From 8a4df6e181009112ee951ed232843f0346d78836 Mon Sep 17 00:00:00 2001 From: Nicolas Patry Date: Wed, 28 Aug 2024 16:34:58 +0200 Subject: [PATCH] Only n_heads / process_group.size() are necessary. --- server/text_generation_server/models/flash_causal_lm.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/server/text_generation_server/models/flash_causal_lm.py b/server/text_generation_server/models/flash_causal_lm.py index 4ee58038..a84ba765 100644 --- a/server/text_generation_server/models/flash_causal_lm.py +++ b/server/text_generation_server/models/flash_causal_lm.py @@ -1001,7 +1001,7 @@ class FlashCausalLM(Model): config.sliding_window = None self.num_layers = config.num_hidden_layers - self.num_heads = config.num_attention_heads + self.num_heads = config.num_attention_heads // self.process_group.size() # Validation is done in the model itself if num_kv_heads is None: num_kv_heads = getattr(config, "num_key_value_heads", None)