Thomas Scialom, Rui Hou, Naman Goyal, Binh Tang, Angela Fan, Vedanuj Goswami, Cristian Ferrer, Sharan Narang, Saghar Hosseini, Pushkar Mishra, Adina Williams, Yixin Nie, and 58 more
We lifted 52 functions out of this paper's own repositories and ran 31 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| IBM/Dromedary | canonical | 16 of 16 |
| squeezeailab/squeezellm | canonical | 5 of 15 |
| meetyou-ai-lab/can-mc-evaluate-llms | — | 4 of 7 |
| Lightning-AI/lit-gpt | — | 4 of 5 |
| glb400/Toy-RecLM | — | 1 of 5 |
| zurichnlp/contradecode | — | 1 of 1 |
| young-geng/easylm | — | 0 of 2 |
| xverse-ai/xverse-13b | — | 0 of 1 |
| Function | Status | Where it lives |
|---|---|---|
| repeat_kv | Ran | IBM/Dromedary/training/llama_with_flash_attn.py pointer only (licence: GPL-3.0) · get_code("30d7eec482ebf6b1") |
| round_to_nearest_pole_sim | Ran | squeezeailab/squeezellm/squeezellm/quant.py code served (permissive licence) · get_code("cdc0ad3109c8984b") |
| Config | Ran | Lightning-AI/lit-gpt/litgpt/model.py code served (permissive licence) · get_code("bf78fefededc09f7") |
| FeedForward | Ran | glb400/Toy-RecLM/model.py code served (permissive licence) · get_code("9715deb8f27d4266") |
| LlamaConfig | Ran | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("1bcd8d227ad19b17") |
| LlamaMLP | Ran | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("50260ad39d7c5e06") |
| LlamaRMSNorm | Ran | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("98b6bb174c074dae") |
| LlamaRotaryEmbedding | Ran | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("26ec53f064adbca5") |
| PromptTemplate | Ran | zurichnlp/contradecode/translation_models/llama.py code served (permissive licence) · get_code("4d273bfb0feb28e0") |
| RMSNorm | Ran | Lightning-AI/lit-gpt/litgpt/model.py code served (permissive licence) · get_code("1e7100681c300ca2") |
| apply_rotary_emb | Ran | IBM/Dromedary/llama_dromedary/llama_dromedary/model.py pointer only (licence: GPL-3.0) · get_code("b47d48e431b34acd") |
| apply_rotary_pos_emb | Ran | IBM/Dromedary/training/llama_with_flash_attn.py pointer only (licence: GPL-3.0) · get_code("e34097675d132bc1") |
| check_indicator_and_length | Ran | Lightning-AI/lit-gpt/litgpt/model.py code served (permissive licence) · get_code("faa77017b4beec30") |
| extract_alpaca_dataset | Ran | IBM/Dromedary/training/data_utils_sft.py pointer only (licence: GPL-3.0) · get_code("b5445674ab17410f") |
| extract_dromedary_dataset | Ran | IBM/Dromedary/training/data_utils_sft.py pointer only (licence: GPL-3.0) · get_code("48456f10725088b5") |
| extract_unnatural_instructions_data | Ran | IBM/Dromedary/training/data_utils_sft.py pointer only (licence: GPL-3.0) · get_code("f82430123a91cf21") |
| find_layers | Ran | squeezeailab/squeezellm/squeezellm/modelutils.py code served (permissive licence) · get_code("a9e7f2cdf016b88b") |
| find_multiple | Ran | Lightning-AI/lit-gpt/litgpt/model.py code served (permissive licence) · get_code("ffe8d3a5e6f4b477") |
| generate_prompt | Ran | IBM/Dromedary/inference/run_stream_chatbot_demo.py pointer only (licence: GPL-3.0) · get_code("55e55d5b9201bf5c") |
| generate_prompt | Ran | IBM/Dromedary/mc_evaluation/evaluate_hhh_eval.py pointer only (licence: GPL-3.0) · get_code("c478c1e77422f0de") |
| get_log_prob | Ran | IBM/Dromedary/mc_evaluation/evaluate_hhh_eval.py pointer only (licence: GPL-3.0) · get_code("0c0f540ee97855aa") |
| get_log_prob | Ran | IBM/Dromedary/mc_evaluation/evaluate_truthfulqa_mc.py pointer only (licence: GPL-3.0) · get_code("47407f4c74f590df") |
| measure_multiple_choice_grade | Ran | IBM/Dromedary/mc_evaluation/evaluate_hhh_eval.py pointer only (licence: GPL-3.0) · get_code("b14dfe7367095841") |
| measure_multiple_choice_grade | Ran | IBM/Dromedary/mc_evaluation/evaluate_truthfulqa_mc.py pointer only (licence: GPL-3.0) · get_code("dfcc630d6e498f6f") |
| precompute_freqs_cis | Ran | IBM/Dromedary/llama_dromedary/llama_dromedary/model.py pointer only (licence: GPL-3.0) · get_code("14a84c2cbfebc413") |
| remove_outliers | Ran | squeezeailab/squeezellm/squeezellm/outliers.py code served (permissive licence) · get_code("303e2b878d149f4a") |
| remove_outliers_by_sensitivity | Ran | squeezeailab/squeezellm/squeezellm/outliers.py code served (permissive licence) · get_code("e3a24217edf7812c") |
| remove_outliers_by_threshold | Ran | squeezeailab/squeezellm/squeezellm/outliers.py code served (permissive licence) · get_code("3d2bc2ac00f3f434") |
| reshape_for_broadcast | Ran | IBM/Dromedary/llama_dromedary/llama_dromedary/model.py pointer only (licence: GPL-3.0) · get_code("70bf6ebaafd266c4") |
| rotate_half | Ran | IBM/Dromedary/training/llama_with_flash_attn.py pointer only (licence: GPL-3.0) · get_code("b99eea6376d1e212") |
| sample_top_p | Ran | IBM/Dromedary/llama_dromedary/llama_dromedary/generation.py pointer only (licence: GPL-3.0) · get_code("8845976729f4c4ee") |
| Attention | Not yet run | glb400/Toy-RecLM/model.py code served (permissive licence) · get_code("4334c46d3961b3fb") |
| FlaxLLaMAModel | Not yet run | young-geng/easylm/EasyLM/models/llama/llama_model.py code served (permissive licence) · get_code("cf00f2eef08595e6") |
| FlaxLLaMAPreTrainedModel | Not yet run | young-geng/easylm/EasyLM/models/llama/llama_model.py code served (permissive licence) · get_code("72032d6f73d90892") |
| LLaMA2_SASRec | Not yet run | glb400/Toy-RecLM/model.py code served (permissive licence) · get_code("4427e517543e4577") |
| LLaMAMLP | Not yet run | Lightning-AI/lit-gpt/litgpt/model.py code served (permissive licence) · get_code("b336e5e4f4b40ceb") |
| LlamaAttention | Not yet run | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("1104581ac75ea9c6") |
| LlamaDecoderLayer | Not yet run | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("660d5e5234e28212") |
| LlamaModel | Not yet run | meetyou-ai-lab/can-mc-evaluate-llms/Embeddings/src/llama/modeling_llama.py code served (permissive licence) · get_code("865ca1b3e9c1aa5c") |
| ModelArgs | Not yet run | glb400/Toy-RecLM/model.py code served (permissive licence) · get_code("75e3a8170f256bb7") |
| TransformerBlock | Not yet run | glb400/Toy-RecLM/model.py code served (permissive licence) · get_code("61aae005c3076a9b") |
| get_c4 | Not yet run | squeezeailab/squeezellm/squeezellm/datautils.py code served (permissive licence) · get_code("92dc709e4e73e133") |
| get_model | Not yet run | squeezeailab/squeezellm/llama.py code served (permissive licence) · get_code("fd802dc39c73084a") |
| get_module_names | Not yet run | squeezeailab/squeezellm/squeezellm/model_parse.py code served (permissive licence) · get_code("b65e0b634a94c717") |
| get_ptb | Not yet run | squeezeailab/squeezellm/squeezellm/datautils.py code served (permissive licence) · get_code("b1e98f962f5ae82d") |
| get_wikitext2 | Not yet run | squeezeailab/squeezellm/squeezellm/datautils.py code served (permissive licence) · get_code("efc347eff314b865") |
| init_model | Not yet run | xverse-ai/xverse-13b/chat_demo.py code served (permissive licence) · get_code("b796bb9465578920") |
| kmeans_fit | Not yet run | squeezeailab/squeezellm/quantization/nuq.py code served (permissive licence) · get_code("77fbadc3b9632874") |
| load_model | Not yet run | squeezeailab/squeezellm/squeezellm/model_parse.py code served (permissive licence) · get_code("b66c43d90a1a3a1b") |
| load_quant | Not yet run | squeezeailab/squeezellm/llama.py code served (permissive licence) · get_code("c0d4637f25eabf00") |
| parse_model | Not yet run | squeezeailab/squeezellm/squeezellm/model_parse.py code served (permissive licence) · get_code("dd982ece7960b018") |
| tiktoken_tokenizer | Not yet run | squeezeailab/squeezellm/models/xgen-7b-8k-base/tokenization_xgen.py code served (permissive licence) · get_code("5445de6acd6035bb") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases. Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closedsource models. We provide a detailed description of our approach to fine-tuning and safety improvements of Llama 2-Chat in order to enable the community to build on our work and contribute to the responsible development of LLMs.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2307.09288")
get_code_for_paper("2307.09288")
have("2307.09288")
Connect an agent — have() is free.