Views
No views yet
1from transformers import AutoModelForCausalLM, AutoTokenizer
2
3model_name = "theblackcat102/pythia-1.4b-deduped-sft-r2"
4
5tokenizer = AutoTokenizer.from_pretrained(model_name)
6model = AutoModelForCausalLM.from_pretrained(model_name).half().eval().cuda()
7
8input_text = """
9<|startoftoken|>system
10You are a helpful assistant<|endoftoken|><|startoftoken|>human
11What's the population of the earth?<|endoftoken|><|startoftoken|>assistant
12
13"""
14inputs = tokenizer(input_text, return_tensors="pt", padding=True).to(0)
15outputs = model.generate(
16 **inputs,
17 early_stopping=True,
18 max_new_tokens=args.max_new_tokens,
19 do_sample=True,
20 top_k=args.top_k,
21 temperature=args.temperature,
22 pad_token_id=tokenizer.eos_token_id,
23 # dialogue_collator.py line 36
24)
25output = tokenizer.decode(outputs[0], truncate_before_pattern=[r"\n\n^#", "^'''", "\n\n\n"])
26print(output)deepspeed trainer_sft.py --configs defaults pythia-1-4b-ost --deepspeeddefaults:
learning_rate: 1e-5
gradient_checkpointing: false
gradient_accumulation_steps: 32
per_device_train_batch_size: 2
per_device_eval_batch_size: 2
weight_decay: 0.00
warmup_steps: 600
eval_steps: 250
save_steps: 250
max_length: 512
num_train_epochs: 2
logging_steps: 10
max_grad_norm: 2.0
save_total_limit: 4
fp16: true
eval_accumulation_steps:
freeze_layer:
datasets:
- oa_private:
data_path: .cache
split: sft
val_split: 0.01
fraction: 1
file: 2023-02-26_oasst_default.jsonl
cache_dir: .cache
loss_fn: CrossEntropyLoss
eval_size:
log_dir: "base"
quantization: false
seq2seqmodel: false
poly_eps: 1.0
fuse_gelu: false
log_wandb: true
samples_mixing: true # uses collator that mixes samples in the batch to create a single sample with possible multiple tasks within
verbose: false
pythia-1-4b-ost:
learning_rate: 1e-6
model_name: EleutherAI/pythia-1.4b-deduped
weight_decay: 0.01
max_length: 1024
warmup_steps: 100
gradient_checkpointing: false
gradient_accumulation_steps: 12
per_device_train_batch_size: 5
per_device_eval_batch_size: 6
eval_steps: 100
save_steps: 100
num_train_epochs: 50
save_total_limit: 4