From c0fd4df88314b79a86df101f5e40ef938efdcf3d Mon Sep 17 00:00:00 2001 From: Eddie-Wang Date: Mon, 10 Jun 2024 03:07:38 +0000 Subject: [PATCH] fix merge --- convert-hf-to-gguf.py | 11 ----------- llama.cpp | 4 ++-- 2 files changed, 2 insertions(+), 13 deletions(-) diff --git a/convert-hf-to-gguf.py b/convert-hf-to-gguf.py index 937d0e328..d8ae13c06 100755 --- a/convert-hf-to-gguf.py +++ b/convert-hf-to-gguf.py @@ -1406,19 +1406,8 @@ class BitnetModel(Model): def set_gguf_parameters(self): super().set_gguf_parameters() - self.gguf_writer.add_name("Bitnet") - self.gguf_writer.add_context_length(self.hparams["max_position_embeddings"]) - self.gguf_writer.add_embedding_length(self.hparams["hidden_size"]) - self.gguf_writer.add_block_count(self.hparams["num_hidden_layers"]) - self.gguf_writer.add_feed_forward_length(self.hparams["intermediate_size"]) - self.gguf_writer.add_rope_dimension_count(self.hparams["hidden_size"] // self.hparams["num_attention_heads"]) - self.gguf_writer.add_head_count(self.hparams["num_attention_heads"]) - self.gguf_writer.add_head_count_kv(self.hparams["num_key_value_heads"]) - self.gguf_writer.add_layer_norm_rms_eps(self.hparams["rms_norm_eps"]) - self.gguf_writer.add_vocab_size(self.hparams["vocab_size"]) self.gguf_writer.add_rope_scaling_type(gguf.RopeScalingType.LINEAR) self.gguf_writer.add_rope_scaling_factor(1.0) - self.gguf_writer.add_rope_freq_base(self.hparams["rope_theta"]) def weight_quant(self, weight): dtype = weight.dtype diff --git a/llama.cpp b/llama.cpp index 61d6ae5a3..5ebdeb024 100644 --- a/llama.cpp +++ b/llama.cpp @@ -11569,14 +11569,14 @@ struct llm_build_context { Qcur = ggml_rope_ext( ctx0, ggml_reshape_3d(ctx0, Qcur, n_embd_head, n_head, n_tokens), inp_pos, nullptr, - n_rot, rope_type, 0, n_orig_ctx, freq_base, freq_scale, + n_rot, rope_type, n_ctx_orig, freq_base, freq_scale, ext_factor, attn_factor, beta_fast, beta_slow ); cb(Qcur, "Qcur", il); Kcur = ggml_rope_ext( ctx0, ggml_reshape_3d(ctx0, Kcur, n_embd_head, n_head_kv, n_tokens), inp_pos, nullptr, - n_rot, rope_type, 0, n_orig_ctx, freq_base, freq_scale, + n_rot, rope_type, n_ctx_orig, freq_base, freq_scale, ext_factor, attn_factor, beta_fast, beta_slow ); cb(Kcur, "Kcur", il);