mlp.up_proj",
130
- "model.layers.13.mlp.down_proj",
131
- "model.layers.14.self_attn.q_proj",
132
- "model.layers.14.self_attn.k_proj",
133
- "model.layers.14.self_attn.v_proj",
134
- "model.layers.14.self_attn.o_proj",
135
- "model.layers.14.mlp.gate_proj",
136
- "model.layers.14.mlp.up_proj",
137
- "model.layers.14.mlp.down_proj",
138
- "model.layers.15.self_attn.q_proj",
139
- "model.layers.15.self_attn.k_proj",
140
- "model.layers.15.self_attn.v_proj",
141
- "model.layers.15.self_attn.o_proj",
142
- "model.layers.15.mlp.gate_proj",
143
- "model.layers.15.mlp.up_proj",
144
- "model.layers.15.mlp.down_proj",
145
- "model.layers.16.self_attn.q_proj",
146
- "model.layers.16.self_attn.k_proj",
147
- "model.layers.16.self_attn.v_proj",
148
- "model.layers.16.self_attn.o_proj",
149
- "model.layers.16.mlp.gate_proj",
150
- "model.layers.16.mlp.up_proj",
151
- "model.layers.16.mlp.down_proj",
152
- "model.layers.17.self_attn.q_proj",
153
- "model.layers.17.self_attn.k_proj",
154
- "model.layers.17.self_attn.v_proj",
155
- "model.layers.17.self_attn.o_proj",
156
- "model.layers.17.mlp.gate_proj",
157
- "model.layers.17.mlp.up_proj",
158
- "model.layers.17.mlp.down_proj",
159
- "model.layers.18.self_attn.q_proj",
160
- "model.layers.18.self_attn.k_proj",
161
- "model.layers.18.self_attn.v_proj",
162
- "model.layers.18.self_attn.o_proj",
163
- "model.layers.18.mlp.gate_proj",
164
- "model.layers.18.mlp.up_proj",
165
- "model.layers.18.mlp.down_proj",
166
- "model.layers.19.self_attn.q_proj",
167
- "model.layers.19.self_attn.k_proj",
168
- "model.layers.19.self_attn.v_proj",
169
- "model.layers.19.self_attn.o_proj",
170
- "model.layers.19.mlp.gate_proj",
171
- "model.layers.19.mlp.up_proj",
172
- "model.layers.19.mlp.down_proj",
173
- "model.layers.20.self_attn.q_proj",
174
- "model.layers.20.self_attn.k_proj",
175
- "model.layers.20.self_attn.v_proj",
176
- "model.layers.20.self_attn.o_proj",
177
- "model.layers.20.mlp.gate_proj",
178
- "model.layers.20.mlp.up_proj",
179
- "model.layers.20.mlp.down_proj",
180
- "model.layers.21.self_attn.q_proj",
181
- "model.layers.21.self_attn.k_proj",
182
- "model.layers.21.self_attn.v_proj",
183
- "model.layers.21.self_attn.o_proj",
184
- "model.layers.21.mlp.gate_proj",
185
- "model.layers.21.mlp.up_proj",
186
- "model.layers.21.mlp.down_proj",
187
- "model.layers.22.self_attn.q_proj",
188
- "model.layers.22.self_attn.k_proj",
189
- "model.layers.22.self_attn.v_proj",
190
- "model.layers.22.self_attn.o_proj",
191
- "model.layers.22.mlp.gate_proj",
192
- "model.layers.22.mlp.up_proj",
193
- "model.layers.22.mlp.down_proj",
194
- "model.layers.23.self_attn.q_proj",
195
- "model.layers.23.self_attn.k_proj",
196
- "model.layers.23.self_attn.v_proj",
197
- "model.layers.23.self_attn.o_proj",
198
- "model.layers.23.mlp.gate_proj",
199
- "model.layers.23.mlp.up_proj",
200
- "model.layers.23.mlp.down_proj",
201
- "model.layers.24.self_attn.q_proj",
202
- "model.layers.24.self_attn.k_proj",
203
- "model.layers.24.self_attn.v_proj",
204
- "model.layers.24.self_attn.o_proj",
205
- "model.layers.24.mlp.gate_proj",
206
- "model.layers.24.mlp.up_proj",
207
- "model.layers.24.mlp.down_proj",
208
- "model.layers.25.self_attn.q_proj",
209
- "model.layers.25.self_attn.k_proj",
210
- "model.layers.25.self_attn.v_proj",
211
- "model.layers.25.self_attn.o_proj",
212
- "model.layers.25.mlp.gate_proj",
213
- "model.layers.25.mlp.up_proj",
214
- "model.layers.25.mlp.down_proj",
215
- "model.layers.26.self_attn.q_proj",
216
- "model.layers.26.self_attn.k_proj",
217
- "model.layers.26.self_attn.v_proj",
218
- "model.layers.26.self_attn.o_proj",
219
- "model.layers.26.mlp.gate_proj",
220
- "model.layers.26.mlp.up_proj",
221
- "model.layers.26.mlp.down_proj",
222
- "model.layers.27.self_attn.q_proj",
223
- "model.layers.27.self_attn.k_proj",
224
- "model.layers.27.self_attn.v_proj",
225
- "model.layers.27.self_attn.o_proj",
226
- "model.layers.27.mlp.gate_proj",
227
- "model.layers.27.mlp.up_proj",
228
- "model.layers.27.mlp.down_proj",
229
- "model.layers.28.self_attn.q_proj",
230
- "model.layers.28.self_attn.k_proj",
231
- "model.layers.28.self_attn.v_proj",
232
- "model.layers.28.self_attn.o_proj",
233
- "model.layers.28.mlp.gate_proj",
234
- "model.layers.28.mlp.up_proj",
235
- "model.layers.28.mlp.down_proj",
236
- "model.layers.29.self_attn.q_proj",
237
- "model.layers.29.self_attn.k_proj",
238
- "model.layers.29.self_attn.v_proj",
239
- "model.layers.29.self_attn.o_proj",
240
- "model.layers.29.mlp.gate_proj",
241
- "model.layers.29.mlp.up_proj",
242
- "model.layers.29.mlp.down_proj",
243
- "model.layers.30.self_attn.q_proj",
244
- "model.layers.30.self_attn.k_proj",
245
- "model.layers.30.self_attn.v_proj",
246
- "model.layers.30.self_attn.o_proj",
247
- "model.layers.30.mlp.gate_proj",
248
- "model.layers.30.mlp.up_proj",
249
- "model.layers.30.mlp.down_proj",
250
- "model.layers.31.self_attn.q_proj",
251
- "model.layers.31.self_attn.k_proj",
252
- "model.layers.31.self_attn.v_proj",
253
- "model.layers.31.self_attn.o_proj",
254
- "model.layers.31.mlp.gate_proj",
255
- "model.layers.31.mlp.up_proj",
256
- "model.layers.31.mlp.down_proj",
257
- "model.layers.32.self_attn.q_proj",
258
- "model.layers.32.self_attn.k_proj",
259
- "model.layers.32.self_attn.v_proj",
260
- "model.layers.32.self_attn.o_proj",
261
- "model.layers.32.mlp.gate_proj",
262
- "model.layers.32.mlp.up_proj",
263
- "model.layers.32.mlp.down_proj",
264
- "model.layers.33.self_attn.q_proj",
265
- "model.layers.33.self_attn.k_proj",
266
- "model.layers.33.self_attn.v_proj",
267
- "model.layers.33.self_attn.o_proj",
268
- "model.layers.33.mlp.gate_proj",
269
- "model.layers.33.mlp.up_proj",
270
- "model.layers.33.mlp.down_proj",
271
- "model.layers.34.self_attn.q_proj",
272
- "model.layers.34.self_attn.k_proj",
273
- "model.layers.34.self_attn.v_proj",
274
- "model.layers.34.self_attn.o_proj",
275
- "model.layers.34.mlp.gate_proj",
276
- "model.layers.34.mlp.up_proj",
277
- "model.layers.34.mlp.down_proj",
278
- "model.layers.35.self_attn.q_proj",
279
- "model.layers.35.self_attn.k_proj",
280
- "model.layers.35.self_attn.v_proj",
281
- "model.layers.35.self_attn.o_proj",
282
- "model.layers.35.mlp.gate_proj",
283
- "model.layers.35.mlp.up_proj",
284
- "model.layers.35.mlp.down_proj",
285
- "model.layers.36.self_attn.q_proj",
286
- "model.layers.36.self_attn.k_proj",
287
- "model.layers.36.self_attn.v_proj",
288
- "model.layers.36.self_attn.o_proj",
289
- "model.layers.36.mlp.gate_proj",
290
- "model.layers.36.mlp.up_proj",
291
- "model.layers.36.mlp.down_proj",
292
- "model.layers.37.self_attn.q_proj",
293
- "model.layers.37.self_attn.k_proj",
294
- "model.layers.37.self_attn.v_proj",
295
- "model.layers.37.self_attn.o_proj",
296
- "model.layers.37.mlp.gate_proj",
297
- "model.layers.37.mlp.up_proj",
298
- "model.layers.37.mlp.down_proj",
299
- "model.layers.38.self_attn.q_proj",
300
- "model.layers.38.self_attn.k_proj",
301
- "model.layers.38.self_attn.v_proj",
302
- "model.layers.38.self_attn.o_proj",
303
- "model.layers.38.mlp.gate_proj",
304
- "model.layers.38.mlp.up_proj",
305
- "model.layers.38.mlp.down_proj",
306
- "model.layers.39.self_attn.q_proj",
307
- "model.layers.39.self_attn.k_proj",
308
- "model.layers.39.self_attn.v_proj",
309
- "model.layers.39.self_attn.o_proj",
310
- "model.layers.39.mlp.gate_proj",
311
- "model.layers.39.mlp.up_proj",
312
- "model.layers.39.mlp.down_proj",
313
- "model.layers.40.self_attn.q_proj",
314
- "model.layers.40.self_attn.k_proj",
315
- "model.layers.40.self_attn.v_proj",
316
- "model.layers.40.self_attn.o_proj",
317
- "model.layers.40.mlp.gate_proj",
318
- "model.layers.40.mlp.up_proj",
319
- "model.layers.40.mlp.down_proj",
320
- "model.layers.41.self_attn.q_proj",
321
- "model.layers.41.self_attn.k_proj",
322
- "model.layers.41.self_attn.v_proj",
323
- "model.layers.41.self_attn.o_proj",
324
- "model.layers.41.mlp.gate_proj",
325
- "model.layers.41.mlp.up_proj",
326
- "model.layers.41.mlp.down_proj",
327
- "model.layers.42.self_attn.q_proj",
328
- "model.layers.42.self_attn.k_proj",
329
- "model.layers.42.self_attn.v_proj",
330
- "model.layers.42.self_attn.o_proj",
331
- "model.layers.42.mlp.gate_proj",
332
- "model.layers.42.mlp.up_proj",
333
- "model.layers.42.mlp.down_proj",
334
- "model.layers.43.self_attn.q_proj",
335
- "model.layers.43.self_attn.k_proj",
336
- "model.layers.43.self_attn.v_proj",
337
- "model.layers.43.self_attn.o_proj",
338
- "model.layers.43.mlp.gate_proj",
339
- "model.layers.43.mlp.up_proj",
340
- "model.layers.43.mlp.down_proj",
341
- "model.layers.44.self_attn.q_proj",
342
- "model.layers.44.self_attn.k_proj",
343
- "model.layers.44.self_attn.v_proj",
344
- "model.layers.44.self_attn.o_proj",
345
- "model.layers.44.mlp.gate_proj",
346
- "model.layers.44.mlp.up_proj",
347
- "model.layers.44.mlp.down_proj",
348
- "model.layers.45.self_attn.q_proj",
349
- "model.layers.45.self_attn.k_proj",
350
- "model.layers.45.self_attn.v_proj",
351
- "model.layers.45.self_attn.o_proj",
352
- "model.layers.45.mlp.gate_proj",
353
- "model.layers.45.mlp.up_proj",
354
- "model.layers.45.mlp.down_proj",
355
- "model.layers.46.self_attn.q_proj",
356
- "model.layers.46.self_attn.k_proj",
357
- "model.layers.46.self_attn.v_proj",
358
- "model.layers.46.self_attn.o_proj",
359
- "model.layers.46.mlp.gate_proj",
360
- "model.layers.46.mlp.up_proj",
361
- "model.layers.46.mlp.down_proj",
362
- "model.layers.47.self_attn.q_proj",
363
- "model.layers.47.self_attn.k_proj",
364
- "model.layers.47.self_attn.v_proj",
365
- "model.layers.47.self_attn.o_proj",
366
- "model.layers.47.mlp.gate_proj",
367
- "model.layers.47.mlp.up_proj",
368
- "model.layers.47.mlp.down_proj",
369
- "model.layers.48.self_attn.q_proj",
370
- "model.layers.48.self_attn.k_proj",
371
- "model.layers.48.self_attn.v_proj",
372
- "model.layers.48.self_attn.o_proj",
373
- "model.layers.48.mlp.gate_proj",
374
- "model.layers.48.mlp.up_proj",
375
- "model.layers.48.mlp.down_proj",
376
- "model.layers.49.self_attn.q_proj",
377
- "model.layers.49.self_attn.k_proj",
378
- "model.layers.49.self_attn.v_proj",
379
- "model.layers.49.self_attn.o_proj",
380
- "model.layers.49.mlp.gate_proj",
381
- "model.layers.49.mlp.up_proj",
382
- "model.layers.49.mlp.down_proj",
383
- "model.layers.50.self_attn.q_proj",
384
- "model.layers.50.self_attn.k_proj",
385
- "model.layers.50.self_attn.v_proj",
386
- "model.layers.50.self_attn.o_proj",
387
- "model.layers.50.mlp.gate_proj",
388
- "model.layers.50.mlp.up_proj",
389
- "model.layers.50.mlp.down_proj",
390
- "model.layers.51.self_attn.q_proj",
391
- "model.layers.51.self_attn.k_proj",
392
- "model.layers.51.self_attn.v_proj",
393
- "model.layers.51.self_attn.o_proj",
394
- "model.layers.51.mlp.gate_proj",
395
- "model.layers.51.mlp.up_proj",
396
- "model.layers.51.mlp.down_proj",
397
- "model.layers.52.self_attn.q_proj",
398
- "model.layers.52.self_attn.k_proj",
399
- "model.layers.52.self_attn.v_proj",
400
- "model.layers.52.self_attn.o_proj",
401
- "model.layers.52.mlp.gate_proj",
402
- "model.layers.52.mlp.up_proj",
403
- "model.layers.52.mlp.down_proj",
404
- "model.layers.53.self_attn.q_proj",
405
- "model.layers.53.self_attn.k_proj",
406
- "model.layers.53.self_attn.v_proj",
407
- "model.layers.53.self_attn.o_proj",
408
- "model.layers.53.mlp.gate_proj",
409
- "model.layers.53.mlp.up_proj",
410
- "model.layers.53.mlp.down_proj",
411
- "model.layers.54.self_attn.q_proj",
412
- "model.layers.54.self_attn.k_proj",
413
- "model.layers.54.self_attn.v_proj",
414
- "model.layers.54.self_attn.o_proj",
415
- "model.layers.54.mlp.gate_proj",
416
- "model.layers.54.mlp.up_proj",
417
- "model.layers.54.mlp.down_proj",
418
- "model.layers.55.self_attn.q_proj",
419
- "model.layers.55.self_attn.k_proj",
420
- "model.layers.55.self_attn.v_proj",
421
- "model.layers.55.self_attn.o_proj",
422
- "model.layers.55.mlp.gate_proj",
423
- "model.layers.55.mlp.up_proj",
424
- "model.layers.55.mlp.down_proj",
425
- "model.layers.56.self_attn.q_proj",
426
- "model.layers.56.self_attn.k_proj",
427
- "model.layers.56.self_attn.v_proj",
428
- "model.layers.56.self_attn.o_proj",
429
- "model.layers.56.mlp.gate_proj",
430
- "model.layers.56.mlp.up_proj",
431
- "model.layers.56.mlp.down_proj",
432
- "model.layers.57.self_attn.q_proj",
433
- "model.layers.57.self_attn.k_proj",
434
- "model.layers.57.self_attn.v_proj",
435
- "model.layers.57.self_attn.o_proj",
436
- "model.layers.57.mlp.gate_proj",
437
- "model.layers.57.mlp.up_proj",
438
- "model.layers.57.mlp.down_proj",
439
- "model.layers.58.self_attn.q_proj",
440
- "model.layers.58.self_attn.k_proj",
441
- "model.layers.58.self_attn.v_proj",
442
- "model.layers.58.self_attn.o_proj",
443
- "model.layers.58.mlp.gate_proj",
444
- "model.layers.58.mlp.up_proj",
445
- "model.layers.58.mlp.down_proj",
446
- "model.layers.59.self_attn.q_proj",
447
- "model.layers.59.self_attn.k_proj",
448
- "model.layers.59.self_attn.v_proj",
449
- "model.layers.59.self_attn.o_proj",
450
- "model.layers.59.mlp.gate_proj",
451
- "model.layers.59.mlp.up_proj",
452
- "model.layers.59.mlp.down_proj",
453
- "model.layers.60.self_attn.q_proj",
454
- "model.layers.60.self_attn.k_proj",
455
- "model.layers.60.self_attn.v_proj",
456
- "model.layers.60.self_attn.o_proj",
457
- "model.layers.60.mlp.gate_proj",
458
- "model.layers.60.mlp.up_proj",
459
- "model.layers.60.mlp.down_proj",
460
- "model.layers.61.self_attn.q_proj",
461
- "model.layers.61.self_attn.k_proj",
462
- "model.layers.61.self_attn.v_proj",
463
- "model.layers.61.self_attn.o_proj",
464
- "model.layers.61.mlp.gate_proj",
465
- "model.layers.61.mlp.up_proj",
466
- "model.layers.61.mlp.down_proj",
467
- "model.layers.62.self_attn.q_proj",
468
- "model.layers.62.self_attn.k_proj",
469
- "model.layers.62.self_attn.v_proj",
470
- "model.layers.62.self_attn.o_proj",
471
- "model.layers.62.mlp.gate_proj",
472
- "model.layers.62.mlp.up_proj",
473
- "model.layers.62.mlp.down_proj",
474
- "model.layers.63.self_attn.q_proj",
475
- "model.layers.63.self_attn.k_proj",
476
- "model.layers.63.self_attn.v_proj",
477
- "model.layers.63.self_attn.o_proj",
478
- "model.layers.63.mlp.gate_proj",
479
- "model.layers.63.mlp.up_proj",
480
- "model.layers.63.mlp.down_proj",
481
  "lm_head"
482
  ],
483
  "registry_requires_subclass": false,
484
- "sparsity_structure": "unstructured",
485
  "targets": [
486
- "model.layers.1.self_attn.q_proj",
487
- "model.layers.1.self_attn.k_proj",
488
- "model.layers.1.self_attn.v_proj",
489
- "model.layers.1.mlp.gate_proj",
490
- "model.layers.1.mlp.up_proj",
491
- "model.layers.1.mlp.down_proj",
492
- "model.layers.2.self_attn.q_proj",
493
- "model.layers.2.self_attn.k_proj",
494
- "model.layers.2.self_attn.v_proj",
495
- "model.layers.2.mlp.gate_proj",
496
- "model.layers.2.mlp.up_proj",
497
- "model.layers.2.mlp.down_proj",
498
- "model.layers.3.self_attn.q_proj",
499
- "model.layers.3.self_attn.k_proj",
500
- "model.layers.3.self_attn.v_proj",
501
- "model.layers.3.mlp.gate_proj",
502
- "model.layers.3.mlp.down_proj",
503
- "model.layers.4.self_attn.q_proj",
504
- "model.layers.4.self_attn.k_proj",
505
- "model.layers.4.mlp.down_proj"
506
  ]
507
  }
508
  },
 
38
  }
39
  }
40
  },
41
+ "format": "marlin-24",
42
  "global_compression_ratio": 1.9011197903965849,
43
  "ignore": [
44
  "lm_head"
 
48
  "quantization_status": "compressed",
49
  "sparsity_config": {
50
  "format": "dense",
51
+ "global_sparsity": 0.49413722527001075,
52
  "ignore": [
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  "lm_head"
54
  ],
55
  "registry_requires_subclass": false,
56
+ "sparsity_structure": "2:4",
57
  "targets": [
58
+ "Linear"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
59
  ]
60
  }
61
  },
model-00001-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:951eb6fd48d8cc2d1a1dcb5d549c57b71a7b417b01e7550b6d742ccf7aca1c89
3
- size 4931437664
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:296340898b17238476619bee829eade6d82bba5c68a39854662f504d4f2afe96
3
+ size 4999862864
model-00002-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a38b41070bf05f9f0e94b1085dddf0ccac81ae60a7081d424f525756825998fc
3
- size 4956016432
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1dab1ce4d122fe437788083b24ae401d02b68286b8deb056da3fdebd7bd5873
3
+ size 4966158448
model-00003-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:79bd5a0c413ae8dd48acf95da56d302da54ae06ea7e6efe542c7073fa414d787
3
- size 4996551064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a80652d354816cb466ecba68b19848e6bb2675f01070f37e8b05b47819f0f47
3
+ size 3783243552
model-00004-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7075a67bf638da1d6ea87be9b6dbcdfe4e8070338fa4cf8813687e84e1e829f9
3
- size 4323099720
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9309786b070bd37d20a3e80a30303da5bb815cf20ee55aa93382adaf65fe41ca
3
+ size 1557135488
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff
 
recipe.yaml CHANGED
@@ -1,6 +1,7 @@
1
  DEFAULT_stage:
2
  DEFAULT_modifiers:
3
  SmoothQuantModifier: {smoothing_strength: 0.65}
 
4
  GPTQModifier:
5
  block_size: 64
6
  dampening_frac: 0.1
 
1
  DEFAULT_stage:
2
  DEFAULT_modifiers:
3
  SmoothQuantModifier: {smoothing_strength: 0.65}
4
+ WandaPruningModifier: {sparsity: 0.5, mask_structure: '2:4', sequential_update: false}
5
  GPTQModifier:
6
  block_size: 64
7
  dampening_frac: 0.1