Model save
Browse files- README.md +14 -25
- adapter_config.json +9 -6
- adapter_model.safetensors +2 -2
- added_tokens.json +4 -0
- config.json +1 -1
- tokenizer.json +18 -0
- tokenizer_config.json +16 -0
- training_args.bin +1 -1
README.md
CHANGED
@@ -2,7 +2,6 @@
|
|
2 |
license: apache-2.0
|
3 |
library_name: peft
|
4 |
tags:
|
5 |
-
- axolotl
|
6 |
- generated_from_trainer
|
7 |
base_model: mistralai/Mixtral-8x7B-Instruct-v0.1
|
8 |
model-index:
|
@@ -25,13 +24,13 @@ chat_template: inst
|
|
25 |
dataset_prepared_path: last_run_prepared
|
26 |
datasets:
|
27 |
- conversation: mistral
|
28 |
-
path:
|
29 |
type: sharegpt
|
30 |
- conversation: mistral
|
31 |
-
path:
|
32 |
type: sharegpt
|
33 |
- conversation: mistral
|
34 |
-
path:
|
35 |
type: sharegpt
|
36 |
debug: null
|
37 |
eval_max_new_tokens: 256
|
@@ -52,6 +51,9 @@ logging_steps: 1
|
|
52 |
lora_alpha: 64
|
53 |
lora_dropout: 0.05
|
54 |
lora_model_dir: null
|
|
|
|
|
|
|
55 |
lora_r: 32
|
56 |
lora_target_linear: true
|
57 |
loss_watchdog_patience: 3
|
@@ -61,22 +63,22 @@ micro_batch_size: 2
|
|
61 |
model_config:
|
62 |
output_router_logits: true
|
63 |
model_type: AutoModelForCausalLM
|
64 |
-
num_epochs:
|
65 |
optimizer: paged_adamw_8bit
|
66 |
-
output_dir:
|
67 |
pad_to_sequence_len: true
|
68 |
sample_packing: true
|
69 |
save_steps: 0.1
|
70 |
sequence_len: 4096
|
71 |
-
special_tokens:
|
72 |
-
conversation: '[c]'
|
73 |
-
function: '[f]'
|
74 |
strict: false
|
75 |
tf32: false
|
76 |
tokenizer_type: AutoTokenizer
|
|
|
|
|
|
|
77 |
train_on_inputs: false
|
78 |
trust_remote_code: true
|
79 |
-
val_set_size: 0.
|
80 |
wandb_log_model: end
|
81 |
wandb_name: more-tools
|
82 |
wandb_project: function-call
|
@@ -89,9 +91,7 @@ weight_decay: 0.0
|
|
89 |
|
90 |
# special-token-all-linear
|
91 |
|
92 |
-
This model is a fine-tuned version of [mistralai/Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) on
|
93 |
-
It achieves the following results on the evaluation set:
|
94 |
-
- Loss: 0.0801
|
95 |
|
96 |
## Model description
|
97 |
|
@@ -122,18 +122,7 @@ The following hyperparameters were used during training:
|
|
122 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
123 |
- lr_scheduler_type: cosine
|
124 |
- lr_scheduler_warmup_steps: 10
|
125 |
-
- num_epochs:
|
126 |
-
|
127 |
-
### Training results
|
128 |
-
|
129 |
-
| Training Loss | Epoch | Step | Validation Loss |
|
130 |
-
|:-------------:|:-----:|:----:|:---------------:|
|
131 |
-
| 2.1829 | 0.01 | 1 | 2.1038 |
|
132 |
-
| 0.091 | 0.8 | 151 | 0.0832 |
|
133 |
-
| 0.0741 | 1.58 | 302 | 0.0801 |
|
134 |
-
| 0.0687 | 2.36 | 453 | 0.0801 |
|
135 |
-
| 0.0654 | 3.14 | 604 | 0.0801 |
|
136 |
-
|
137 |
|
138 |
### Framework versions
|
139 |
|
|
|
2 |
license: apache-2.0
|
3 |
library_name: peft
|
4 |
tags:
|
|
|
5 |
- generated_from_trainer
|
6 |
base_model: mistralai/Mixtral-8x7B-Instruct-v0.1
|
7 |
model-index:
|
|
|
24 |
dataset_prepared_path: last_run_prepared
|
25 |
datasets:
|
26 |
- conversation: mistral
|
27 |
+
path: 722c5ac4af9b4ac09cdd250a281284f5/./data/with_function_response/original_clean/function_used_training.jsonl
|
28 |
type: sharegpt
|
29 |
- conversation: mistral
|
30 |
+
path: 722c5ac4af9b4ac09cdd250a281284f5/./data/with_function_response/original_clean/function_not_used_training.jsonl
|
31 |
type: sharegpt
|
32 |
- conversation: mistral
|
33 |
+
path: 722c5ac4af9b4ac09cdd250a281284f5/./data/with_function_response/parallel_call/parallel_data_training.jsonl
|
34 |
type: sharegpt
|
35 |
debug: null
|
36 |
eval_max_new_tokens: 256
|
|
|
51 |
lora_alpha: 64
|
52 |
lora_dropout: 0.05
|
53 |
lora_model_dir: null
|
54 |
+
lora_modules_to_save:
|
55 |
+
- embed_tokens
|
56 |
+
- lm_head
|
57 |
lora_r: 32
|
58 |
lora_target_linear: true
|
59 |
loss_watchdog_patience: 3
|
|
|
63 |
model_config:
|
64 |
output_router_logits: true
|
65 |
model_type: AutoModelForCausalLM
|
66 |
+
num_epochs: 1
|
67 |
optimizer: paged_adamw_8bit
|
68 |
+
output_dir: 722c5ac4af9b4ac09cdd250a281284f5/model
|
69 |
pad_to_sequence_len: true
|
70 |
sample_packing: true
|
71 |
save_steps: 0.1
|
72 |
sequence_len: 4096
|
|
|
|
|
|
|
73 |
strict: false
|
74 |
tf32: false
|
75 |
tokenizer_type: AutoTokenizer
|
76 |
+
tokens:
|
77 |
+
- '[f]'
|
78 |
+
- '[c]'
|
79 |
train_on_inputs: false
|
80 |
trust_remote_code: true
|
81 |
+
val_set_size: 0.05
|
82 |
wandb_log_model: end
|
83 |
wandb_name: more-tools
|
84 |
wandb_project: function-call
|
|
|
91 |
|
92 |
# special-token-all-linear
|
93 |
|
94 |
+
This model is a fine-tuned version of [mistralai/Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) on an unknown dataset.
|
|
|
|
|
95 |
|
96 |
## Model description
|
97 |
|
|
|
122 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
123 |
- lr_scheduler_type: cosine
|
124 |
- lr_scheduler_warmup_steps: 10
|
125 |
+
- num_epochs: 1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
126 |
|
127 |
### Framework versions
|
128 |
|
adapter_config.json
CHANGED
@@ -13,20 +13,23 @@
|
|
13 |
"lora_dropout": 0.05,
|
14 |
"megatron_config": null,
|
15 |
"megatron_core": "megatron.core",
|
16 |
-
"modules_to_save":
|
|
|
|
|
|
|
17 |
"peft_type": "LORA",
|
18 |
"r": 32,
|
19 |
"rank_pattern": {},
|
20 |
"revision": null,
|
21 |
"target_modules": [
|
22 |
-
"
|
23 |
-
"o_proj",
|
24 |
-
"q_proj",
|
25 |
"k_proj",
|
|
|
|
|
26 |
"v_proj",
|
|
|
27 |
"w3",
|
28 |
-
"
|
29 |
-
"gate"
|
30 |
],
|
31 |
"task_type": "CAUSAL_LM",
|
32 |
"use_dora": false,
|
|
|
13 |
"lora_dropout": 0.05,
|
14 |
"megatron_config": null,
|
15 |
"megatron_core": "megatron.core",
|
16 |
+
"modules_to_save": [
|
17 |
+
"embed_tokens",
|
18 |
+
"lm_head"
|
19 |
+
],
|
20 |
"peft_type": "LORA",
|
21 |
"r": 32,
|
22 |
"rank_pattern": {},
|
23 |
"revision": null,
|
24 |
"target_modules": [
|
25 |
+
"w2",
|
|
|
|
|
26 |
"k_proj",
|
27 |
+
"q_proj",
|
28 |
+
"w1",
|
29 |
"v_proj",
|
30 |
+
"gate",
|
31 |
"w3",
|
32 |
+
"o_proj"
|
|
|
33 |
],
|
34 |
"task_type": "CAUSAL_LM",
|
35 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:752f1017a27bd38e1d2b3f9d6687df0af638719a10af88e6dc48d059a5d71d14
|
3 |
+
size 3511040576
|
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[c]": 32001,
|
3 |
+
"[f]": 32000
|
4 |
+
}
|
config.json
CHANGED
@@ -40,5 +40,5 @@
|
|
40 |
"torch_dtype": "bfloat16",
|
41 |
"transformers_version": "4.38.2",
|
42 |
"use_cache": false,
|
43 |
-
"vocab_size":
|
44 |
}
|
|
|
40 |
"torch_dtype": "bfloat16",
|
41 |
"transformers_version": "4.38.2",
|
42 |
"use_cache": false,
|
43 |
+
"vocab_size": 32002
|
44 |
}
|
tokenizer.json
CHANGED
@@ -29,6 +29,24 @@
|
|
29 |
"rstrip": false,
|
30 |
"normalized": false,
|
31 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
}
|
33 |
],
|
34 |
"normalizer": {
|
|
|
29 |
"rstrip": false,
|
30 |
"normalized": false,
|
31 |
"special": true
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"id": 32000,
|
35 |
+
"content": "[f]",
|
36 |
+
"single_word": false,
|
37 |
+
"lstrip": false,
|
38 |
+
"rstrip": false,
|
39 |
+
"normalized": true,
|
40 |
+
"special": false
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"id": 32001,
|
44 |
+
"content": "[c]",
|
45 |
+
"single_word": false,
|
46 |
+
"lstrip": false,
|
47 |
+
"rstrip": false,
|
48 |
+
"normalized": true,
|
49 |
+
"special": false
|
50 |
}
|
51 |
],
|
52 |
"normalizer": {
|
tokenizer_config.json
CHANGED
@@ -25,6 +25,22 @@
|
|
25 |
"rstrip": false,
|
26 |
"single_word": false,
|
27 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
}
|
29 |
},
|
30 |
"additional_special_tokens": [],
|
|
|
25 |
"rstrip": false,
|
26 |
"single_word": false,
|
27 |
"special": true
|
28 |
+
},
|
29 |
+
"32000": {
|
30 |
+
"content": "[f]",
|
31 |
+
"lstrip": false,
|
32 |
+
"normalized": true,
|
33 |
+
"rstrip": false,
|
34 |
+
"single_word": false,
|
35 |
+
"special": false
|
36 |
+
},
|
37 |
+
"32001": {
|
38 |
+
"content": "[c]",
|
39 |
+
"lstrip": false,
|
40 |
+
"normalized": true,
|
41 |
+
"rstrip": false,
|
42 |
+
"single_word": false,
|
43 |
+
"special": false
|
44 |
}
|
45 |
},
|
46 |
"additional_special_tokens": [],
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5688
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3943d56b3a1e58cef6a9201d4b1b47c36798809881f5ae91a6c377c0528808e3
|
3 |
size 5688
|