sampling : add support for backend sampling

common/arg.cpp

@@ -1513,6 +1513,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
             params.sampling.grammar = json_schema_to_grammar(json::parse(schema));
         }
     ).set_sparam());
+    add_opt(common_arg(
+        {"--backend-sampling"},
+        "enable backend sampling (default: disabled)",
+        [](common_params & params) {
+            params.sampling.backend_sampling = true;
+        }
+    ).set_sparam());
     add_opt(common_arg(
         {"--pooling"}, "{none,mean,cls,last,rank}",
         "pooling type for embeddings, use model default if unspecified",

common/common.cpp

@@ -950,31 +950,40 @@ std::vector<common_file_info> fs_list_files(const std::string & path) {
 // Model utils
 //
 
-static inline void common_init_sampler_from_model(
+// TODO: move to common/sampling
+static void common_init_sampler_from_model(
     const llama_model * model,
     common_params_sampling & sparams) {
 
     const uint64_t config = sparams.user_sampling_config;
 
     auto get_int32 = [&](const char * key, int32_t & dst, uint64_t user_config) {
-        if (config & user_config) return;
+        if (config & user_config) {
+            return;
+        }
 
         char buf[64] = {0};
         if (llama_model_meta_val_str(model, key, buf, sizeof(buf)) > 0) {
             char * end = nullptr;
             int32_t v = strtol(buf, &end, 10);
-            if (end && end != buf) dst = v;
+            if (end && end != buf) {
+                dst = v;
+            }
         }
     };
 
     auto get_float = [&](const char * key, float & dst, uint64_t user_config) {
-        if (config & user_config) return;
+        if (config & user_config) {
+            return;
+        }
 
         char buf[128] = {0};
         if (llama_model_meta_val_str(model, key, buf, sizeof(buf)) > 0) {
             char * end = nullptr;
             float v = strtof(buf, &end);
-            if (end && end != buf) dst = v;
+            if (end && end != buf) {
+                dst = v;
+            }
         }
     };
 
@@ -1002,31 +1011,130 @@ static inline void common_init_sampler_from_model(
     get_float(llama_model_meta_key_str(LLAMA_MODEL_META_KEY_SAMPLING_MIROSTAT_ETA),    sparams.mirostat_eta,    common_params_sampling_config::COMMON_PARAMS_SAMPLING_CONFIG_MIROSTAT_ETA);
 }
 
-struct common_init_result common_init_from_params(common_params & params) {
-    common_init_result iparams;
-    auto mparams = common_model_params_to_llama(params);
+struct common_init_result::impl {
+    impl() = default;
+    ~impl() = default;
+
+    llama_model_ptr   model;
+    llama_context_ptr context;
+
+    std::vector<llama_adapter_lora_ptr> lora;
+
+    std::vector<common_sampler_ptr> samplers;
+    std::vector<llama_sampler_seq_config> samplers_seq_config;
+};
+
+common_init_result::common_init_result(common_params & params) :
+    pimpl(new impl{}) {
+    const auto mparams = common_model_params_to_llama(params);
 
     llama_model * model = llama_model_load_from_file(params.model.path.c_str(), mparams);
     if (model == NULL) {
-        LOG_ERR("%s: failed to load model '%s', try reducing --n-gpu-layers if you're running out of VRAM\n",
-            __func__, params.model.path.c_str());
-        return iparams;
+        return;
     }
 
-    common_init_sampler_from_model(model, params.sampling);
+    pimpl->model.reset(model);
 
     const llama_vocab * vocab = llama_model_get_vocab(model);
 
+    // updates params.sampling
+    // TODO: fix naming
+    common_init_sampler_from_model(model, params.sampling);
+
     auto cparams = common_context_params_to_llama(params);
 
+    if (params.sampling.ignore_eos && llama_vocab_eos(vocab) == LLAMA_TOKEN_NULL) {
+        LOG_WRN("%s: warning: vocab does not have an EOS token, ignoring --ignore-eos\n", __func__);
+        params.sampling.ignore_eos = false;
+    }
+
+    // initialize once
+    for (llama_token i = 0; i < llama_vocab_n_tokens(vocab); i++) {
+        if (llama_vocab_is_eog(vocab, i)) {
+            LOG_INF("%s: added %s logit bias = %f\n", __func__, common_token_to_piece(vocab, i).c_str(), -INFINITY);
+            params.sampling.logit_bias_eog.push_back({i, -INFINITY});
+        }
+    }
+
+    if (params.sampling.ignore_eos) {
+        // add EOG biases to the active set of logit biases
+        params.sampling.logit_bias.insert(
+                params.sampling.logit_bias.end(),
+                params.sampling.logit_bias_eog.begin(), params.sampling.logit_bias_eog.end());
+    }
+
+    //if (params.sampling.penalty_last_n == -1) {
+    //    LOG_INF("%s: setting penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
+    //    params.sampling.penalty_last_n = llama_n_ctx(lctx);
+    //}
+
+    //if (params.sampling.dry_penalty_last_n == -1) {
+    //    LOG_INF("%s: setting dry_penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
+    //    params.sampling.dry_penalty_last_n = llama_n_ctx(lctx);
+    //}
+
+    // init the backend samplers as part of the context creation
+    pimpl->samplers.resize(cparams.n_seq_max);
+    pimpl->samplers_seq_config.resize(cparams.n_seq_max);
+
+    for (int i = 0; i < (int) cparams.n_seq_max; ++i) {
+        pimpl->samplers[i].reset(common_sampler_init(model, params.sampling));
+        llama_sampler * backend_chain = common_sampler_chain_backend(pimpl->samplers[i].get());
+        pimpl->samplers_seq_config[i] = { i, backend_chain };
+    }
+
+    cparams.samplers   = pimpl->samplers_seq_config.data();
+    cparams.n_samplers = pimpl->samplers_seq_config.size();
+
     llama_context * lctx = llama_init_from_model(model, cparams);
     if (lctx == NULL) {
         LOG_ERR("%s: failed to create context with model '%s', try reducing --n-gpu-layers if you're running out of VRAM\n",
+                __func__, params.model.path.c_str());
+        return;
+    }
+
+    pimpl->context.reset(lctx);
+}
+
+llama_model * common_init_result::model() {
+    return pimpl->model.get();
+}
+
+llama_context * common_init_result::context() {
+    return pimpl->context.get();
+}
+
+common_sampler * common_init_result::sampler(llama_seq_id seq_id) {
+    return pimpl->samplers[seq_id].get();
+}
+
+std::vector<llama_adapter_lora_ptr> & common_init_result::lora() {
+    return pimpl->lora;
+}
+
+void common_init_result::free_context() {
+    pimpl->context.reset();
+}
+
+common_init_result_ptr common_init_from_params(common_params & params) {
+    common_init_result_ptr res(new common_init_result(params));
+
+    llama_model * model = res->model();
+    if (model == NULL) {
+        LOG_ERR("%s: failed to load model '%s', try reducing --n-gpu-layers if you're running out of VRAM\n",
             __func__, params.model.path.c_str());
-        llama_model_free(model);
-        return iparams;
+        return res;
     }
 
+    llama_context * lctx = res->context();
+    if (lctx == NULL) {
+        LOG_ERR("%s: failed to create context with model '%s', try reducing --n-gpu-layers if you're running out of VRAM\n",
+            __func__, params.model.path.c_str());
+        return res;
+    }
+
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     if (params.ctx_shift && !llama_memory_can_shift(llama_get_memory(lctx))) {
         LOG_WRN("%s: KV cache shifting is not supported for this context, disabling KV cache shifting\n", __func__);
         params.ctx_shift = false;
@@ -1038,10 +1146,7 @@ struct common_init_result common_init_from_params(common_params & params) {
 
         const auto cvec = common_control_vector_load(params.control_vectors);
         if (cvec.n_embd == -1) {
-            llama_free(lctx);
-            llama_model_free(model);
-
-            return iparams;
+            return res;
         }
 
         int err = llama_apply_adapter_cvec(
@@ -1052,10 +1157,7 @@ struct common_init_result common_init_from_params(common_params & params) {
                 params.control_vector_layer_start,
                 params.control_vector_layer_end);
         if (err) {
-            llama_free(lctx);
-            llama_model_free(model);
-
-            return iparams;
+            return res;
         }
     }
 
@@ -1079,10 +1181,7 @@ struct common_init_result common_init_from_params(common_params & params) {
         }
 
         if (!ok) {
-            llama_free(lctx);
-            llama_model_free(model);
-
-            return iparams;
+            return res;
         }
     }
 
@@ -1092,9 +1191,7 @@ struct common_init_result common_init_from_params(common_params & params) {
         lora.reset(llama_adapter_lora_init(model, la.path.c_str()));
         if (lora == nullptr) {
             LOG_ERR("%s: failed to apply lora adapter '%s'\n", __func__, la.path.c_str());
-            llama_free(lctx);
-            llama_model_free(model);
-            return iparams;
+            return res;
         }
 
         char buf[1024];
@@ -1103,43 +1200,13 @@ struct common_init_result common_init_from_params(common_params & params) {
         la.task_name = buf;
         llama_adapter_meta_val_str(la.ptr, "adapter.lora.prompt_prefix", buf, sizeof(buf));
         la.prompt_prefix = buf;
-        iparams.lora.emplace_back(std::move(lora)); // copy to list of loaded adapters
+        res->lora().emplace_back(std::move(lora)); // copy to list of loaded adapters
     }
 
     if (!params.lora_init_without_apply) {
         common_set_adapter_lora(lctx, params.lora_adapters);
     }
 
-    if (params.sampling.ignore_eos && llama_vocab_eos(vocab) == LLAMA_TOKEN_NULL) {
-        LOG_WRN("%s: warning: vocab does not have an EOS token, ignoring --ignore-eos\n", __func__);
-        params.sampling.ignore_eos = false;
-    }
-
-    // initialize once
-    for (llama_token i = 0; i < llama_vocab_n_tokens(vocab); i++) {
-        if (llama_vocab_is_eog(vocab, i)) {
-            LOG_INF("%s: added %s logit bias = %f\n", __func__, common_token_to_piece(lctx, i).c_str(), -INFINITY);
-            params.sampling.logit_bias_eog.push_back({i, -INFINITY});
-        }
-    }
-
-    if (params.sampling.ignore_eos) {
-        // add EOG biases to the active set of logit biases
-        params.sampling.logit_bias.insert(
-                params.sampling.logit_bias.end(),
-                params.sampling.logit_bias_eog.begin(), params.sampling.logit_bias_eog.end());
-    }
-
-    if (params.sampling.penalty_last_n == -1) {
-        LOG_INF("%s: setting penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
-        params.sampling.penalty_last_n = llama_n_ctx(lctx);
-    }
-
-    if (params.sampling.dry_penalty_last_n == -1) {
-        LOG_INF("%s: setting dry_penalty_last_n to ctx_size = %d\n", __func__, llama_n_ctx(lctx));
-        params.sampling.dry_penalty_last_n = llama_n_ctx(lctx);
-    }
-
     if (params.warmup) {
         LOG_WRN("%s: warming up the model with an empty run - please wait ... (--no-warmup to disable)\n", __func__);
 
@@ -1178,12 +1245,11 @@ struct common_init_result common_init_from_params(common_params & params) {
         llama_set_warmup(lctx, false);
     }
 
-    iparams.model.reset(model);
-    iparams.context.reset(lctx);
-
-    return iparams;
+    return res;
 }
 
+common_init_result::~common_init_result() = default;
+
 std::string get_model_endpoint() {
     const char * model_endpoint_env = getenv("MODEL_ENDPOINT");
     // We still respect the use of environment-variable "HF_ENDPOINT" for backward-compatibility.
@@ -1192,7 +1258,9 @@ std::string get_model_endpoint() {
     std::string model_endpoint = "https://huggingface.co/";
     if (endpoint_env) {
         model_endpoint = endpoint_env;
-        if (model_endpoint.back() != '/') model_endpoint += '/';
+        if (model_endpoint.back() != '/') {
+            model_endpoint += '/';
+        }
     }
     return model_endpoint;
 }

common/common.h

@@ -192,7 +192,6 @@ struct common_params_sampling {
 
     std::vector<std::string> dry_sequence_breakers = {"\n", ":", "\"", "*"};     // default sequence breakers for DRY
 
-
     std::vector<enum common_sampler_type> samplers = {
         COMMON_SAMPLER_TYPE_PENALTIES,
         COMMON_SAMPLER_TYPE_DRY,
@@ -213,6 +212,18 @@ struct common_params_sampling {
     std::vector<llama_logit_bias> logit_bias;     // logit biases to apply
     std::vector<llama_logit_bias> logit_bias_eog; // pre-calculated logit biases for EOG tokens
 
+    bool backend_sampling = false; // enable backend sampling
+
+    bool has_logit_bias() const {
+        return !logit_bias.empty();
+    }
+
+    bool is_disabled(enum common_sampler_type type) const;
+
+    // remove disabled samplers
+    // TODO: temporary until all samplers have llama_sampler_backend_ API [LLAMA_SAMPLER_BACKEND]
+    void filter_disabled();
+
     // print the parameters into a string
     std::string print() const;
 };
@@ -648,15 +659,29 @@ std::vector<common_file_info> fs_list_files(const std::string & path);
 // Model utils
 //
 
+struct common_sampler;
+
 // note: defines object's lifetime
 struct common_init_result {
-    llama_model_ptr   model;
-    llama_context_ptr context;
+    common_init_result(common_params & params);
+    ~common_init_result();
+
+    llama_model * model();
+    llama_context * context();
+    common_sampler * sampler(llama_seq_id seq_id);
+
+    std::vector<llama_adapter_lora_ptr> & lora();
 
-    std::vector<llama_adapter_lora_ptr> lora;
+    void free_context();
+
+private:
+    struct impl;
+    std::unique_ptr<impl> pimpl;
 };
 
-struct common_init_result     common_init_from_params(common_params & params);
+using common_init_result_ptr = std::unique_ptr<common_init_result>;
+
+common_init_result_ptr common_init_from_params(common_params & params);
 
 struct llama_model_params     common_model_params_to_llama  (      common_params & params);
 struct llama_context_params   common_context_params_to_llama(const common_params & params);