Xiaowen-dg commited on
Commit
930d2e8
1 Parent(s): ffd25ab

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +654 -2
README.md CHANGED
@@ -15,8 +15,660 @@ extra_gated_description: If you want to learn more about how we process your per
15
  data, please read our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
16
  model-index:
17
  - name: Mistral-Nemo-Instruct-2407
18
- results: []
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19
  ---
 
 
 
 
 
 
20
 
21
  # Model Card for Mistral-Nemo-Instruct-2407
22
 
@@ -271,4 +923,4 @@ make the model finely respect guardrails, allowing for deployment in environment
271
 
272
  ## The Mistral AI Team
273
 
274
- Albert Jiang, Alexandre Sablayrolles, Alexis Tacnet, Alok Kothari, Antoine Roux, Arthur Mensch, Audrey Herblin-Stoop, Augustin Garreau, Austin Birky, Bam4d, Baptiste Bout, Baudouin de Monicault, Blanche Savary, Carole Rambaud, Caroline Feldman, Devendra Singh Chaplot, Diego de las Casas, Eleonore Arcelin, Emma Bou Hanna, Etienne Metzger, Gaspard Blanchet, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Harizo Rajaona, Henri Roussez, Hichem Sattouf, Ian Mack, Jean-Malo Delignon, Jessica Chudnovsky, Justus Murke, Kartik Khandelwal, Lawrence Stewart, Louis Martin, Louis Ternon, Lucile Saulnier, Lélio Renard Lavaud, Margaret Jennings, Marie Pellat, Marie Torelli, Marie-Anne Lachaux, Marjorie Janiewicz, Mickaël Seznec, Nicolas Schuhl, Niklas Muhs, Olivier de Garrigues, Patrick von Platen, Paul Jacob, Pauline Buche, Pavan Kumar Reddy, Perry Savas, Pierre Stock, Romain Sauvestre, Sagar Vaze, Sandeep Subramanian, Saurabh Garg, Sophia Yang, Szymon Antoniak, Teven Le Scao, Thibault Schueller, Thibaut Lavril, Thomas Wang, Théophile Gervet, Timothée Lacroix, Valera Nemychnikova, Wendy Shang, William El Sayed, William Marshall
 
15
  data, please read our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
16
  model-index:
17
  - name: Mistral-Nemo-Instruct-2407
18
+ results:
19
+ - task:
20
+ type: squad_answerable-judge
21
+ dataset:
22
+ name: squad_answerable
23
+ type: multi-choices
24
+ metrics:
25
+ - type: judge_match
26
+ value: '0.685'
27
+ args:
28
+ results:
29
+ squad_answerable-judge:
30
+ exact_match,strict_match: 0.6852522530110334
31
+ exact_match_stderr,strict_match: 0.004262305820311226
32
+ alias: squad_answerable-judge
33
+ context_has_answer-judge:
34
+ exact_match,strict_match: 0.7906976744186046
35
+ exact_match_stderr,strict_match: 0.04412480456048906
36
+ alias: context_has_answer-judge
37
+ group_subtasks:
38
+ context_has_answer-judge: []
39
+ squad_answerable-judge: []
40
+ configs:
41
+ context_has_answer-judge:
42
+ task: context_has_answer-judge
43
+ group: dg
44
+ dataset_path: DataGuard/eval-multi-choices
45
+ dataset_name: context_has_answer_judge
46
+ test_split: test
47
+ doc_to_text: '<s>[INST]You are asked to determine if a question has the
48
+ answer in the context, and answer with a simple Yes or No.
49
+
50
+
51
+ Example:
52
+
53
+ Question: How is the weather today? Context: How is the traffic today?
54
+ It is horrible. Does the question have the answer in the Context?
55
+
56
+ Answer: No
57
+
58
+ Question: How is the weather today? Context: Is the weather good today?
59
+ Yes, it is sunny. Does the question have the answer in the Context?
60
+
61
+ Answer: Yes
62
+
63
+
64
+ Question: {{question}}
65
+
66
+ Context: {{similar_question}} {{similar_answer}}
67
+
68
+ Does the question have the answer in the Context?
69
+
70
+ [/INST]'
71
+ doc_to_target: '{{''Yes'' if is_relevant in [''Yes'', 1] else ''No''}}'
72
+ description: ''
73
+ target_delimiter: ' '
74
+ fewshot_delimiter: '
75
+
76
+
77
+ '
78
+ metric_list:
79
+ - metric: exact_match
80
+ output_type: generate_until
81
+ generation_kwargs:
82
+ until:
83
+ - <|im_end|>
84
+ do_sample: false
85
+ temperature: 0.3
86
+ repeats: 1
87
+ filter_list:
88
+ - name: strict_match
89
+ filter:
90
+ - function: regex
91
+ regex_pattern: Yes|No
92
+ group_select: -1
93
+ - function: take_first
94
+ should_decontaminate: false
95
+ squad_answerable-judge:
96
+ task: squad_answerable-judge
97
+ group: dg
98
+ dataset_path: DataGuard/eval-multi-choices
99
+ dataset_name: squad_answerable_judge
100
+ test_split: test
101
+ doc_to_text: '<s>[INST]You are asked to determine if a question has the
102
+ answer in the context, and answer with a simple Yes or No.
103
+
104
+
105
+ Example:
106
+
107
+ Question: How is the weather today? Context: The traffic is horrible.
108
+ Does the question have the answer in the Context?
109
+
110
+ Answer: No
111
+
112
+ Question: How is the weather today? Context: The weather is good. Does
113
+ the question have the answer in the Context?
114
+
115
+ Answer: Yes
116
+
117
+
118
+ Question: {{question}}
119
+
120
+ Context: {{context}}
121
+
122
+ Does the question have the answer in the Context?
123
+
124
+ [/INST]'
125
+ doc_to_target: '{{''Yes'' if is_relevant in [''Yes'', 1] else ''No''}}'
126
+ description: ''
127
+ target_delimiter: ' '
128
+ fewshot_delimiter: '
129
+
130
+
131
+ '
132
+ metric_list:
133
+ - metric: exact_match
134
+ output_type: generate_until
135
+ generation_kwargs:
136
+ until:
137
+ - <|im_end|>
138
+ do_sample: false
139
+ temperature: 0.3
140
+ repeats: 1
141
+ filter_list:
142
+ - name: strict_match
143
+ filter:
144
+ - function: regex
145
+ regex_pattern: Yes|No
146
+ group_select: -1
147
+ - function: take_first
148
+ should_decontaminate: false
149
+ versions:
150
+ context_has_answer-judge: Yaml
151
+ squad_answerable-judge: Yaml
152
+ n-shot: {}
153
+ config:
154
+ model: vllm
155
+ model_args: pretrained=mistralai/Mistral-Nemo-Instruct-2407,tensor_parallel_size=1,dtype=auto,gpu_memory_utilization=0.8,max_model_len=2048,trust_remote_code=True
156
+ batch_size: auto
157
+ batch_sizes: []
158
+ bootstrap_iters: 100000
159
+ git_hash: cddf85d
160
+ pretty_env_info: 'PyTorch version: 2.4.0+cu121
161
+
162
+ Is debug build: False
163
+
164
+ CUDA used to build PyTorch: 12.1
165
+
166
+ ROCM used to build PyTorch: N/A
167
+
168
+
169
+ OS: Ubuntu 22.04.3 LTS (x86_64)
170
+
171
+ GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
172
+
173
+ Clang version: Could not collect
174
+
175
+ CMake version: version 3.25.0
176
+
177
+ Libc version: glibc-2.35
178
+
179
+
180
+ Python version: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] (64-bit
181
+ runtime)
182
+
183
+ Python platform: Linux-5.4.0-149-generic-x86_64-with-glibc2.35
184
+
185
+ Is CUDA available: True
186
+
187
+ CUDA runtime version: 11.8.89
188
+
189
+ CUDA_MODULE_LOADING set to: LAZY
190
+
191
+ GPU models and configuration: GPU 0: NVIDIA L40
192
+
193
+ Nvidia driver version: 535.54.03
194
+
195
+ cuDNN version: Could not collect
196
+
197
+ HIP runtime version: N/A
198
+
199
+ MIOpen runtime version: N/A
200
+
201
+ Is XNNPACK available: True
202
+
203
+
204
+ CPU:
205
+
206
+ Architecture: x86_64
207
+
208
+ CPU op-mode(s): 32-bit, 64-bit
209
+
210
+ Address sizes: 48 bits physical, 48 bits virtual
211
+
212
+ Byte Order: Little Endian
213
+
214
+ CPU(s): 256
215
+
216
+ On-line CPU(s) list: 0-254
217
+
218
+ Off-line CPU(s) list: 255
219
+
220
+ Vendor ID: AuthenticAMD
221
+
222
+ Model name: AMD EPYC 7773X 64-Core Processor
223
+
224
+ CPU family: 25
225
+
226
+ Model: 1
227
+
228
+ Thread(s) per core: 2
229
+
230
+ Core(s) per socket: 64
231
+
232
+ Socket(s): 2
233
+
234
+ Stepping: 2
235
+
236
+ Frequency boost: enabled
237
+
238
+ CPU max MHz: 2200.0000
239
+
240
+ CPU min MHz: 0.0000
241
+
242
+ BogoMIPS: 4400.14
243
+
244
+ Flags: fpu vme de pse tsc msr pae mce cx8 apic
245
+ sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx
246
+ mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc
247
+ cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1
248
+ sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy
249
+ svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit
250
+ wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3
251
+ cdp_l3 invpcid_single hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase
252
+ bmi1 avx2 smep bmi2 invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni
253
+ xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local
254
+ clzero irperf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale
255
+ vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload
256
+ vgif umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca
257
+
258
+ Virtualization: AMD-V
259
+
260
+ L1d cache: 4 MiB (128 instances)
261
+
262
+ L1i cache: 4 MiB (128 instances)
263
+
264
+ L2 cache: 64 MiB (128 instances)
265
+
266
+ L3 cache: 1.5 GiB (16 instances)
267
+
268
+ NUMA node(s): 16
269
+
270
+ NUMA node0 CPU(s): 0-7,128-135
271
+
272
+ NUMA node1 CPU(s): 8-15,136-143
273
+
274
+ NUMA node2 CPU(s): 16-23,144-151
275
+
276
+ NUMA node3 CPU(s): 24-31,152-159
277
+
278
+ NUMA node4 CPU(s): 32-39,160-167
279
+
280
+ NUMA node5 CPU(s): 40-47,168-175
281
+
282
+ NUMA node6 CPU(s): 48-55,176-183
283
+
284
+ NUMA node7 CPU(s): 56-63,184-191
285
+
286
+ NUMA node8 CPU(s): 64-71,192-199
287
+
288
+ NUMA node9 CPU(s): 72-79,200-207
289
+
290
+ NUMA node10 CPU(s): 80-87,208-215
291
+
292
+ NUMA node11 CPU(s): 88-95,216-223
293
+
294
+ NUMA node12 CPU(s): 96-103,224-231
295
+
296
+ NUMA node13 CPU(s): 104-111,232-239
297
+
298
+ NUMA node14 CPU(s): 112-119,240-247
299
+
300
+ NUMA node15 CPU(s): 120-127,248-254
301
+
302
+ Vulnerability Itlb multihit: Not affected
303
+
304
+ Vulnerability L1tf: Not affected
305
+
306
+ Vulnerability Mds: Not affected
307
+
308
+ Vulnerability Meltdown: Not affected
309
+
310
+ Vulnerability Mmio stale data: Not affected
311
+
312
+ Vulnerability Retbleed: Not affected
313
+
314
+ Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled
315
+ via prctl and seccomp
316
+
317
+ Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and
318
+ __user pointer sanitization
319
+
320
+ Vulnerability Spectre v2: Mitigation; Retpolines, IBPB conditional,
321
+ IBRS_FW, STIBP always-on, RSB filling, PBRSB-eIBRS Not affected
322
+
323
+ Vulnerability Srbds: Not affected
324
+
325
+ Vulnerability Tsx async abort: Not affected
326
+
327
+
328
+ Versions of relevant libraries:
329
+
330
+ [pip3] numpy==1.24.1
331
+
332
+ [pip3] torch==2.4.0
333
+
334
+ [pip3] torchaudio==2.0.2+cu118
335
+
336
+ [pip3] torchvision==0.19.0
337
+
338
+ [pip3] triton==3.0.0
339
+
340
+ [conda] Could not collect'
341
+ transformers_version: 4.44.1
342
+ - task:
343
+ type: context_has_answer-judge
344
+ dataset:
345
+ name: context_has_answer
346
+ type: multi-choices
347
+ metrics:
348
+ - type: judge_match
349
+ value: '0.791'
350
+ args:
351
+ results:
352
+ squad_answerable-judge:
353
+ exact_match,strict_match: 0.6852522530110334
354
+ exact_match_stderr,strict_match: 0.004262305820311226
355
+ alias: squad_answerable-judge
356
+ context_has_answer-judge:
357
+ exact_match,strict_match: 0.7906976744186046
358
+ exact_match_stderr,strict_match: 0.04412480456048906
359
+ alias: context_has_answer-judge
360
+ group_subtasks:
361
+ context_has_answer-judge: []
362
+ squad_answerable-judge: []
363
+ configs:
364
+ context_has_answer-judge:
365
+ task: context_has_answer-judge
366
+ group: dg
367
+ dataset_path: DataGuard/eval-multi-choices
368
+ dataset_name: context_has_answer_judge
369
+ test_split: test
370
+ doc_to_text: '<s>[INST]You are asked to determine if a question has the
371
+ answer in the context, and answer with a simple Yes or No.
372
+
373
+
374
+ Example:
375
+
376
+ Question: How is the weather today? Context: How is the traffic today?
377
+ It is horrible. Does the question have the answer in the Context?
378
+
379
+ Answer: No
380
+
381
+ Question: How is the weather today? Context: Is the weather good today?
382
+ Yes, it is sunny. Does the question have the answer in the Context?
383
+
384
+ Answer: Yes
385
+
386
+
387
+ Question: {{question}}
388
+
389
+ Context: {{similar_question}} {{similar_answer}}
390
+
391
+ Does the question have the answer in the Context?
392
+
393
+ [/INST]'
394
+ doc_to_target: '{{''Yes'' if is_relevant in [''Yes'', 1] else ''No''}}'
395
+ description: ''
396
+ target_delimiter: ' '
397
+ fewshot_delimiter: '
398
+
399
+
400
+ '
401
+ metric_list:
402
+ - metric: exact_match
403
+ output_type: generate_until
404
+ generation_kwargs:
405
+ until:
406
+ - <|im_end|>
407
+ do_sample: false
408
+ temperature: 0.3
409
+ repeats: 1
410
+ filter_list:
411
+ - name: strict_match
412
+ filter:
413
+ - function: regex
414
+ regex_pattern: Yes|No
415
+ group_select: -1
416
+ - function: take_first
417
+ should_decontaminate: false
418
+ squad_answerable-judge:
419
+ task: squad_answerable-judge
420
+ group: dg
421
+ dataset_path: DataGuard/eval-multi-choices
422
+ dataset_name: squad_answerable_judge
423
+ test_split: test
424
+ doc_to_text: '<s>[INST]You are asked to determine if a question has the
425
+ answer in the context, and answer with a simple Yes or No.
426
+
427
+
428
+ Example:
429
+
430
+ Question: How is the weather today? Context: The traffic is horrible.
431
+ Does the question have the answer in the Context?
432
+
433
+ Answer: No
434
+
435
+ Question: How is the weather today? Context: The weather is good. Does
436
+ the question have the answer in the Context?
437
+
438
+ Answer: Yes
439
+
440
+
441
+ Question: {{question}}
442
+
443
+ Context: {{context}}
444
+
445
+ Does the question have the answer in the Context?
446
+
447
+ [/INST]'
448
+ doc_to_target: '{{''Yes'' if is_relevant in [''Yes'', 1] else ''No''}}'
449
+ description: ''
450
+ target_delimiter: ' '
451
+ fewshot_delimiter: '
452
+
453
+
454
+ '
455
+ metric_list:
456
+ - metric: exact_match
457
+ output_type: generate_until
458
+ generation_kwargs:
459
+ until:
460
+ - <|im_end|>
461
+ do_sample: false
462
+ temperature: 0.3
463
+ repeats: 1
464
+ filter_list:
465
+ - name: strict_match
466
+ filter:
467
+ - function: regex
468
+ regex_pattern: Yes|No
469
+ group_select: -1
470
+ - function: take_first
471
+ should_decontaminate: false
472
+ versions:
473
+ context_has_answer-judge: Yaml
474
+ squad_answerable-judge: Yaml
475
+ n-shot: {}
476
+ config:
477
+ model: vllm
478
+ model_args: pretrained=mistralai/Mistral-Nemo-Instruct-2407,tensor_parallel_size=1,dtype=auto,gpu_memory_utilization=0.8,max_model_len=2048,trust_remote_code=True
479
+ batch_size: auto
480
+ batch_sizes: []
481
+ bootstrap_iters: 100000
482
+ git_hash: cddf85d
483
+ pretty_env_info: 'PyTorch version: 2.4.0+cu121
484
+
485
+ Is debug build: False
486
+
487
+ CUDA used to build PyTorch: 12.1
488
+
489
+ ROCM used to build PyTorch: N/A
490
+
491
+
492
+ OS: Ubuntu 22.04.3 LTS (x86_64)
493
+
494
+ GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
495
+
496
+ Clang version: Could not collect
497
+
498
+ CMake version: version 3.25.0
499
+
500
+ Libc version: glibc-2.35
501
+
502
+
503
+ Python version: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] (64-bit
504
+ runtime)
505
+
506
+ Python platform: Linux-5.4.0-149-generic-x86_64-with-glibc2.35
507
+
508
+ Is CUDA available: True
509
+
510
+ CUDA runtime version: 11.8.89
511
+
512
+ CUDA_MODULE_LOADING set to: LAZY
513
+
514
+ GPU models and configuration: GPU 0: NVIDIA L40
515
+
516
+ Nvidia driver version: 535.54.03
517
+
518
+ cuDNN version: Could not collect
519
+
520
+ HIP runtime version: N/A
521
+
522
+ MIOpen runtime version: N/A
523
+
524
+ Is XNNPACK available: True
525
+
526
+
527
+ CPU:
528
+
529
+ Architecture: x86_64
530
+
531
+ CPU op-mode(s): 32-bit, 64-bit
532
+
533
+ Address sizes: 48 bits physical, 48 bits virtual
534
+
535
+ Byte Order: Little Endian
536
+
537
+ CPU(s): 256
538
+
539
+ On-line CPU(s) list: 0-254
540
+
541
+ Off-line CPU(s) list: 255
542
+
543
+ Vendor ID: AuthenticAMD
544
+
545
+ Model name: AMD EPYC 7773X 64-Core Processor
546
+
547
+ CPU family: 25
548
+
549
+ Model: 1
550
+
551
+ Thread(s) per core: 2
552
+
553
+ Core(s) per socket: 64
554
+
555
+ Socket(s): 2
556
+
557
+ Stepping: 2
558
+
559
+ Frequency boost: enabled
560
+
561
+ CPU max MHz: 2200.0000
562
+
563
+ CPU min MHz: 0.0000
564
+
565
+ BogoMIPS: 4400.14
566
+
567
+ Flags: fpu vme de pse tsc msr pae mce cx8 apic
568
+ sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx
569
+ mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc
570
+ cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1
571
+ sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy
572
+ svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit
573
+ wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3
574
+ cdp_l3 invpcid_single hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase
575
+ bmi1 avx2 smep bmi2 invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni
576
+ xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local
577
+ clzero irperf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale
578
+ vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload
579
+ vgif umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca
580
+
581
+ Virtualization: AMD-V
582
+
583
+ L1d cache: 4 MiB (128 instances)
584
+
585
+ L1i cache: 4 MiB (128 instances)
586
+
587
+ L2 cache: 64 MiB (128 instances)
588
+
589
+ L3 cache: 1.5 GiB (16 instances)
590
+
591
+ NUMA node(s): 16
592
+
593
+ NUMA node0 CPU(s): 0-7,128-135
594
+
595
+ NUMA node1 CPU(s): 8-15,136-143
596
+
597
+ NUMA node2 CPU(s): 16-23,144-151
598
+
599
+ NUMA node3 CPU(s): 24-31,152-159
600
+
601
+ NUMA node4 CPU(s): 32-39,160-167
602
+
603
+ NUMA node5 CPU(s): 40-47,168-175
604
+
605
+ NUMA node6 CPU(s): 48-55,176-183
606
+
607
+ NUMA node7 CPU(s): 56-63,184-191
608
+
609
+ NUMA node8 CPU(s): 64-71,192-199
610
+
611
+ NUMA node9 CPU(s): 72-79,200-207
612
+
613
+ NUMA node10 CPU(s): 80-87,208-215
614
+
615
+ NUMA node11 CPU(s): 88-95,216-223
616
+
617
+ NUMA node12 CPU(s): 96-103,224-231
618
+
619
+ NUMA node13 CPU(s): 104-111,232-239
620
+
621
+ NUMA node14 CPU(s): 112-119,240-247
622
+
623
+ NUMA node15 CPU(s): 120-127,248-254
624
+
625
+ Vulnerability Itlb multihit: Not affected
626
+
627
+ Vulnerability L1tf: Not affected
628
+
629
+ Vulnerability Mds: Not affected
630
+
631
+ Vulnerability Meltdown: Not affected
632
+
633
+ Vulnerability Mmio stale data: Not affected
634
+
635
+ Vulnerability Retbleed: Not affected
636
+
637
+ Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled
638
+ via prctl and seccomp
639
+
640
+ Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and
641
+ __user pointer sanitization
642
+
643
+ Vulnerability Spectre v2: Mitigation; Retpolines, IBPB conditional,
644
+ IBRS_FW, STIBP always-on, RSB filling, PBRSB-eIBRS Not affected
645
+
646
+ Vulnerability Srbds: Not affected
647
+
648
+ Vulnerability Tsx async abort: Not affected
649
+
650
+
651
+ Versions of relevant libraries:
652
+
653
+ [pip3] numpy==1.24.1
654
+
655
+ [pip3] torch==2.4.0
656
+
657
+ [pip3] torchaudio==2.0.2+cu118
658
+
659
+ [pip3] torchvision==0.19.0
660
+
661
+ [pip3] triton==3.0.0
662
+
663
+ [conda] Could not collect'
664
+ transformers_version: 4.44.1
665
  ---
666
+ ### Needle in a Haystack Evaluation Heatmap
667
+
668
+ ![Needle in a Haystack Evaluation Heatmap EN](./niah_heatmap_en.png)
669
+
670
+ ![Needle in a Haystack Evaluation Heatmap DE](./niah_heatmap_de.png)
671
+
672
 
673
  # Model Card for Mistral-Nemo-Instruct-2407
674
 
 
923
 
924
  ## The Mistral AI Team
925
 
926
+ Albert Jiang, Alexandre Sablayrolles, Alexis Tacnet, Alok Kothari, Antoine Roux, Arthur Mensch, Audrey Herblin-Stoop, Augustin Garreau, Austin Birky, Bam4d, Baptiste Bout, Baudouin de Monicault, Blanche Savary, Carole Rambaud, Caroline Feldman, Devendra Singh Chaplot, Diego de las Casas, Eleonore Arcelin, Emma Bou Hanna, Etienne Metzger, Gaspard Blanchet, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Harizo Rajaona, Henri Roussez, Hichem Sattouf, Ian Mack, Jean-Malo Delignon, Jessica Chudnovsky, Justus Murke, Kartik Khandelwal, Lawrence Stewart, Louis Martin, Louis Ternon, Lucile Saulnier, Lélio Renard Lavaud, Margaret Jennings, Marie Pellat, Marie Torelli, Marie-Anne Lachaux, Marjorie Janiewicz, Mickaël Seznec, Nicolas Schuhl, Niklas Muhs, Olivier de Garrigues, Patrick von Platen, Paul Jacob, Pauline Buche, Pavan Kumar Reddy, Perry Savas, Pierre Stock, Romain Sauvestre, Sagar Vaze, Sandeep Subramanian, Saurabh Garg, Sophia Yang, Szymon Antoniak, Teven Le Scao, Thibault Schueller, Thibaut Lavril, Thomas Wang, Théophile Gervet, Timothée Lacroix, Valera Nemychnikova, Wendy Shang, William El Sayed, William Marshall