fhdz commited on
Commit
25731a6
1 Parent(s): 003bccf

Initial commit

Browse files
inference.yaml ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ transforms: [sentencepiece]
2
+
3
+ #### Subword
4
+ src_subword_model: "<model_dir>/tokenizer.model"
5
+ tgt_subword_model: "<model_dir>/tokenizer.model"
6
+
7
+ # Model info
8
+ model: "<model_dir>/mistral-instruct-v0.2-onmt-awq-gemv.pt"
9
+
10
+
11
+ # Inference
12
+ seed: 42
13
+ max_length: 256
14
+ gpu: 0
15
+ batch_type: sents
16
+ batch_size: 4
17
+ world_size: 1
18
+ gpu_ranks: [0]
19
+ #parallel_mode: "tensor_parallel"
20
+ precision: fp16
21
+ #random_sampling_topk: 1
22
+ #random_sampling_topp: 0.6
23
+ #random_sampling_temp: 0.9
24
+ beam_size: 1
25
+ n_best: 1
26
+ profile: false
27
+ report_time: true
28
+ src: None
mistral-instruct-v0.2-onmt-awq-gemv.00.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7c24e774166aa5bf5293f6a2c603336dfec29d4f3fdcf8733a2de679622d4bf
3
+ size 4150957864
mistral-instruct-v0.2-onmt-awq-gemv.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17a185e64df160df9049ef26adf67c1b80d00d5a64b497e19fa313f3d09fad19
3
+ size 1054160
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
vocab.txt ADDED
The diff for this file is too large to render. See raw diff