unitorch.cli.models.peft¤
BloomLoraForClassification¤
Tip
classification/peft/lora/bloom
is the section for configuration of BloomLoraForClassification.
Bases: BloomLoraForClassification
Initialize the BloomLoraForClassification model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config_path |
str
|
The path to the model configuration file. |
required |
lora_r |
int
|
The number of Lora ranks. Defaults to 16. |
16
|
lora_alpha |
int
|
The Lora alpha value. Defaults to 32. |
32
|
lora_dropout |
float
|
The Lora dropout rate. Defaults to 0.05. |
0.05
|
fan_in_fan_out |
bool
|
Whether to use fan-in/fan-out weight initialization. Defaults to True. |
True
|
target_modules |
Union[List[str], str]
|
The target modules for Lora regularization. Defaults to ["q_proj", "v_proj"]. |
['query_key_value']
|
num_classes |
int
|
The number of classes. Defaults to 1. |
1
|
gradient_checkpointing |
bool
|
Whether to use gradient checkpointing during training. Defaults to False. |
False
|
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 |
|
forward ¤
forward(
input_ids: Tensor,
attention_mask: Optional[Tensor] = None,
position_ids: Optional[Tensor] = None,
)
Perform forward pass of the BloomLoraForClassification model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
The input IDs. |
required |
attention_mask |
Tensor
|
The attention mask. |
None
|
position_ids |
Tensor
|
The position IDs. |
None
|
Returns:
Name | Type | Description |
---|---|---|
ClassificationOutputs |
The output of the classification task. |
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 |
|
from_core_configure
classmethod
¤
from_core_configure(config, **kwargs)
Create an instance of BloomLoraForClassification from a core configuration.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config |
The core configuration. |
required | |
**kwargs |
Additional keyword arguments. |
{}
|
Returns:
Name | Type | Description |
---|---|---|
BloomLoraForClassification |
The initialized BloomLoraForClassification instance. |
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
|
BloomLoraForGeneration¤
Tip
core/model/generation/peft/lora/bloom
is the section for configuration of BloomLoraForGeneration.
Bases: BloomLoraForGeneration
BloomLora model for generation tasks.
Initialize the BloomLoraForGeneration model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config_path |
str
|
The path to the model configuration file. |
required |
lora_r |
int
|
The number of Lora ranks. Defaults to 16. |
16
|
lora_alpha |
int
|
The Lora alpha value. Defaults to 32. |
32
|
lora_dropout |
float
|
The Lora dropout rate. Defaults to 0.05. |
0.05
|
fan_in_fan_out |
bool
|
Whether to use fan-in/fan-out weight initialization. Defaults to True. |
True
|
target_modules |
Union[List[str], str]
|
The target modules for Lora regularization. Defaults to ["q_proj", "v_proj"]. |
['query_key_value']
|
gradient_checkpointing |
bool
|
Whether to use gradient checkpointing during training. Defaults to False. |
False
|
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 |
|
forward ¤
forward(
input_ids: Tensor,
attention_mask: Optional[Tensor] = None,
position_ids: Optional[Tensor] = None,
)
Perform forward pass of the BloomLoraForGeneration model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
The input IDs. |
required |
attention_mask |
Tensor
|
The attention mask. |
None
|
position_ids |
Tensor
|
The position IDs. |
None
|
Returns:
Name | Type | Description |
---|---|---|
GenerationOutputs |
The output of the generation task. |
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 |
|
from_core_configure
classmethod
¤
from_core_configure(config, **kwargs)
Create an instance of BloomLoraForGeneration from a core configuration.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config |
The core configuration. |
required | |
**kwargs |
Additional keyword arguments. |
{}
|
Returns:
Name | Type | Description |
---|---|---|
BloomLoraForGeneration |
The initialized BloomLoraForGeneration instance. |
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 |
|
generate ¤
generate(
input_ids: Tensor,
num_beams: Optional[int] = 5,
decoder_start_token_id: Optional[int] = 1,
decoder_end_token_id: Optional[
Union[int, List[int]]
] = 2,
num_return_sequences: Optional[int] = 1,
min_gen_seq_length: Optional[int] = 0,
max_gen_seq_length: Optional[int] = 48,
repetition_penalty: Optional[float] = 1.0,
no_repeat_ngram_size: Optional[int] = 0,
early_stopping: Optional[bool] = True,
length_penalty: Optional[float] = 1.0,
num_beam_groups: Optional[int] = 1,
diversity_penalty: Optional[float] = 0.0,
do_sample: Optional[bool] = False,
temperature: Optional[float] = 1.0,
top_k: Optional[int] = 50,
top_p: Optional[float] = 1.0,
)
Generate sequences using the Bloom model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
Input token IDs. |
required |
num_beams |
int
|
Number of beams for beam search. Defaults to 5. |
5
|
decoder_start_token_id |
int
|
Decoder start token ID. Defaults to 0. |
1
|
decoder_end_token_id |
int or List[int]
|
The ID(s) of the decoder end token(s). Defaults to 1. |
2
|
num_return_sequences |
int
|
Number of generated sequences to return. Defaults to 1. |
1
|
min_gen_seq_length |
int
|
Minimum generation sequence length. Defaults to 0. |
0
|
max_gen_seq_length |
int
|
Maximum generation sequence length. Defaults to 48. |
48
|
repetition_penalty |
float
|
Repetition penalty. Defaults to 1.0. |
1.0
|
no_repeat_ngram_size |
int
|
Size of n-grams to prevent repetition. Defaults to 0. |
0
|
early_stopping |
bool
|
Whether to perform early stopping. Defaults to True. |
True
|
length_penalty |
float
|
Length penalty. Defaults to 1.0. |
1.0
|
num_beam_groups |
int
|
Number of beam groups for diverse beam search. Defaults to 1. |
1
|
diversity_penalty |
float
|
Diversity penalty for diverse beam search. Defaults to 0.0. |
0.0
|
do_sample |
bool
|
Whether to use sampling for generation. Defaults to False. |
False
|
temperature |
float
|
Sampling temperature. Defaults to 1.0. |
1.0
|
top_k |
int
|
Top-k sampling parameter. Defaults to 50. |
50
|
top_p |
float
|
Top-p sampling parameter. Defaults to 1.0. |
1.0
|
Returns:
Name | Type | Description |
---|---|---|
GenerationOutputs |
The generation outputs. |
Source code in src/unitorch/cli/models/peft/modeling_bloom.py
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 |
|
LlamaLoraForClassification¤
Tip
core/model/classification/peft/lora/llama
is the section for configuration of LlamaLoraForClassification.
Bases: LlamaLoraForClassification
LlamaLora model for classification tasks.
Initialize the LlamaLoraForClassification model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config_path |
str
|
The path to the model configuration file. |
required |
lora_r |
int
|
The number of Lora ranks. Defaults to 16. |
16
|
lora_alpha |
int
|
The Lora alpha value. Defaults to 32. |
32
|
lora_dropout |
float
|
The Lora dropout rate. Defaults to 0.05. |
0.05
|
fan_in_fan_out |
bool
|
Whether to use fan-in/fan-out weight initialization. Defaults to True. |
True
|
target_modules |
Union[List[str], str]
|
The target modules for Lora regularization. Defaults to ["q_proj", "v_proj"]. |
['q_proj', 'v_proj']
|
num_classes |
int
|
The number of classes. Defaults to 1. |
1
|
gradient_checkpointing |
bool
|
Whether to use gradient checkpointing during training. Defaults to False. |
False
|
Source code in src/unitorch/cli/models/peft/modeling_llama.py
30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
|
forward ¤
forward(
input_ids: Tensor,
attention_mask: Optional[Tensor] = None,
position_ids: Optional[Tensor] = None,
)
Perform forward pass of the LlamaLoraForClassification model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
The input IDs. |
required |
attention_mask |
Tensor
|
The attention mask. |
None
|
position_ids |
Tensor
|
The position IDs. |
None
|
Returns:
Name | Type | Description |
---|---|---|
ClassificationOutputs |
The output of the classification task. |
Source code in src/unitorch/cli/models/peft/modeling_llama.py
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 |
|
from_core_configure
classmethod
¤
from_core_configure(config, **kwargs)
Create an instance of LlamaLoraForClassification from a core configuration.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config |
The core configuration. |
required | |
**kwargs |
Additional keyword arguments. |
{}
|
Returns:
Name | Type | Description |
---|---|---|
LlamaLoraForClassification |
The initialized LlamaLoraForClassification instance. |
Source code in src/unitorch/cli/models/peft/modeling_llama.py
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 |
|
LlamaLoraForGeneration¤
Tip
core/model/generation/peft/lora/llama
is the section for configuration of LlamaLoraForGeneration.
Bases: LlamaLoraForGeneration
LlamaLora model for generation tasks.
Initialize the LlamaLoraForGeneration model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config_path |
str
|
The path to the model configuration file. |
required |
lora_r |
int
|
The number of Lora ranks. Defaults to 16. |
16
|
lora_alpha |
int
|
The Lora alpha value. Defaults to 32. |
32
|
lora_dropout |
float
|
The Lora dropout rate. Defaults to 0.05. |
0.05
|
fan_in_fan_out |
bool
|
Whether to use fan-in/fan-out weight initialization. Defaults to True. |
True
|
target_modules |
Union[List[str], str]
|
The target modules for Lora regularization. Defaults to ["q_proj", "v_proj"]. |
['q_proj', 'v_proj']
|
gradient_checkpointing |
bool
|
Whether to use gradient checkpointing during training. Defaults to False. |
False
|
Source code in src/unitorch/cli/models/peft/modeling_llama.py
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 |
|
forward ¤
forward(
input_ids: Tensor,
attention_mask: Optional[Tensor] = None,
position_ids: Optional[Tensor] = None,
)
Perform forward pass of the LlamaLoraForGeneration model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
The input IDs. |
required |
attention_mask |
Tensor
|
The attention mask. |
None
|
position_ids |
Tensor
|
The position IDs. |
None
|
Returns:
Name | Type | Description |
---|---|---|
GenerationOutputs |
The output of the generation task. |
Source code in src/unitorch/cli/models/peft/modeling_llama.py
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 |
|
from_core_configure
classmethod
¤
from_core_configure(config, **kwargs)
Create an instance of LlamaLoraForGeneration from a core configuration.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
config |
The core configuration. |
required | |
**kwargs |
Additional keyword arguments. |
{}
|
Returns:
Name | Type | Description |
---|---|---|
LlamaLoraForGeneration |
The initialized LlamaLoraForGeneration instance. |
Source code in src/unitorch/cli/models/peft/modeling_llama.py
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 |
|
generate ¤
generate(
input_ids: Tensor,
num_beams: Optional[int] = 5,
decoder_start_token_id: Optional[int] = 1,
decoder_end_token_id: Optional[
Union[int, List[int]]
] = 2,
num_return_sequences: Optional[int] = 1,
min_gen_seq_length: Optional[int] = 0,
max_gen_seq_length: Optional[int] = 48,
repetition_penalty: Optional[float] = 1.0,
no_repeat_ngram_size: Optional[int] = 0,
early_stopping: Optional[bool] = True,
length_penalty: Optional[float] = 1.0,
num_beam_groups: Optional[int] = 1,
diversity_penalty: Optional[float] = 0.0,
do_sample: Optional[bool] = False,
temperature: Optional[float] = 1.0,
top_k: Optional[int] = 50,
top_p: Optional[float] = 1.0,
)
Generate sequences using the Llama model.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_ids |
Tensor
|
Input token IDs. |
required |
num_beams |
int
|
Number of beams for beam search. Defaults to 5. |
5
|
decoder_start_token_id |
int
|
Decoder start token ID. Defaults to 0. |
1
|
decoder_end_token_id |
int or List[int]
|
The ID(s) of the decoder end token(s). Defaults to 1. |
2
|
num_return_sequences |
int
|
Number of generated sequences to return. Defaults to 1. |
1
|
min_gen_seq_length |
int
|
Minimum generation sequence length. Defaults to 0. |
0
|
max_gen_seq_length |
int
|
Maximum generation sequence length. Defaults to 48. |
48
|
repetition_penalty |
float
|
Repetition penalty. Defaults to 1.0. |
1.0
|
no_repeat_ngram_size |
int
|
Size of n-grams to prevent repetition. Defaults to 0. |
0
|
early_stopping |
bool
|
Whether to perform early stopping. Defaults to True. |
True
|
length_penalty |
float
|
Length penalty. Defaults to 1.0. |
1.0
|
num_beam_groups |
int
|
Number of beam groups for diverse beam search. Defaults to 1. |
1
|
diversity_penalty |
float
|
Diversity penalty for diverse beam search. Defaults to 0.0. |
0.0
|
do_sample |
bool
|
Whether to use sampling for generation. Defaults to False. |
False
|
temperature |
float
|
Sampling temperature. Defaults to 1.0. |
1.0
|
top_k |
int
|
Top-k sampling parameter. Defaults to 50. |
50
|
top_p |
float
|
Top-p sampling parameter. Defaults to 1.0. |
1.0
|
Returns:
Name | Type | Description |
---|---|---|
GenerationOutputs |
The generation outputs. |
Source code in src/unitorch/cli/models/peft/modeling_llama.py
299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 |
|