Implement repetition-penalty for generation

Author: tridao

README.md

@@ -134,8 +134,8 @@ Other configurable options include the top-p (nucleus sampling) probability, and
 To test generation latency (e.g. batch size = 1) with different sampling strategies:
 
 ```
-python benchmarks/benchmark_generation_mamba_simple.py --model-name "state-spaces/mamba-2.8b" --prompt "My cat wrote all this CUDA code for a new language model and" --topp 0.9 --temperature 0.5
-python benchmarks/benchmark_generation_mamba_simple.py --model-name "EleutherAI/pythia-2.8b" --prompt "My cat wrote all this CUDA code for a new language model and" --topp 0.9 --temperature 0.5
+python benchmarks/benchmark_generation_mamba_simple.py --model-name "state-spaces/mamba-2.8b" --prompt "My cat wrote all this CUDA code for a new language model and" --topp 0.9 --temperature 0.7 --repetition-penalty 1.2
+python benchmarks/benchmark_generation_mamba_simple.py --model-name "EleutherAI/pythia-2.8b" --prompt "My cat wrote all this CUDA code for a new language model and" --topp 0.9 --temperature 0.7 --repetition-penalty 1.2
 ```
 
 To test generation throughput with random prompts (e.g. large batch size):

benchmarks/benchmark_generation_mamba_simple.py

@@ -22,6 +22,7 @@
 parser.add_argument("--temperature", type=float, default=1.0)
 parser.add_argument("--topk", type=int, default=1)
 parser.add_argument("--topp", type=float, default=1.0)
+parser.add_argument("--repetition-penalty", type=float, default=1.0)
 parser.add_argument("--batch", type=int, default=1)
 args = parser.parse_args()
 
@@ -61,6 +62,7 @@
         temperature=args.temperature,
         top_k=args.topk,
         top_p=args.topp,
+        repetition_penalty=args.repetition_penalty,
     )
 else:
     fn = lambda: model.generate(
@@ -73,6 +75,7 @@
         temperature=args.temperature,
         top_k=args.topk,
         top_p=args.topp,
+        repetition_penalty=args.repetition_penalty,
     )
 out = fn()
 if args.prompt is not None:

mamba_ssm/utils/generation.py

@@ -60,6 +60,20 @@ def modify_logits_for_top_p_filtering(logits, top_p):
     logits.masked_fill_(indices_to_remove, float("-inf"))
 
 
+def modify_logit_for_repetition_penalty(logits, prev_output_tokens, repetition_penalty=1.0):
+    """Apply repetition penalty. See https://arxiv.org/abs/1909.05858
+    logits: (batch_size, vocab_size)
+    prev_output_tokens: (batch_size, seq_len)
+    """
+    if repetition_penalty == 1.0:
+        return logits
+    score = torch.gather(logits, 1, prev_output_tokens)
+    # if score < 0 then repetition penalty has to be multiplied to reduce the previous token probability
+    score = torch.where(score < 0, score * repetition_penalty, score / repetition_penalty)
+    logits.scatter_(1, prev_output_tokens, score)
+    return logits
+
+
 def sample(logits, top_k=1, top_p=0.0, temperature=1.0):
     """Sample from top-k logits.
     Arguments:
@@ -97,6 +111,7 @@ def decode(
     top_k=1,
     top_p=0.0,
     temperature=1.0,
+    repetition_penalty=1.0,
     eos_token_id=None,
     teacher_outputs=None,
     vocab_size=None,
@@ -186,10 +201,18 @@ def should_stop(current_token, inference_params):
     if enable_timing:
         start.record()
     scores, sequences = [], [input_ids]
+    sequences_cat = input_ids
     while not should_stop(sequences[-1], inference_params):
         scores.append(get_logits(sequences[-1], inference_params))
         inference_params.seqlen_offset += sequences[-1].shape[1]
-        sampled_tokens = sample_tokens(scores[-1], inference_params)
+        if repetition_penalty == 1.0:
+            sampled_tokens = sample_tokens(scores[-1], inference_params)
+        else:
+            logits = modify_logit_for_repetition_penalty(
+                scores[-1].clone(), sequences_cat, repetition_penalty
+            )
+            sampled_tokens = sample_tokens(logits, inference_params)
+            sequences_cat = torch.cat([sequences_cat, sampled_tokens], dim=1)
         sequences.append(sampled_tokens)
         if streamer is not None:
             streamer.put(sampled_tokens.cpu())