Bump llama-cpp-python to 0.2.18 (2nd attempt) (#4637)
* Update requirements*.txt * Add back seed
This commit is contained in:
oobabooga
GitHub
parent
9d6f79db74
commit
e0ca49ed9c
15 changed files with 108 additions and 116 deletions
|
|
@ -39,7 +39,7 @@ class LlamacppHF(PreTrainedModel):
|
|||
'n_tokens': self.model.n_tokens,
|
||||
'input_ids': self.model.input_ids,
|
||||
'scores': self.model.scores,
|
||||
'ctx': self.model.ctx
|
||||
'ctx': self.model._ctx
|
||||
}
|
||||
|
||||
if shared.args.cfg_cache:
|
||||
|
|
@ -65,7 +65,7 @@ class LlamacppHF(PreTrainedModel):
|
|||
'n_tokens': self.model.n_tokens,
|
||||
'input_ids': self.model.input_ids,
|
||||
'scores': self.model.scores,
|
||||
'ctx': self.model.ctx
|
||||
'ctx': self.model._ctx
|
||||
})
|
||||
|
||||
def save_negative_cache(self):
|
||||
|
|
@ -73,20 +73,20 @@ class LlamacppHF(PreTrainedModel):
|
|||
'n_tokens': self.model.n_tokens,
|
||||
'input_ids': self.model.input_ids,
|
||||
'scores': self.model.scores,
|
||||
'ctx': self.model.ctx
|
||||
'ctx': self.model._ctx
|
||||
})
|
||||
|
||||
def load_cache(self):
|
||||
self.model.n_tokens = self.llamacpp_cache['n_tokens']
|
||||
self.model.input_ids = self.llamacpp_cache['input_ids']
|
||||
self.model.scores = self.llamacpp_cache['scores']
|
||||
self.model.ctx = self.llamacpp_cache['ctx']
|
||||
self.model._ctx = self.llamacpp_cache['ctx']
|
||||
|
||||
def load_negative_cache(self):
|
||||
self.model.n_tokens = self.llamacpp_cache_negative['n_tokens']
|
||||
self.model.input_ids = self.llamacpp_cache_negative['input_ids']
|
||||
self.model.scores = self.llamacpp_cache_negative['scores']
|
||||
self.model.ctx = self.llamacpp_cache_negative['ctx']
|
||||
self.model._ctx = self.llamacpp_cache_negative['ctx']
|
||||
|
||||
@property
|
||||
def device(self) -> torch.device:
|
||||
|
|
@ -192,7 +192,6 @@ class LlamacppHF(PreTrainedModel):
|
|||
params = {
|
||||
'model_path': str(model_file),
|
||||
'n_ctx': shared.args.n_ctx,
|
||||
'seed': int(shared.args.llama_cpp_seed),
|
||||
'n_threads': shared.args.threads or None,
|
||||
'n_threads_batch': shared.args.threads_batch or None,
|
||||
'n_batch': shared.args.n_batch,
|
||||
|
|
|
|||
|
|
@ -74,7 +74,6 @@ class LlamaCppModel:
|
|||
params = {
|
||||
'model_path': str(path),
|
||||
'n_ctx': shared.args.n_ctx,
|
||||
'seed': int(shared.args.llama_cpp_seed),
|
||||
'n_threads': shared.args.threads or None,
|
||||
'n_threads_batch': shared.args.threads_batch or None,
|
||||
'n_batch': shared.args.n_batch,
|
||||
|
|
@ -144,15 +143,16 @@ class LlamaCppModel:
|
|||
max_tokens=state['max_new_tokens'],
|
||||
temperature=state['temperature'],
|
||||
top_p=state['top_p'],
|
||||
top_k=state['top_k'],
|
||||
repeat_penalty=state['repetition_penalty'],
|
||||
presence_penalty=state['presence_penalty'],
|
||||
frequency_penalty=state['frequency_penalty'],
|
||||
presence_penalty=state['presence_penalty'],
|
||||
repeat_penalty=state['repetition_penalty'],
|
||||
top_k=state['top_k'],
|
||||
stream=True,
|
||||
seed=int(state['seed']) if state['seed'] != -1 else None,
|
||||
tfs_z=state['tfs'],
|
||||
mirostat_mode=int(state['mirostat_mode']),
|
||||
mirostat_tau=state['mirostat_tau'],
|
||||
mirostat_eta=state['mirostat_eta'],
|
||||
stream=True,
|
||||
logits_processor=logit_processors,
|
||||
grammar=self.grammar
|
||||
)
|
||||
|
|
|
|||
|
|
@ -99,7 +99,6 @@ loaders_and_params = OrderedDict({
|
|||
'no_mmap',
|
||||
'mlock',
|
||||
'no_mul_mat_q',
|
||||
'llama_cpp_seed',
|
||||
'alpha_value',
|
||||
'rope_freq_base',
|
||||
'compress_pos_emb',
|
||||
|
|
@ -366,6 +365,7 @@ loaders_samplers = {
|
|||
'repetition_penalty',
|
||||
'presence_penalty',
|
||||
'frequency_penalty',
|
||||
'seed',
|
||||
'mirostat_mode',
|
||||
'mirostat_tau',
|
||||
'mirostat_eta',
|
||||
|
|
|
|||
|
|
@ -112,7 +112,6 @@ parser.add_argument('--no-mmap', action='store_true', help='Prevent mmap from be
|
|||
parser.add_argument('--mlock', action='store_true', help='Force the system to keep the model in RAM.')
|
||||
parser.add_argument('--n-gpu-layers', type=int, default=0, help='Number of layers to offload to the GPU.')
|
||||
parser.add_argument('--tensor_split', type=str, default=None, help='Split the model across multiple GPUs. Comma-separated list of proportions. Example: 18,17.')
|
||||
parser.add_argument('--llama_cpp_seed', type=int, default=0, help='Seed for llama-cpp models. Default is 0 (random).')
|
||||
parser.add_argument('--numa', action='store_true', help='Activate NUMA task allocation for llama.cpp.')
|
||||
parser.add_argument('--logits_all', action='store_true', help='Needs to be set for perplexity evaluation to work. Otherwise, ignore it, as it makes prompt processing slower.')
|
||||
parser.add_argument('--cache-capacity', type=str, help='Maximum cache capacity (llama-cpp-python). Examples: 2000MiB, 2GiB. When provided without units, bytes will be assumed.')
|
||||
|
|
@ -182,6 +181,7 @@ parser.add_argument('--no-stream', action='store_true', help='DEPRECATED')
|
|||
parser.add_argument('--mul_mat_q', action='store_true', help='DEPRECATED')
|
||||
parser.add_argument('--api-blocking-port', type=int, default=5000, help='DEPRECATED')
|
||||
parser.add_argument('--api-streaming-port', type=int, default=5005, help='DEPRECATED')
|
||||
parser.add_argument('--llama_cpp_seed', type=int, default=0, help='DEPRECATED')
|
||||
parser.add_argument('--use_fast', action='store_true', help='DEPRECATED')
|
||||
|
||||
args = parser.parse_args()
|
||||
|
|
|
|||
|
|
@ -80,7 +80,6 @@ def list_model_elements():
|
|||
'n_gpu_layers',
|
||||
'tensor_split',
|
||||
'n_ctx',
|
||||
'llama_cpp_seed',
|
||||
'gpu_split',
|
||||
'max_seq_len',
|
||||
'compress_pos_emb',
|
||||
|
|
|
|||
|
|
@ -120,7 +120,6 @@ def create_ui():
|
|||
shared.gradio['load_in_4bit'] = gr.Checkbox(label="load-in-4bit", value=shared.args.load_in_4bit)
|
||||
shared.gradio['use_double_quant'] = gr.Checkbox(label="use_double_quant", value=shared.args.use_double_quant)
|
||||
shared.gradio['tensor_split'] = gr.Textbox(label='tensor_split', info='Split the model across multiple GPUs, comma-separated list of proportions, e.g. 18,17')
|
||||
shared.gradio['llama_cpp_seed'] = gr.Number(label='Seed (0 for random)', value=shared.args.llama_cpp_seed)
|
||||
shared.gradio['trust_remote_code'] = gr.Checkbox(label="trust-remote-code", value=shared.args.trust_remote_code, info='To enable this option, start the web UI with the --trust-remote-code flag. It is necessary for some models.', interactive=shared.args.trust_remote_code)
|
||||
shared.gradio['cfg_cache'] = gr.Checkbox(label="cfg-cache", value=shared.args.cfg_cache, info='Create an additional cache for CFG negative prompts.')
|
||||
shared.gradio['logits_all'] = gr.Checkbox(label="logits_all", value=shared.args.logits_all, info='Needs to be set for perplexity evaluation to work. Otherwise, ignore it, as it makes prompt processing slower.')
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue