README.md

metadata

license: llama2

What is it?

This is a quantized version of h2oai/h2ogpt-4096-llama2-13b-chat, formatted in GGUF format to be run with llama.cpp and similar inference tools.

Available Formats

Format	Bits	Use case
q8_0	8	Original quant method, 8-bit.

Currently in conversion

Format	Bits	Use case
q3_K_L	3	New k-quant method. Uses GGML_TYPE_Q5_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K
q3_K_M	3	New k-quant method. Uses GGML_TYPE_Q4_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K
q3_K_S	3	New k-quant method. Uses GGML_TYPE_Q3_K for all tensors
q4_0	4	Original quant method, 4-bit.
q4_1	4	Original quant method, 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models.
q4_K_M	4	New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q4_K
q4_K_S	4	New k-quant method. Uses GGML_TYPE_Q4_K for all tensors
q5_0	5	Original quant method, 5-bit. Higher accuracy, higher resource usage and slower inference.
q5_1	5	Original quant method, 5-bit. Even higher accuracy, resource usage and slower inference.
q5_K_M	5	New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q5_K
q5_K_S	5	New k-quant method. Uses GGML_TYPE_Q5_K for all tensors
q6_K	6	New k-quant method. Uses GGML_TYPE_Q8_K for all tensors - 6-bit quantization

Original Model Card

---

inference: false language: - en license: llama2 model_type: llama pipeline_tag: text-generation tags: - facebook - meta - pytorch - llama - llama-2 - h2ogpt

h2oGPT clone of Meta's Llama 2 13B Chat.

Try it live on our h2oGPT demo with side-by-side LLM comparisons and private document chat!

See how it compares to other models on our LLM Leaderboard!

See more at H2O.ai

Model Architecture

LlamaForCausalLM(
  (model): LlamaModel(
    (embed_tokens): Embedding(32000, 5120, padding_idx=0)
    (layers): ModuleList(
      (0-39): 40 x LlamaDecoderLayer(
        (self_attn): LlamaAttention(
          (q_proj): Linear(in_features=5120, out_features=5120, bias=False)
          (k_proj): Linear(in_features=5120, out_features=5120, bias=False)
          (v_proj): Linear(in_features=5120, out_features=5120, bias=False)
          (o_proj): Linear(in_features=5120, out_features=5120, bias=False)
          (rotary_emb): LlamaRotaryEmbedding()
        )
        (mlp): LlamaMLP(
          (gate_proj): Linear(in_features=5120, out_features=13824, bias=False)
          (up_proj): Linear(in_features=5120, out_features=13824, bias=False)
          (down_proj): Linear(in_features=13824, out_features=5120, bias=False)
          (act_fn): SiLUActivation()
        )
        (input_layernorm): LlamaRMSNorm()
        (post_attention_layernorm): LlamaRMSNorm()
      )
    )
    (norm): LlamaRMSNorm()
  )
  (lm_head): Linear(in_features=5120, out_features=32000, bias=False)
)