The latest series of Code-Specific Qwen models, with significant improvements in code generation, code reasoning, and code fixing.
tools
0.5b
1.5b
3b
7b
14b
32b
966.2K Pulls Updated 5 weeks ago
29d8c98fa6b0 · 986MB
-
general.architectureqwen2
-
general.base_model.0.nameQwen2.5 Coder 1.5B
-
general.base_model.0.organizationQwen
-
general.base_model.0.repo_urlhttps://huggingface.co/Qwen/Qwen2.5-Coder-1.5B
-
general.base_model.count1
-
general.basenameQwen2.5-Coder
-
general.file_type15
-
general.finetuneInstruct
-
general.languages[en]
-
general.licenseapache-2.0
-
general.license.linkhttps://huggingface.co/Qwen/Qwen2.5-Coder-1.5B-Instruct/blob/main/LICENSE
-
general.nameQwen2.5 Coder 1.5B Instruct
-
general.quantization_version2
-
general.size_label1.5B
-
general.tags[code, codeqwen, chat, qwen, qwen-coder, ...]
-
general.typemodel
-
qwen2.attention.head_count12
-
qwen2.attention.head_count_kv2
-
qwen2.attention.layer_norm_rms_epsilon1e-06
-
qwen2.block_count28
-
qwen2.context_length32768
-
qwen2.embedding_length1536
-
qwen2.feed_forward_length8960
-
qwen2.rope.freq_base1e+06
-
tokenizer.ggml.add_bos_tokenfalse
-
tokenizer.ggml.bos_token_id151643
-
tokenizer.ggml.eos_token_id151645
-
tokenizer.ggml.merges[Ġ Ġ, ĠĠ ĠĠ, i n, Ġ t, ĠĠĠĠ ĠĠĠĠ, ...]
-
tokenizer.ggml.modelgpt2
-
tokenizer.ggml.padding_token_id151643
-
tokenizer.ggml.preqwen2
-
tokenizer.ggml.token_type[1, 1, 1, 1, 1, ...]
-
tokenizer.ggml.tokens[!, ", #, $, %, ...]
-
NameTypeShape
-
token_embd.weightQ6_K[1536, 151936]
-
blk.0.attn_k.biasF32[256]
-
blk.0.attn_k.weightQ4_K[1536, 256]
-
blk.0.attn_norm.weightF32[1536]
-
blk.0.attn_output.weightQ4_K[1536, 1536]
-
blk.0.attn_q.biasF32[1536]
-
blk.0.attn_q.weightQ4_K[1536, 1536]
-
blk.0.attn_v.biasF32[256]
-
blk.0.attn_v.weightQ6_K[1536, 256]
-
blk.0.ffn_down.weightQ6_K[8960, 1536]
-
blk.0.ffn_gate.weightQ4_K[1536, 8960]
-
blk.0.ffn_norm.weightF32[1536]
-
blk.0.ffn_up.weightQ4_K[1536, 8960]
-
blk.1.attn_k.biasF32[256]
-
blk.1.attn_k.weightQ4_K[1536, 256]
-
blk.1.attn_norm.weightF32[1536]
-
blk.1.attn_output.weightQ4_K[1536, 1536]
-
blk.1.attn_q.biasF32[1536]
-
blk.1.attn_q.weightQ4_K[1536, 1536]
-
blk.1.attn_v.biasF32[256]
-
blk.1.attn_v.weightQ6_K[1536, 256]
-
blk.1.ffn_down.weightQ6_K[8960, 1536]
-
blk.1.ffn_gate.weightQ4_K[1536, 8960]
-
blk.1.ffn_norm.weightF32[1536]
-
blk.1.ffn_up.weightQ4_K[1536, 8960]
-
blk.2.attn_k.biasF32[256]
-
blk.2.attn_k.weightQ4_K[1536, 256]
-
blk.2.attn_norm.weightF32[1536]
-
blk.2.attn_output.weightQ4_K[1536, 1536]
-
blk.2.attn_q.biasF32[1536]
-
blk.2.attn_q.weightQ4_K[1536, 1536]
-
blk.2.attn_v.biasF32[256]
-
blk.2.attn_v.weightQ4_K[1536, 256]
-
blk.2.ffn_down.weightQ4_K[8960, 1536]
-
blk.2.ffn_gate.weightQ4_K[1536, 8960]
-
blk.2.ffn_norm.weightF32[1536]
-
blk.2.ffn_up.weightQ4_K[1536, 8960]
-
blk.3.attn_k.biasF32[256]
-
blk.3.attn_k.weightQ4_K[1536, 256]
-
blk.3.attn_norm.weightF32[1536]
-
blk.3.attn_output.weightQ4_K[1536, 1536]
-
blk.3.attn_q.biasF32[1536]
-
blk.3.attn_q.weightQ4_K[1536, 1536]
-
blk.3.attn_v.biasF32[256]
-
blk.3.attn_v.weightQ4_K[1536, 256]
-
blk.3.ffn_down.weightQ4_K[8960, 1536]
-
blk.3.ffn_gate.weightQ4_K[1536, 8960]
-
blk.3.ffn_norm.weightF32[1536]
-
blk.3.ffn_up.weightQ4_K[1536, 8960]
-
blk.4.attn_k.biasF32[256]
-
blk.4.attn_k.weightQ4_K[1536, 256]
-
blk.4.attn_norm.weightF32[1536]
-
blk.4.attn_output.weightQ4_K[1536, 1536]
-
blk.4.attn_q.biasF32[1536]
-
blk.4.attn_q.weightQ4_K[1536, 1536]
-
blk.4.attn_v.biasF32[256]
-
blk.4.attn_v.weightQ4_K[1536, 256]
-
blk.4.ffn_down.weightQ4_K[8960, 1536]
-
blk.4.ffn_gate.weightQ4_K[1536, 8960]
-
blk.4.ffn_norm.weightF32[1536]
-
blk.4.ffn_up.weightQ4_K[1536, 8960]
-
blk.5.attn_k.biasF32[256]
-
blk.5.attn_k.weightQ4_K[1536, 256]
-
blk.5.attn_norm.weightF32[1536]
-
blk.5.attn_output.weightQ4_K[1536, 1536]
-
blk.5.attn_q.biasF32[1536]
-
blk.5.attn_q.weightQ4_K[1536, 1536]
-
blk.5.attn_v.biasF32[256]
-
blk.5.attn_v.weightQ6_K[1536, 256]
-
blk.5.ffn_down.weightQ6_K[8960, 1536]
-
blk.5.ffn_gate.weightQ4_K[1536, 8960]
-
blk.5.ffn_norm.weightF32[1536]
-
blk.5.ffn_up.weightQ4_K[1536, 8960]
-
blk.6.attn_k.biasF32[256]
-
blk.6.attn_k.weightQ4_K[1536, 256]
-
blk.6.attn_norm.weightF32[1536]
-
blk.6.attn_output.weightQ4_K[1536, 1536]
-
blk.6.attn_q.biasF32[1536]
-
blk.6.attn_q.weightQ4_K[1536, 1536]
-
blk.6.attn_v.biasF32[256]
-
blk.6.attn_v.weightQ6_K[1536, 256]
-
blk.6.ffn_down.weightQ6_K[8960, 1536]
-
blk.6.ffn_gate.weightQ4_K[1536, 8960]
-
blk.6.ffn_norm.weightF32[1536]
-
blk.6.ffn_up.weightQ4_K[1536, 8960]
-
blk.7.attn_k.biasF32[256]
-
blk.7.attn_k.weightQ4_K[1536, 256]
-
blk.7.attn_norm.weightF32[1536]
-
blk.7.attn_output.weightQ4_K[1536, 1536]
-
blk.7.attn_q.biasF32[1536]
-
blk.7.attn_q.weightQ4_K[1536, 1536]
-
blk.7.attn_v.biasF32[256]
-
blk.7.attn_v.weightQ6_K[1536, 256]
-
blk.7.ffn_down.weightQ6_K[8960, 1536]
-
blk.7.ffn_gate.weightQ4_K[1536, 8960]
-
blk.7.ffn_norm.weightF32[1536]
-
blk.7.ffn_up.weightQ4_K[1536, 8960]
-
blk.8.attn_k.biasF32[256]
-
blk.8.attn_k.weightQ4_K[1536, 256]
-
blk.8.attn_norm.weightF32[1536]
-
blk.8.attn_output.weightQ4_K[1536, 1536]
-
blk.8.attn_q.biasF32[1536]
-
blk.8.attn_q.weightQ4_K[1536, 1536]
-
blk.8.attn_v.biasF32[256]
-
blk.8.attn_v.weightQ6_K[1536, 256]
-
blk.8.ffn_down.weightQ6_K[8960, 1536]
-
blk.8.ffn_gate.weightQ4_K[1536, 8960]
-
blk.8.ffn_norm.weightF32[1536]
-
blk.8.ffn_up.weightQ4_K[1536, 8960]
-
blk.9.attn_k.biasF32[256]
-
blk.9.attn_k.weightQ4_K[1536, 256]
-
blk.9.attn_norm.weightF32[1536]
-
blk.9.attn_output.weightQ4_K[1536, 1536]
-
blk.9.attn_q.biasF32[1536]
-
blk.9.attn_q.weightQ4_K[1536, 1536]
-
blk.9.attn_v.biasF32[256]
-
blk.9.attn_v.weightQ6_K[1536, 256]
-
blk.9.ffn_down.weightQ6_K[8960, 1536]
-
blk.9.ffn_gate.weightQ4_K[1536, 8960]
-
blk.9.ffn_norm.weightF32[1536]
-
blk.9.ffn_up.weightQ4_K[1536, 8960]
-
blk.10.attn_k.biasF32[256]
-
blk.10.attn_k.weightQ4_K[1536, 256]
-
blk.10.attn_norm.weightF32[1536]
-
blk.10.attn_output.weightQ4_K[1536, 1536]
-
blk.10.attn_q.biasF32[1536]
-
blk.10.attn_q.weightQ4_K[1536, 1536]
-
blk.10.attn_v.biasF32[256]
-
blk.10.attn_v.weightQ6_K[1536, 256]
-
blk.10.ffn_down.weightQ6_K[8960, 1536]
-
blk.10.ffn_gate.weightQ4_K[1536, 8960]
-
blk.10.ffn_norm.weightF32[1536]
-
blk.10.ffn_up.weightQ4_K[1536, 8960]
-
blk.11.attn_k.biasF32[256]
-
blk.11.attn_k.weightQ4_K[1536, 256]
-
blk.11.attn_norm.weightF32[1536]
-
blk.11.attn_output.weightQ4_K[1536, 1536]
-
blk.11.attn_q.biasF32[1536]
-
blk.11.attn_q.weightQ4_K[1536, 1536]
-
blk.11.attn_v.biasF32[256]
-
blk.11.attn_v.weightQ4_K[1536, 256]
-
blk.11.ffn_down.weightQ4_K[8960, 1536]
-
blk.11.ffn_gate.weightQ4_K[1536, 8960]
-
blk.11.ffn_norm.weightF32[1536]
-
blk.11.ffn_up.weightQ4_K[1536, 8960]
-
blk.12.attn_k.biasF32[256]
-
blk.12.attn_k.weightQ4_K[1536, 256]
-
blk.12.attn_norm.weightF32[1536]
-
blk.12.attn_output.weightQ4_K[1536, 1536]
-
blk.12.attn_q.biasF32[1536]
-
blk.12.attn_q.weightQ4_K[1536, 1536]
-
blk.12.attn_v.biasF32[256]
-
blk.12.attn_v.weightQ4_K[1536, 256]
-
blk.12.ffn_down.weightQ4_K[8960, 1536]
-
blk.12.ffn_gate.weightQ4_K[1536, 8960]
-
blk.12.ffn_norm.weightF32[1536]
-
blk.12.ffn_up.weightQ4_K[1536, 8960]
-
blk.13.attn_k.biasF32[256]
-
blk.13.attn_k.weightQ4_K[1536, 256]
-
blk.13.attn_norm.weightF32[1536]
-
blk.13.attn_output.weightQ4_K[1536, 1536]
-
blk.13.attn_q.biasF32[1536]
-
blk.13.attn_q.weightQ4_K[1536, 1536]
-
blk.13.attn_v.biasF32[256]
-
blk.13.attn_v.weightQ6_K[1536, 256]
-
blk.13.ffn_down.weightQ6_K[8960, 1536]
-
blk.13.ffn_gate.weightQ4_K[1536, 8960]
-
blk.13.ffn_norm.weightF32[1536]
-
blk.13.ffn_up.weightQ4_K[1536, 8960]
-
blk.14.attn_k.biasF32[256]
-
blk.14.attn_k.weightQ4_K[1536, 256]
-
blk.14.attn_norm.weightF32[1536]
-
blk.14.attn_output.weightQ4_K[1536, 1536]
-
blk.14.attn_q.biasF32[1536]
-
blk.14.attn_q.weightQ4_K[1536, 1536]
-
blk.14.attn_v.biasF32[256]
-
blk.14.attn_v.weightQ4_K[1536, 256]
-
blk.14.ffn_down.weightQ4_K[8960, 1536]
-
blk.14.ffn_gate.weightQ4_K[1536, 8960]
-
blk.14.ffn_norm.weightF32[1536]
-
blk.14.ffn_up.weightQ4_K[1536, 8960]
-
blk.15.attn_k.biasF32[256]
-
blk.15.attn_k.weightQ4_K[1536, 256]
-
blk.15.attn_norm.weightF32[1536]
-
blk.15.attn_output.weightQ4_K[1536, 1536]
-
blk.15.attn_q.biasF32[1536]
-
blk.15.attn_q.weightQ4_K[1536, 1536]
-
blk.15.attn_v.biasF32[256]
-
blk.15.attn_v.weightQ4_K[1536, 256]
-
blk.15.ffn_down.weightQ4_K[8960, 1536]
-
blk.15.ffn_gate.weightQ4_K[1536, 8960]
-
blk.15.ffn_norm.weightF32[1536]
-
blk.15.ffn_up.weightQ4_K[1536, 8960]
-
blk.16.attn_k.biasF32[256]
-
blk.16.attn_k.weightQ4_K[1536, 256]
-
blk.16.attn_norm.weightF32[1536]
-
blk.16.attn_output.weightQ4_K[1536, 1536]
-
blk.16.attn_q.biasF32[1536]
-
blk.16.attn_q.weightQ4_K[1536, 1536]
-
blk.16.attn_v.biasF32[256]
-
blk.16.attn_v.weightQ6_K[1536, 256]
-
blk.16.ffn_down.weightQ6_K[8960, 1536]
-
blk.16.ffn_gate.weightQ4_K[1536, 8960]
-
blk.16.ffn_norm.weightF32[1536]
-
blk.16.ffn_up.weightQ4_K[1536, 8960]
-
blk.17.attn_k.biasF32[256]
-
blk.17.attn_k.weightQ4_K[1536, 256]
-
blk.17.attn_norm.weightF32[1536]
-
blk.17.attn_output.weightQ4_K[1536, 1536]
-
blk.17.attn_q.biasF32[1536]
-
blk.17.attn_q.weightQ4_K[1536, 1536]
-
blk.17.attn_v.biasF32[256]
-
blk.17.attn_v.weightQ4_K[1536, 256]
-
blk.17.ffn_down.weightQ4_K[8960, 1536]
-
blk.17.ffn_gate.weightQ4_K[1536, 8960]
-
blk.17.ffn_norm.weightF32[1536]
-
blk.17.ffn_up.weightQ4_K[1536, 8960]
-
blk.18.attn_k.biasF32[256]
-
blk.18.attn_k.weightQ4_K[1536, 256]
-
blk.18.attn_norm.weightF32[1536]
-
blk.18.attn_output.weightQ4_K[1536, 1536]
-
blk.18.attn_q.biasF32[1536]
-
blk.18.attn_q.weightQ4_K[1536, 1536]
-
blk.18.attn_v.biasF32[256]
-
blk.18.attn_v.weightQ4_K[1536, 256]
-
blk.18.ffn_down.weightQ4_K[8960, 1536]
-
blk.18.ffn_gate.weightQ4_K[1536, 8960]
-
blk.18.ffn_norm.weightF32[1536]
-
blk.18.ffn_up.weightQ4_K[1536, 8960]
-
blk.19.attn_k.biasF32[256]
-
blk.19.attn_k.weightQ4_K[1536, 256]
-
blk.19.attn_norm.weightF32[1536]
-
blk.19.attn_output.weightQ4_K[1536, 1536]
-
blk.19.attn_q.biasF32[1536]
-
blk.19.attn_q.weightQ4_K[1536, 1536]
-
blk.19.attn_v.biasF32[256]
-
blk.19.attn_v.weightQ6_K[1536, 256]
-
blk.19.ffn_down.weightQ6_K[8960, 1536]
-
blk.19.ffn_gate.weightQ4_K[1536, 8960]
-
blk.19.ffn_norm.weightF32[1536]
-
blk.19.ffn_up.weightQ4_K[1536, 8960]
-
blk.20.attn_k.biasF32[256]
-
blk.20.attn_k.weightQ4_K[1536, 256]
-
blk.20.attn_norm.weightF32[1536]
-
blk.20.attn_output.weightQ4_K[1536, 1536]
-
blk.20.attn_q.biasF32[1536]
-
blk.20.attn_q.weightQ4_K[1536, 1536]
-
blk.20.attn_v.biasF32[256]
-
blk.20.attn_v.weightQ4_K[1536, 256]
-
blk.20.ffn_down.weightQ4_K[8960, 1536]
-
blk.20.ffn_gate.weightQ4_K[1536, 8960]
-
blk.20.ffn_norm.weightF32[1536]
-
blk.20.ffn_up.weightQ4_K[1536, 8960]
-
blk.21.attn_k.biasF32[256]
-
blk.21.attn_k.weightQ4_K[1536, 256]
-
blk.21.attn_norm.weightF32[1536]
-
blk.21.attn_output.weightQ4_K[1536, 1536]
-
blk.21.attn_q.biasF32[1536]
-
blk.21.attn_q.weightQ4_K[1536, 1536]
-
blk.21.attn_v.biasF32[256]
-
blk.21.attn_v.weightQ6_K[1536, 256]
-
blk.21.ffn_down.weightQ6_K[8960, 1536]
-
blk.21.ffn_gate.weightQ4_K[1536, 8960]
-
blk.21.ffn_norm.weightF32[1536]
-
blk.21.ffn_up.weightQ4_K[1536, 8960]
-
blk.22.attn_k.biasF32[256]
-
blk.22.attn_k.weightQ4_K[1536, 256]
-
blk.22.attn_norm.weightF32[1536]
-
blk.22.attn_output.weightQ4_K[1536, 1536]
-
blk.22.attn_q.biasF32[1536]
-
blk.22.attn_q.weightQ4_K[1536, 1536]
-
blk.22.attn_v.biasF32[256]
-
blk.22.attn_v.weightQ4_K[1536, 256]
-
blk.22.ffn_down.weightQ4_K[8960, 1536]
-
blk.22.ffn_gate.weightQ4_K[1536, 8960]
-
blk.22.ffn_norm.weightF32[1536]
-
blk.22.ffn_up.weightQ4_K[1536, 8960]
-
blk.23.attn_k.biasF32[256]
-
blk.23.attn_k.weightQ4_K[1536, 256]
-
blk.23.attn_norm.weightF32[1536]
-
blk.23.attn_output.weightQ4_K[1536, 1536]
-
blk.23.attn_q.biasF32[1536]
-
blk.23.attn_q.weightQ4_K[1536, 1536]
-
blk.23.attn_v.biasF32[256]
-
blk.23.attn_v.weightQ4_K[1536, 256]
-
blk.23.ffn_down.weightQ4_K[8960, 1536]
-
blk.23.ffn_gate.weightQ4_K[1536, 8960]
-
blk.23.ffn_norm.weightF32[1536]
-
blk.23.ffn_up.weightQ4_K[1536, 8960]
-
blk.24.attn_k.biasF32[256]
-
blk.24.attn_k.weightQ4_K[1536, 256]
-
blk.24.attn_norm.weightF32[1536]
-
blk.24.attn_output.weightQ4_K[1536, 1536]
-
blk.24.attn_q.biasF32[1536]
-
blk.24.attn_q.weightQ4_K[1536, 1536]
-
blk.24.attn_v.biasF32[256]
-
blk.24.attn_v.weightQ6_K[1536, 256]
-
blk.24.ffn_down.weightQ6_K[8960, 1536]
-
blk.24.ffn_gate.weightQ4_K[1536, 8960]
-
blk.24.ffn_norm.weightF32[1536]
-
blk.24.ffn_up.weightQ4_K[1536, 8960]
-
blk.25.attn_k.biasF32[256]
-
blk.25.attn_k.weightQ4_K[1536, 256]
-
blk.25.attn_norm.weightF32[1536]
-
blk.25.attn_output.weightQ4_K[1536, 1536]
-
blk.25.attn_q.biasF32[1536]
-
blk.25.attn_q.weightQ4_K[1536, 1536]
-
blk.25.attn_v.biasF32[256]
-
blk.25.attn_v.weightQ4_K[1536, 256]
-
blk.25.ffn_down.weightQ4_K[8960, 1536]
-
blk.25.ffn_gate.weightQ4_K[1536, 8960]
-
blk.25.ffn_norm.weightF32[1536]
-
blk.25.ffn_up.weightQ4_K[1536, 8960]
-
blk.26.attn_k.biasF32[256]
-
blk.26.attn_k.weightQ4_K[1536, 256]
-
blk.26.attn_norm.weightF32[1536]
-
blk.26.attn_output.weightQ4_K[1536, 1536]
-
blk.26.attn_q.biasF32[1536]
-
blk.26.attn_q.weightQ4_K[1536, 1536]
-
blk.26.attn_v.biasF32[256]
-
blk.26.attn_v.weightQ4_K[1536, 256]
-
blk.26.ffn_down.weightQ4_K[8960, 1536]
-
blk.26.ffn_gate.weightQ4_K[1536, 8960]
-
blk.26.ffn_norm.weightF32[1536]
-
blk.26.ffn_up.weightQ4_K[1536, 8960]
-
blk.27.attn_k.biasF32[256]
-
blk.27.attn_k.weightQ4_K[1536, 256]
-
blk.27.attn_norm.weightF32[1536]
-
blk.27.attn_output.weightQ4_K[1536, 1536]
-
blk.27.attn_q.biasF32[1536]
-
blk.27.attn_q.weightQ4_K[1536, 1536]
-
blk.27.attn_v.biasF32[256]
-
blk.27.attn_v.weightQ6_K[1536, 256]
-
blk.27.ffn_down.weightQ6_K[8960, 1536]
-
blk.27.ffn_gate.weightQ4_K[1536, 8960]
-
blk.27.ffn_norm.weightF32[1536]
-
blk.27.ffn_up.weightQ4_K[1536, 8960]
-
output_norm.weightF32[1536]
Metadata
Tensor
blk.0
blk.1
blk.2
blk.3
blk.4
blk.5
blk.6
blk.7
blk.8
blk.9
blk.10
blk.11
blk.12
blk.13
blk.14
blk.15
blk.16
blk.17
blk.18
blk.19
blk.20
blk.21
blk.22
blk.23
blk.24
blk.25
blk.26
blk.27