bigdl-core-cpp 2.6.0b20241204__py3-none-win_amd64.whl → 2.6.0b20241211__py3-none-win_amd64.whl

bigdl/cpp/gguf-py/gguf/quants.py

@@ -574,6 +574,87 @@ class Q6_K(__Quant, qtype=GGMLQuantizationType.Q6_K):
         return (d * q).reshape((n_blocks, QK_K))
 
 
+class TQ1_0(__Quant, qtype=GGMLQuantizationType.TQ1_0):
+    @classmethod
+    def quantize_blocks(cls, blocks: np.ndarray) -> np.ndarray:
+        n_blocks = blocks.shape[0]
+
+        d = abs(blocks).max(axis=-1, keepdims=True)
+        with np.errstate(divide="ignore"):
+            id = np.where(d == 0, 0, 1 / d)
+        qs = np_roundf(blocks * id)
+        qs = (qs.astype(np.int8) + np.int8(1)).astype(np.uint8)
+
+        qs0, qs1, qh = qs[..., :(32 * 5)], qs[..., (32 * 5):(48 * 5)], qs[..., (48 * 5):]
+        qs0 = qs0.reshape((n_blocks, -1, 5, 32)) * np.array([81, 27, 9, 3, 1], dtype=np.uint8).reshape((1, 1, 5, 1))
+        qs0 = np.sum(qs0, axis=-2).reshape((n_blocks, -1))
+        qs1 = qs1.reshape((n_blocks, -1, 5, 16)) * np.array([81, 27, 9, 3, 1], dtype=np.uint8).reshape((1, 1, 5, 1))
+        qs1 = np.sum(qs1, axis=-2).reshape((n_blocks, -1))
+        qh = qh.reshape((n_blocks, -1, 4, 4)) * np.array([81, 27, 9, 3], dtype=np.uint8).reshape((1, 1, 4, 1))
+        qh = np.sum(qh, axis=-2).reshape((n_blocks, -1))
+        qs = np.concatenate([qs0, qs1, qh], axis=-1)
+        qs = (qs.astype(np.uint16) * 256 + (243 - 1)) // 243
+
+        qs = qs.astype(np.uint8)
+        d = d.astype(np.float16).view(np.uint8)
+
+        return np.concatenate([qs, d], axis=-1)
+
+    @classmethod
+    def dequantize_blocks(cls, blocks: np.ndarray) -> np.ndarray:
+        n_blocks = blocks.shape[0]
+
+        qs, rest = np.hsplit(blocks, [(QK_K - 4 * QK_K // 64) // 5])
+        qh, d = np.hsplit(rest, [QK_K // 64])
+
+        d = d.view(np.float16).astype(np.float32)
+
+        qs0, qs1 = qs[..., :32], qs[..., 32:]
+        qs0 = qs0.reshape((n_blocks, -1, 1, 32)) * np.array([1, 3, 9, 27, 81], dtype=np.uint8).reshape((1, 1, 5, 1))
+        qs0 = qs0.reshape((n_blocks, -1))
+        qs1 = qs1.reshape((n_blocks, -1, 1, 16)) * np.array([1, 3, 9, 27, 81], dtype=np.uint8).reshape((1, 1, 5, 1))
+        qs1 = qs1.reshape((n_blocks, -1))
+        qh = qh.reshape((n_blocks, -1, 1, 4)) * np.array([1, 3, 9, 27], dtype=np.uint8).reshape((1, 1, 4, 1))
+        qh = qh.reshape((n_blocks, -1))
+        qs = np.concatenate([qs0, qs1, qh], axis=-1)
+        qs = ((qs.astype(np.uint16) * 3) >> 8).astype(np.int8) - np.int8(1)
+
+        return (d * qs.astype(np.float32))
+
+
+class TQ2_0(__Quant, qtype=GGMLQuantizationType.TQ2_0):
+    @classmethod
+    def quantize_blocks(cls, blocks: np.ndarray) -> np.ndarray:
+        n_blocks = blocks.shape[0]
+
+        d = abs(blocks).max(axis=-1, keepdims=True)
+        with np.errstate(divide="ignore"):
+            id = np.where(d == 0, 0, 1 / d)
+        qs = np_roundf(blocks * id)
+        qs = (qs.astype(np.int8) + np.int8(1)).astype(np.uint8)
+
+        qs = qs.reshape((n_blocks, -1, 4, 32)) << np.array([0, 2, 4, 6], dtype=np.uint8).reshape((1, 1, 4, 1))
+        qs = qs[..., 0, :] | qs[..., 1, :] | qs[..., 2, :] | qs[..., 3, :]
+        qs = qs.reshape((n_blocks, -1))
+
+        d = d.astype(np.float16).view(np.uint8)
+
+        return np.concatenate([qs, d], axis=-1)
+
+    @classmethod
+    def dequantize_blocks(cls, blocks: np.ndarray) -> np.ndarray:
+        n_blocks = blocks.shape[0]
+
+        qs, d = np.hsplit(blocks, [QK_K // 4])
+
+        d = d.view(np.float16).astype(np.float32)
+
+        qs = qs.reshape((n_blocks, -1, 1, 32)) >> np.array([0, 2, 4, 6], dtype=np.uint8).reshape((1, 1, 4, 1))
+        qs = (qs & 0x03).reshape((n_blocks, -1)).astype(np.int8) - np.int8(1)
+
+        return (d * qs.astype(np.float32))
+
+
 class IQ2_XXS(__Quant, qtype=GGMLQuantizationType.IQ2_XXS):
     ksigns: bytes = (
         b"\x00\x81\x82\x03\x84\x05\x06\x87\x88\x09\x0a\x8b\x0c\x8d\x8e\x0f"

bigdl/cpp/gguf-py/gguf/tensor_mapping.py

@@ -13,7 +13,7 @@ class TensorNameMap:
             "transformer.wte",                           # gpt2 gpt-j mpt refact qwen dbrx jais exaone
             "transformer.word_embeddings",               # falcon
             "word_embeddings",                           # bloom
-            "model.embed_tokens",                        # llama-hf nemotron
+            "model.embed_tokens",                        # llama-hf nemotron olmoe
             "tok_embeddings",                            # llama-pth
             "embeddings.word_embeddings",                # bert nomic-bert
             "language_model.embedding.word_embeddings",  # persimmon
@@ -27,6 +27,7 @@ class TensorNameMap:
             "embedding.word_embeddings",                 # chatglm
             "transformer.token_embeddings",              # openelm
             "shared",                                    # t5
+            "rwkv.embeddings",                           # rwkv
         ),
 
         # Token type embeddings
@@ -40,6 +41,7 @@ class TensorNameMap:
             "embeddings.LayerNorm",       # bert
             "emb_ln",                     # nomic-bert
             "transformer.norm",           # openelm
+            "rwkv.blocks.0.pre_ln",       # rwkv
         ),
 
         # Position embeddings
@@ -52,18 +54,19 @@ class TensorNameMap:
         # Output
         MODEL_TENSOR.OUTPUT: (
             "embed_out",                 # gptneox
-            "lm_head",                   # gpt2 mpt falcon llama-hf baichuan qwen mamba dbrx jais nemotron exaone
+            "lm_head",                   # gpt2 mpt falcon llama-hf baichuan qwen mamba dbrx jais nemotron exaone olmoe
             "output",                    # llama-pth bloom internlm2
             "word_embeddings_for_head",  # persimmon
             "lm_head.linear",            # phi2
             "output_layer",              # chatglm
+            "head",                      # rwkv
         ),
 
         # Output norm
         MODEL_TENSOR.OUTPUT_NORM: (
             "gpt_neox.final_layer_norm",               # gptneox
             "transformer.ln_f",                        # gpt2 gpt-j falcon jais exaone
-            "model.norm",                              # llama-hf baichuan internlm2
+            "model.norm",                              # llama-hf baichuan internlm2 olmoe
             "norm",                                    # llama-pth
             "transformer.norm_f",                      # mpt dbrx
             "ln_f",                                    # refact bloom qwen gpt2
@@ -76,6 +79,7 @@ class TensorNameMap:
             "encoder.final_layernorm",                 # chatglm
             "transformer.norm",                        # openelm
             "model.norm",                              # nemotron
+            "rwkv.ln_out",                             # rwkv
         ),
 
         # Rope frequencies
@@ -83,6 +87,9 @@ class TensorNameMap:
             "rope.freqs",  # llama-pth
             "rotary_pos_emb.inv_freq",  # chatglm
         ),
+
+        MODEL_TENSOR.ROPE_FACTORS_LONG: (),
+        MODEL_TENSOR.ROPE_FACTORS_SHORT: (),
     }
 
     block_mappings_cfg: dict[MODEL_TENSOR, tuple[str, ...]] = {
@@ -94,7 +101,7 @@ class TensorNameMap:
             "transformer.h.{bid}.input_layernorm",                  # falcon7b
             "h.{bid}.input_layernorm",                              # bloom
             "transformer.h.{bid}.ln_mlp",                           # falcon40b
-            "model.layers.{bid}.input_layernorm",                   # llama-hf nemotron
+            "model.layers.{bid}.input_layernorm",                   # llama-hf nemotron olmoe
             "layers.{bid}.attention_norm",                          # llama-pth
             "language_model.encoder.layers.{bid}.input_layernorm",  # persimmon
             "model.layers.{bid}.ln1",                               # yi
@@ -108,12 +115,14 @@ class TensorNameMap:
             "transformer.blocks.{bid}.norm_attn_norm.norm_1",       # dbrx
             "encoder.layers.{bid}.input_layernorm",                 # chatglm
             "transformer.layers.{bid}.attn_norm",                   # openelm
+            "rwkv.blocks.{bid}.ln1",                                # rwkv
         ),
 
         # Attention norm 2
         MODEL_TENSOR.ATTN_NORM_2: (
-            "transformer.h.{bid}.ln_attn",  # falcon40b
+            "transformer.h.{bid}.ln_attn",                  # falcon40b
             "encoder.layer.{bid}.layer_norm_1",             # jina-v2-code
+            "rwkv.blocks.{bid}.ln2",                        # rwkv
         ),
 
         # Attention query-key-value
@@ -136,7 +145,7 @@ class TensorNameMap:
 
         # Attention query
         MODEL_TENSOR.ATTN_Q: (
-            "model.layers.{bid}.self_attn.q_proj",                       # llama-hf nemotron
+            "model.layers.{bid}.self_attn.q_proj",                       # llama-hf nemotron olmoe
             "layers.{bid}.attention.wq",                                 # llama-pth
             "encoder.layer.{bid}.attention.self.query",                  # bert
             "transformer.h.{bid}.attn.q_proj",                           # gpt-j
@@ -148,7 +157,7 @@ class TensorNameMap:
 
         # Attention key
         MODEL_TENSOR.ATTN_K: (
-            "model.layers.{bid}.self_attn.k_proj",                     # llama-hf nemotron
+            "model.layers.{bid}.self_attn.k_proj",                     # llama-hf nemotron olmoe
             "layers.{bid}.attention.wk",                               # llama-pth
             "encoder.layer.{bid}.attention.self.key",                  # bert
             "transformer.h.{bid}.attn.k_proj",                         # gpt-j
@@ -161,7 +170,7 @@ class TensorNameMap:
 
         # Attention value
         MODEL_TENSOR.ATTN_V: (
-            "model.layers.{bid}.self_attn.v_proj",                       # llama-hf nemotron
+            "model.layers.{bid}.self_attn.v_proj",                       # llama-hf nemotron olmoe
             "layers.{bid}.attention.wv",                                 # llama-pth
             "encoder.layer.{bid}.attention.self.value",                  # bert
             "transformer.h.{bid}.attn.v_proj",                           # gpt-j
@@ -179,7 +188,7 @@ class TensorNameMap:
             "transformer.blocks.{bid}.attn.out_proj",                       # mpt
             "transformer.h.{bid}.self_attention.dense",                     # falcon
             "h.{bid}.self_attention.dense",                                 # bloom
-            "model.layers.{bid}.self_attn.o_proj",                          # llama-hf nemotron
+            "model.layers.{bid}.self_attn.o_proj",                          # llama-hf nemotron olmoe
             "layers.{bid}.attention.wo",                                    # llama-pth
             "encoder.layer.{bid}.attention.output.dense",                   # bert
             "transformer.h.{bid}.attn.out_proj",                            # gpt-j
@@ -223,7 +232,7 @@ class TensorNameMap:
             "transformer.h.{bid}.ln_2",                                      # gpt2 refact qwen jais exaone
             "h.{bid}.post_attention_layernorm",                              # bloom
             "transformer.blocks.{bid}.norm_2",                               # mpt
-            "model.layers.{bid}.post_attention_layernorm",                   # llama-hf nemotron
+            "model.layers.{bid}.post_attention_layernorm",                   # llama-hf nemotron olmoe
             "layers.{bid}.ffn_norm",                                         # llama-pth
             "language_model.encoder.layers.{bid}.post_attention_layernorm",  # persimmon
             "model.layers.{bid}.ln2",                                        # yi
@@ -245,11 +254,12 @@ class TensorNameMap:
         ),
 
         MODEL_TENSOR.FFN_GATE_INP: (
-            "layers.{bid}.feed_forward.gate",             # mixtral
-            "model.layers.{bid}.block_sparse_moe.gate",   # mixtral
-            "model.layers.{bid}.mlp.gate",                # qwen2moe
-            "transformer.decoder_layer.{bid}.router",     # Grok
-            "transformer.blocks.{bid}.ffn.router.layer",  # dbrx
+            "layers.{bid}.feed_forward.gate",                   # mixtral
+            "model.layers.{bid}.block_sparse_moe.gate",         # mixtral
+            "model.layers.{bid}.mlp.gate",                      # qwen2moe olmoe
+            "transformer.decoder_layer.{bid}.router",           # Grok
+            "transformer.blocks.{bid}.ffn.router.layer",        # dbrx
+            "model.layers.{bid}.block_sparse_moe.router.layer", # granitemoe
         ),
 
         MODEL_TENSOR.FFN_GATE_INP_SHEXP: (
@@ -289,7 +299,7 @@ class TensorNameMap:
             "layers.{bid}.feed_forward.experts.w3",          # mixtral (merged)
             "transformer.decoder_layer.{bid}.moe.linear_v",  # Grok (merged)
             "transformer.blocks.{bid}.ffn.experts.mlp.v1",   # dbrx
-            "model.layers.{bid}.mlp.experts.up_proj",        # qwen2moe (merged)
+            "model.layers.{bid}.mlp.experts.up_proj",        # qwen2moe olmoe (merged)
         ),
 
         MODEL_TENSOR.FFN_UP_SHEXP: (
@@ -321,7 +331,7 @@ class TensorNameMap:
             "layers.{bid}.feed_forward.experts.w1",         # mixtral (merged)
             "transformer.decoder_layer.{bid}.moe.linear",   # Grok (merged)
             "transformer.blocks.{bid}.ffn.experts.mlp.w1",  # dbrx
-            "model.layers.{bid}.mlp.experts.gate_proj",     # qwen2moe (merged)
+            "model.layers.{bid}.mlp.experts.gate_proj",     # qwen2moe olmoe (merged)
         ),
 
         MODEL_TENSOR.FFN_GATE_SHEXP: (
@@ -358,10 +368,11 @@ class TensorNameMap:
         ),
 
         MODEL_TENSOR.FFN_DOWN_EXP: (
-            "layers.{bid}.feed_forward.experts.w2",          # mixtral (merged)
-            "transformer.decoder_layer.{bid}.moe.linear_1",  # Grok (merged)
-            "transformer.blocks.{bid}.ffn.experts.mlp.w2",   # dbrx
-            "model.layers.{bid}.mlp.experts.down_proj",      # qwen2moe (merged)
+            "layers.{bid}.feed_forward.experts.w2",              # mixtral (merged)
+            "transformer.decoder_layer.{bid}.moe.linear_1",      # Grok (merged)
+            "transformer.blocks.{bid}.ffn.experts.mlp.w2",       # dbrx
+            "model.layers.{bid}.mlp.experts.down_proj",          # qwen2moe olmoe (merged)
+            "model.layers.{bid}.block_sparse_moe.output_linear", # granitemoe
         ),
 
         MODEL_TENSOR.FFN_DOWN_SHEXP: (
@@ -372,7 +383,7 @@ class TensorNameMap:
         MODEL_TENSOR.ATTN_Q_NORM: (
             "language_model.encoder.layers.{bid}.self_attention.q_layernorm",
             "model.layers.{bid}.self_attn.q_layernorm",                       # persimmon
-            "model.layers.{bid}.self_attn.q_norm",                            # cohere
+            "model.layers.{bid}.self_attn.q_norm",                            # cohere olmoe chameleon
             "transformer.blocks.{bid}.attn.q_ln",                             # sea-lion
             "encoder.layer.{bid}.attention.self.layer_norm_q",                # jina-bert-v2
             "transformer.layers.{bid}.attn.q_norm",                           # openelm
@@ -381,7 +392,7 @@ class TensorNameMap:
         MODEL_TENSOR.ATTN_K_NORM: (
             "language_model.encoder.layers.{bid}.self_attention.k_layernorm",
             "model.layers.{bid}.self_attn.k_layernorm",                       # persimmon
-            "model.layers.{bid}.self_attn.k_norm",                            # cohere
+            "model.layers.{bid}.self_attn.k_norm",                            # cohere olmoe chameleon
             "transformer.blocks.{bid}.attn.k_ln",                             # sea-lion
             "encoder.layer.{bid}.attention.self.layer_norm_k",                # jina-bert-v2
             "transformer.layers.{bid}.attn.k_norm",                           # openelm
@@ -434,6 +445,98 @@ class TensorNameMap:
             "backbone.layers.{bid}.mixer.out_proj",
         ),
 
+        MODEL_TENSOR.TIME_MIX_W1: (
+            "rwkv.blocks.{bid}.attention.time_maa_w1",  # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_W2: (
+            "rwkv.blocks.{bid}.attention.time_maa_w2",  # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_X: (
+            "rwkv.blocks.{bid}.attention.time_maa_x",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_K: (
+            "rwkv.blocks.{bid}.attention.time_maa_k",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_V: (
+            "rwkv.blocks.{bid}.attention.time_maa_v",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_R: (
+            "rwkv.blocks.{bid}.attention.time_maa_r",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_G: (
+            "rwkv.blocks.{bid}.attention.time_maa_g",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LERP_W: (
+            "rwkv.blocks.{bid}.attention.time_maa_w",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_FIRST: (
+            "rwkv.blocks.{bid}.attention.time_faaaa",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_DECAY: (
+            "rwkv.blocks.{bid}.attention.time_decay",   # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_DECAY_W1: (
+            "rwkv.blocks.{bid}.attention.time_decay_w1",  # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_DECAY_W2: (
+            "rwkv.blocks.{bid}.attention.time_decay_w2",  # rwkv v6
+        ),
+
+        MODEL_TENSOR.TIME_MIX_KEY: (
+            "rwkv.blocks.{bid}.attention.key", # rwkv
+        ),
+
+        MODEL_TENSOR.TIME_MIX_VALUE: (
+            "rwkv.blocks.{bid}.attention.value", # rwkv
+        ),
+
+        MODEL_TENSOR.TIME_MIX_RECEPTANCE: (
+            "rwkv.blocks.{bid}.attention.receptance", # rwkv
+        ),
+
+        MODEL_TENSOR.TIME_MIX_GATE: (
+            "rwkv.blocks.{bid}.attention.gate", # rwkv
+        ),
+
+        MODEL_TENSOR.TIME_MIX_LN: (
+            "rwkv.blocks.{bid}.attention.ln_x", # rwkv
+        ),
+
+        MODEL_TENSOR.TIME_MIX_OUTPUT: (
+            "rwkv.blocks.{bid}.attention.output", # rwkv
+        ),
+
+        MODEL_TENSOR.CHANNEL_MIX_LERP_K: (
+            "rwkv.blocks.{bid}.feed_forward.time_maa_k", # rwkv v6
+        ),
+
+        MODEL_TENSOR.CHANNEL_MIX_LERP_R: (
+            "rwkv.blocks.{bid}.feed_forward.time_maa_r", # rwkv v6
+        ),
+
+        MODEL_TENSOR.CHANNEL_MIX_KEY: (
+            "rwkv.blocks.{bid}.feed_forward.key", # rwkv
+        ),
+
+        MODEL_TENSOR.CHANNEL_MIX_RECEPTANCE: (
+            "rwkv.blocks.{bid}.feed_forward.receptance", # rwkv
+        ),
+
+        MODEL_TENSOR.CHANNEL_MIX_VALUE: (
+            "rwkv.blocks.{bid}.feed_forward.value", # rwkv
+        ),
+
         MODEL_TENSOR.ATTN_Q_A: (
             "model.layers.{bid}.self_attn.q_a_proj", # deepseek2
         ),
@@ -579,6 +682,15 @@ class TensorNameMap:
         MODEL_TENSOR.ENC_OUTPUT_NORM: (
             "encoder.final_layer_norm", # t5
         ),
+
+        MODEL_TENSOR.CLS: (
+            "classifier",       # jina
+            "classifier.dense", # roberta
+        ),
+
+        MODEL_TENSOR.CLS_OUT: (
+            "classifier.out_proj", # roberta
+        ),
     }
 
     # architecture-specific block mappings

{bigdl_core_cpp-2.6.0b20241204.dist-info → bigdl_core_cpp-2.6.0b20241211.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: bigdl-core-cpp
-Version: 2.6.0b20241204
+Version: 2.6.0b20241211
 Summary: Large Language Model Develop Toolkit
 Author: BigDL Authors
 License: Apache License, Version 2.0

bigdl_core_cpp-2.6.0b20241211.dist-info/RECORD

@@ -0,0 +1,45 @@
+bigdl/cpp/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+bigdl/cpp/convert_hf_to_gguf.py,sha256=QcBpqyIsrqLlLFwCp0Z8M3IzMobHygjQY0ZgvFoF_u0,207430
+bigdl/cpp/convert_hf_to_gguf_update.py,sha256=O1NH13YPWT9Af778goJOg8pccbrc5cOgwYcPOIOqYq0,16612
+bigdl/cpp/convert_llama_ggml_to_gguf.py,sha256=0dKjRhmFzvWV4e-cuLmaeW14JrWUtZwerBmz8mYyMvI,19556
+bigdl/cpp/convert_lora_to_gguf.py,sha256=qBJSMA_w3cIN_Mi5pNsi4zI1P5GYIeRi4nZPTpAs8QQ,15461
+bigdl/cpp/gguf-py/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+bigdl/cpp/gguf-py/gguf/__init__.py,sha256=h5GWs6SMXYR8giWZ7MTZzAc3hYsIJF-HAkdxtgXLOPo,228
+bigdl/cpp/gguf-py/gguf/constants.py,sha256=8_u4WadRGm7XeN0hxaIzDUgGajyUdHB4XsbslumYS2U,58733
+bigdl/cpp/gguf-py/gguf/gguf.py,sha256=V5jY968TEJn6GJHVdjzH0_aIkZ1QC967vPdHDKDoxZw,491
+bigdl/cpp/gguf-py/gguf/gguf_reader.py,sha256=N3LnQQ30t-S0U85-EvZZzIBfHzo0XuyFVUltdg7Sj3c,12680
+bigdl/cpp/gguf-py/gguf/gguf_writer.py,sha256=tHtbF0ogWwfclKCvO7VyGKoJuCieFrrBx-vVpnruoQA,37092
+bigdl/cpp/gguf-py/gguf/lazy.py,sha256=YIYxGBWD-oKXU4HOvpHs9eiEn81HUgeSmt1mmHJlbdM,8814
+bigdl/cpp/gguf-py/gguf/metadata.py,sha256=wtquhynkyH8R7m3zxgeSUe2bnaekJi6HoCMiYJfJBmk,26232
+bigdl/cpp/gguf-py/gguf/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+bigdl/cpp/gguf-py/gguf/quants.py,sha256=2z6vcK-kBefqZbYNmSEVmdZF_tXHeVb5NC6jCbBdgKc,62040
+bigdl/cpp/gguf-py/gguf/tensor_mapping.py,sha256=oW7E5hnCWy7IeiQeN0v7xoLWeSatDUgEmKq_ZYlLV8s,35299
+bigdl/cpp/gguf-py/gguf/utility.py,sha256=LAwLstUlosYQ1oX9LlJZ-1uCmwyNtOFcJfXpkLnju0k,3003
+bigdl/cpp/gguf-py/gguf/vocab.py,sha256=FtNcm8M5aX9RIr6rRR6UXsUlKMagRUC2xnIWb-xu6rI,19511
+bigdl/cpp/libs/common.lib,sha256=KiEclGjWr4Re-koAlu_EJ0dx_eJWOZ_-cxynqjEhiMs,4205058
+bigdl/cpp/libs/ggml.dll,sha256=ZYnnmLc1Ncyn3rDEJrW6LwWYLHU_Obz80O3TGukLZIs,6219776
+bigdl/cpp/libs/llama-batched.exe,sha256=YiRTTJYU5HzIBRSABWfUsouCYZbUO8A33BJ-sZMN7NM,838656
+bigdl/cpp/libs/llama-bench.exe,sha256=sCqTPsHULyNLaelX2zsJf5eytVUFTJ0CGIQqhCjbKxA,290304
+bigdl/cpp/libs/llama-cli.exe,sha256=CfOwSxxDB9Mnjmv3f--NcQ8o8mnJNn35oIpAiMboAwI,929280
+bigdl/cpp/libs/llama-embedding.exe,sha256=25gJuxi33L9YGz6rqJaB-RwNrPoKJV7TmqkpiFD5D3w,860672
+bigdl/cpp/libs/llama-gguf.exe,sha256=i9_oh1DcUMvQeYo6vJVxha_clodUD2cmGiLiEajco5w,66560
+bigdl/cpp/libs/llama-llava-cli.exe,sha256=Qz63VFbVJos_YcBOJ9lZmz4ls4_WOkb-Ja4Zbxl5qwg,1091072
+bigdl/cpp/libs/llama-lookup.exe,sha256=Xu6BHMpzrXd0KCqbIsxd2v1wYG0tVYl3BQMzCedus74,893952
+bigdl/cpp/libs/llama-ls-sycl-device.exe,sha256=n3r_ttydRhyAfwo9pXZC3Cx81ffJZJg5Iu9R9ObkM-w,10240
+bigdl/cpp/libs/llama-minicpmv-cli.exe,sha256=nzshnx0yvRmS8vf4JQcdllN5gnE55yXEKcDvRlVZoc4,1089024
+bigdl/cpp/libs/llama-perplexity.exe,sha256=2PImsgokgDW2koGU61FADV5BnUhlBAbzZ153bde7aZ4,997376
+bigdl/cpp/libs/llama-quantize.exe,sha256=Zx5Nofwfm9ITX-b1X2ka9VB8oEeTDcq9FLIUKAB9A6I,139264
+bigdl/cpp/libs/llama-server.exe,sha256=_DEZ1za1cqW7lg9FqeTf7UMUc-QOVmuvAKaYxyMkDJY,1960960
+bigdl/cpp/libs/llama-simple.exe,sha256=9CI4Y7tfoXn0zOkUGtNJXcouLpxqqyYdyYKkypxMPUU,834048
+bigdl/cpp/libs/llama-speculative.exe,sha256=yAFNeNEpxzSEbrEqRZlG9b2h975tGyYWlwOO3GvAmew,896512
+bigdl/cpp/libs/llama-tokenize.exe,sha256=445caGydC-Tof4OX9xfyYUItyKrCE6dTZJ8mWpb8CFc,102400
+bigdl/cpp/libs/llama.dll,sha256=na9ZxrYLqG3CmutTL7SCFF__kt2MbbCOQOG-YkWwo6M,2471936
+bigdl/cpp/libs/llava_shared.dll,sha256=Rxxf-Mq63MDfas2omN-A8Y4Fjk06GmmcaGc1uD69a3g,369152
+bigdl/cpp/libs/ollama.exe,sha256=8IcEJkzPcoE0r3K2sAacTX6aVfvKI-xu-1ysEOQ8yJA,64879199
+bigdl_core_cpp-2.6.0b20241211.data/scripts/init-llama-cpp.bat,sha256=U0h6RifZxL3GGJp-0dxdZapQIvXUATSj644CURJL-lg,751
+bigdl_core_cpp-2.6.0b20241211.data/scripts/init-llama-cpp.ps1,sha256=JFOylLxO4MKpllHhdbPuJ1xHi9azxDpzdJns8JtZpkU,501
+bigdl_core_cpp-2.6.0b20241211.data/scripts/init-ollama.bat,sha256=2rScaP2-_yWuAnc86NyX-i6pg7BxjvLc96Kt7HuvQVM,603
+bigdl_core_cpp-2.6.0b20241211.dist-info/METADATA,sha256=eNkNqvxgOYJelp2IlQ17j-t19SmJf1pcKGmQfhyXaBw,652
+bigdl_core_cpp-2.6.0b20241211.dist-info/WHEEL,sha256=z8gukVdnGwjcwo0VnsfJMrhPu5QJT68VcMWmAgvAufw,97
+bigdl_core_cpp-2.6.0b20241211.dist-info/top_level.txt,sha256=iGuLfZARD_qANcIMfy0tbbrC3EtCg6BSiH8icc3dLWs,6
+bigdl_core_cpp-2.6.0b20241211.dist-info/RECORD,,

bigdl_core_cpp-2.6.0b20241204.dist-info/RECORD

@@ -1,54 +0,0 @@
-bigdl/cpp/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bigdl/cpp/convert_hf_to_gguf.py,sha256=5DxO33LLwJX4aYpWMwTTTvnZd2JinTWatBwnW54A8iQ,189773
-bigdl/cpp/convert_hf_to_gguf_update.py,sha256=pKKPaDe8Dhsvcu_ofSPVEgZ6Ojgk8P9bmDIFi1Hm7lo,15503
-bigdl/cpp/convert_llama_ggml_to_gguf.py,sha256=VyHM3jMYwzM5uQByh-W2DKHEXiwQDk8RBonpdbBL5l8,19734
-bigdl/cpp/convert_lora_to_gguf.py,sha256=bc-D5-lINVo6SXrt-Lws8wdWeIfwcdA7GBptP3MttcM,14775
-bigdl/cpp/gguf-py/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bigdl/cpp/gguf-py/gguf/__init__.py,sha256=h5GWs6SMXYR8giWZ7MTZzAc3hYsIJF-HAkdxtgXLOPo,228
-bigdl/cpp/gguf-py/gguf/constants.py,sha256=J1dMukNMfNKF_yEbjvOSQpYODfifY05TwdiTQqHw27E,50556
-bigdl/cpp/gguf-py/gguf/gguf.py,sha256=V5jY968TEJn6GJHVdjzH0_aIkZ1QC967vPdHDKDoxZw,491
-bigdl/cpp/gguf-py/gguf/gguf_reader.py,sha256=N3LnQQ30t-S0U85-EvZZzIBfHzo0XuyFVUltdg7Sj3c,12680
-bigdl/cpp/gguf-py/gguf/gguf_writer.py,sha256=VZneSoXRxmxCFP55CMIqBws4XgboeC8sJeGNoMHy6Uc,35976
-bigdl/cpp/gguf-py/gguf/lazy.py,sha256=kckbqp8tj7NMkDNwePxwI_1WxK6qH5pMk9p7Lu8cj6A,8816
-bigdl/cpp/gguf-py/gguf/metadata.py,sha256=wtquhynkyH8R7m3zxgeSUe2bnaekJi6HoCMiYJfJBmk,26232
-bigdl/cpp/gguf-py/gguf/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bigdl/cpp/gguf-py/gguf/quants.py,sha256=8broI4A53_Zbd8nQ-a6Qcw2TY4T8XgqHhHFq-J6E9eY,58524
-bigdl/cpp/gguf-py/gguf/tensor_mapping.py,sha256=V2UiBqnPVfFeOK_sg9JY6mXqOLBhCsHGJy7kPHn5jMQ,31548
-bigdl/cpp/gguf-py/gguf/utility.py,sha256=LAwLstUlosYQ1oX9LlJZ-1uCmwyNtOFcJfXpkLnju0k,3003
-bigdl/cpp/gguf-py/gguf/vocab.py,sha256=FtNcm8M5aX9RIr6rRR6UXsUlKMagRUC2xnIWb-xu6rI,19511
-bigdl/cpp/libs/common.lib,sha256=fu4v0lhqD5X44oleXPd2r_Hb7vM6oxu7GWgHZEN5wBA,4057932
-bigdl/cpp/libs/ggml.dll,sha256=Yqvqkmai6yfxHPNj9tQIGQy5UpDxQr3ovxWN6dXNPjo,5747200
-bigdl/cpp/libs/llama-batched.exe,sha256=Gmapigvr9e5mRnTKQQoMiOPM0MQ2Oo-Jbv_Aj3C7uBI,771072
-bigdl/cpp/libs/llama-bench.exe,sha256=bPgiGEdSVK1UTDaTu1UjAx09k4rU9ddnWd5MLFI0G6U,336896
-bigdl/cpp/libs/llama-cli.exe,sha256=zOZkZowVzARtz_uGGMcZjgN2HWYiPNy-r-hYLnH-S9o,906752
-bigdl/cpp/libs/llama-embedding.exe,sha256=d4Z12U85skGuHWP2OnsLZb3GWrq7Imf0jlXEFe67aGo,785408
-bigdl/cpp/libs/llama-gguf.exe,sha256=_SFeDN1R9l-6JlAD4cqDpHvJJQPz9umomskj4WDV1EU,66560
-bigdl/cpp/libs/llama-llava-cli.exe,sha256=fgf_QoWWvEJUE1jivwfNv-4kNTKlUUpknBWrVO2x41M,1091072
-bigdl/cpp/libs/llama-lookup.exe,sha256=jk7gecgU4-JMPQzvuZ_5Ua_DADcREsZJJKJDaODv3S4,851968
-bigdl/cpp/libs/llama-ls-sycl-device.exe,sha256=0T3nyCsh3ipx3MAg2jnKJ9s8zpFcHqUp2VVTKgPFWPU,10240
-bigdl/cpp/libs/llama-minicpmv-cli.exe,sha256=kAXSj0uQPKQTyilHF9v8O3uzEd_3Fz682UKJQEPBx4Y,1082880
-bigdl/cpp/libs/llama-perplexity.exe,sha256=QSbQlsT097fzly9acGyuRrFbj-WRPHnoH1aSnfpNuUg,915456
-bigdl/cpp/libs/llama-quantize.exe,sha256=CMhHE0XPHcX3yGf1dDnRVEsYu6A5p_u9Ehm_2xnwfJk,227840
-bigdl/cpp/libs/llama-server.exe,sha256=vzkcVuu80vwff2jTbnAbh5oqtCgftNDzhX5H1yaIHss,2143744
-bigdl/cpp/libs/llama-simple.exe,sha256=dfsypF6HfCj3mzLxGGgb-MNdlb0doLcqJfzTITYKi7o,763392
-bigdl/cpp/libs/llama-speculative.exe,sha256=Z-QEL9Ica-cigbvXJV-z0ouAzzySSY5QbbGAvJU1IEU,849920
-bigdl/cpp/libs/llama-tokenize.exe,sha256=qd8XY05zRbbmu9S2YXE-NCjZjYg8HRUkG_fJM0iPz9I,200704
-bigdl/cpp/libs/llama.dll,sha256=sjV8oupuQX40hruymke6Gr8hpBmbBUnp1_F3XtHevQg,2217984
-bigdl/cpp/libs/llava_shared.dll,sha256=1gxdRqpsm18_PDOqj6WYhf9QXAG49nz0bqyc-L2LkgE,404480
-bigdl/cpp/libs/ollama.exe,sha256=kRJZ_YpcUh8dAwmRNX9T_GdPB2Wqkt2B5aknoCm4Gn8,65494851
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu/ggml.dll,sha256=EcgDjYmlAZ1y2onVuQaVzFwmVKORQbAYapdQFrEi77E,5703680
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu/llama.dll,sha256=ccXu6uypObtwy7EOlsGBoFp6RCWqjfnwziCgT_rcfZk,2217984
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu/ollama_llama_server.exe,sha256=djHmEslhDOBSliETcM6xB3fzkyaC_O46w-s3kZOVVEk,1376768
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx/ggml.dll,sha256=X8NcVQRPqasgn6EgagaN3rLshBRZTWB1SeLamJsXXEc,5703680
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx/llama.dll,sha256=-9vr01rKVrFo3g4nvTEOCFeVOj7RMlJed2rLrWaZMV8,2217984
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx/ollama_llama_server.exe,sha256=1gl5SaFXe0W4kq0CJ14ffSl3abx4ux9wiqXfgkxk7lc,1376768
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx2/ggml.dll,sha256=08UQw5mL9lA2QtfOV-Io9cLFhShGk5U6umJkQroSCPU,5703680
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx2/llama.dll,sha256=iqPp4mL-l_v_Bi2rRHxdqraqW89vPyfgL6rR3C18djI,2217984
-bigdl/cpp/libs/dist/windows-amd64/lib/ollama/runners/cpu_avx2/ollama_llama_server.exe,sha256=uFHFQrpYs6a-nYPbotrkX1Soum3T-gyPiR549E1ukiI,1376768
-bigdl_core_cpp-2.6.0b20241204.data/scripts/init-llama-cpp.bat,sha256=U0h6RifZxL3GGJp-0dxdZapQIvXUATSj644CURJL-lg,751
-bigdl_core_cpp-2.6.0b20241204.data/scripts/init-llama-cpp.ps1,sha256=JFOylLxO4MKpllHhdbPuJ1xHi9azxDpzdJns8JtZpkU,501
-bigdl_core_cpp-2.6.0b20241204.data/scripts/init-ollama.bat,sha256=2rScaP2-_yWuAnc86NyX-i6pg7BxjvLc96Kt7HuvQVM,603
-bigdl_core_cpp-2.6.0b20241204.dist-info/METADATA,sha256=zybu715duUbDtl0TbBZfoY1k1Rv1FGsHGLd0son1f8g,652
-bigdl_core_cpp-2.6.0b20241204.dist-info/WHEEL,sha256=z8gukVdnGwjcwo0VnsfJMrhPu5QJT68VcMWmAgvAufw,97
-bigdl_core_cpp-2.6.0b20241204.dist-info/top_level.txt,sha256=iGuLfZARD_qANcIMfy0tbbrC3EtCg6BSiH8icc3dLWs,6
-bigdl_core_cpp-2.6.0b20241204.dist-info/RECORD,,