llama : remove sampling from llama_context

ggml-ci

Author: ggerganov

common/common.cpp

@@ -264,6 +264,10 @@ bool gpt_params_parse_ex(int argc, char ** argv, gpt_params & params) {
         params.kv_overrides.back().key[0] = 0;
     }
 
+    if (params.sparams.seed == LLAMA_DEFAULT_SEED) {
+        params.sparams.seed = time(NULL);
+    }
+
     return true;
 }
 
@@ -294,8 +298,6 @@ bool gpt_params_find_arg(int argc, char ** argv, const std::string & arg, gpt_pa
 
     if (arg == "-s" || arg == "--seed") {
         CHECK_ARG
-        // TODO: this is temporary, in the future the sampling state will be moved fully to llama_sampling_context.
-        params.seed = std::stoul(argv[i]);
         sparams.seed = std::stoul(argv[i]);
         return true;
     }
@@ -1404,7 +1406,6 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
     options.push_back({ "*",           "       --verbose-prompt",       "print a verbose prompt before generation (default: %s)", params.verbose_prompt ? "true" : "false" });
     options.push_back({ "*",           "       --no-display-prompt",    "don't print prompt at generation (default: %s)", !params.display_prompt ? "true" : "false" });
     options.push_back({ "*",           "-co,   --color",                "colorise output to distinguish prompt and user input from generations (default: %s)", params.use_color ? "true" : "false" });
-    options.push_back({ "*",           "-s,    --seed SEED",            "RNG seed (default: %d, use random seed for < 0)", params.seed });
     options.push_back({ "*",           "-t,    --threads N",            "number of threads to use during generation (default: %d)", params.n_threads });
     options.push_back({ "*",           "-tb,   --threads-batch N",      "number of threads to use during batch and prompt processing (default: same as --threads)" });
     options.push_back({ "speculative", "-td,   --threads-draft N",      "number of threads to use during generation (default: same as --threads)" });
@@ -1455,6 +1456,7 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
                                        "       --spm-infill",           "use Suffix/Prefix/Middle pattern for infill (instead of Prefix/Suffix/Middle) as some models prefer this. (default: %s)", params.spm_infill ? "enabled" : "disabled" });
 
     options.push_back({ "sampling" });
+    options.push_back({ "*",           "-s,    --seed SEED",            "RNG seed (default: %d, use random seed for < 0)", sparams.seed });
     options.push_back({ "*",           "       --samplers SAMPLERS",    "samplers that will be used for generation in the order, separated by \';\'\n"
                                                                         "(default: %s)", sampler_type_names.c_str() });
     options.push_back({ "*",           "       --sampling-seq SEQUENCE",
@@ -2199,7 +2201,6 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
     cparams.n_ubatch          = params.n_ubatch;
     cparams.n_threads         = params.n_threads;
     cparams.n_threads_batch   = params.n_threads_batch == -1 ? params.n_threads : params.n_threads_batch;
-    cparams.seed              = params.seed;
     cparams.logits_all        = params.logits_all;
     cparams.embeddings        = params.embedding;
     cparams.rope_scaling_type = params.rope_scaling_type;
@@ -3210,7 +3211,6 @@ void yaml_dump_non_result_info(FILE * stream, const gpt_params & params, const l
 
     fprintf(stream, "rope_freq_base: %f # default: 10000.0\n", params.rope_freq_base);
     fprintf(stream, "rope_freq_scale: %f # default: 1.0\n", params.rope_freq_scale);
-    fprintf(stream, "seed: %u # default: -1 (random seed)\n", params.seed);
     fprintf(stream, "simple_io: %s # default: false\n", params.simple_io ? "true" : "false");
     fprintf(stream, "cont_batching: %s # default: false\n", params.cont_batching ? "true" : "false");
     fprintf(stream, "flash_attn: %s # default: false\n", params.flash_attn ? "true" : "false");

common/common.h

@@ -59,8 +59,6 @@ enum dimre_method {
 };
 
 struct gpt_params {
-    uint32_t seed                 = LLAMA_DEFAULT_SEED; // RNG seed
-
     int32_t n_threads             = cpu_get_num_math();
     int32_t n_threads_draft       =    -1;
     int32_t n_threads_batch       =    -1; // number of threads to use for batch processing (-1 = use n_threads)

common/sampling.cpp

@@ -3,19 +3,10 @@
 #include <random>
 
 struct llama_sampling_context * llama_sampling_init(const struct llama_sampling_params & params, const struct llama_model * model) {
-    auto result = llama_sampling_init(params, llama_sampling_init(model, params.grammar.c_str(), "root"));
-
-    result->owned = true;
-
-    return result;
-}
-
-struct llama_sampling_context * llama_sampling_init(const struct llama_sampling_params & params, struct llama_sampling * smpl) {
     struct llama_sampling_context * result = new llama_sampling_context();
 
     result->params = params;
-    result->owned  = false;
-    result->smpl   = smpl;
+    result->smpl   = llama_sampling_init(model, params.grammar.c_str(), "root");
 
     result->prev.resize(params.n_prev);
 
@@ -27,9 +18,7 @@ struct llama_sampling_context * llama_sampling_init(const struct llama_sampling_
 }
 
 void llama_sampling_free(struct llama_sampling_context * ctx) {
-    if (ctx->owned) {
-        llama_sampling_free(ctx->smpl);
-    }
+    llama_sampling_free(ctx->smpl);
 
     delete ctx;
 }

common/sampling.h

@@ -71,8 +71,6 @@ struct llama_sampling_context {
     // mirostat sampler state
     float mirostat_mu;
 
-    bool owned;
-
     llama_sampling * smpl;
 
     // TODO: replace with ring-buffer
@@ -86,7 +84,6 @@ struct llama_sampling_context {
 
 // Create a new sampling context instance.
 struct llama_sampling_context * llama_sampling_init(const struct llama_sampling_params & params, const struct llama_model * model);
-struct llama_sampling_context * llama_sampling_init(const struct llama_sampling_params & params, struct llama_sampling * smpl);
 
 void llama_sampling_free(struct llama_sampling_context * ctx);
 

examples/batched/batched.cpp

@@ -64,7 +64,7 @@ int main(int argc, char ** argv) {
     ctx_params.n_batch = std::max(n_predict, n_parallel);
 
     llama_context * ctx = llama_new_context_with_model(model, ctx_params);
-    llama_sampling * smpl = llama_get_sampling(ctx);
+    llama_sampling * smpl = llama_sampling_init(model, nullptr, nullptr);
 
     if (ctx == NULL) {
         fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__);

examples/embedding/embedding.cpp

@@ -68,13 +68,7 @@ int main(int argc, char ** argv) {
 
     print_build_info();
 
-    if (params.seed == LLAMA_DEFAULT_SEED) {
-        params.seed = time(NULL);
-    }
-
-    fprintf(stderr, "%s: seed  = %u\n", __func__, params.seed);
-
-    std::mt19937 rng(params.seed);
+    LOG_TEE("%s: seed = %u\n", __func__, params.sparams.seed);
 
     llama_backend_init();
     llama_numa_init(params.numa);

examples/eval-callback/eval-callback.cpp

@@ -151,8 +151,6 @@ int main(int argc, char ** argv) {
 
     print_build_info();
 
-    std::mt19937 rng(params.seed);
-
     llama_backend_init();
     llama_numa_init(params.numa);
 

examples/gritlm/gritlm.cpp

@@ -92,11 +92,10 @@ static std::vector<std::vector<float>> encode(llama_context * ctx, const std::ve
     return result;
 }
 
-static std::string generate(llama_context * ctx, const std::string & prompt, bool stream) {
+static std::string generate(llama_context * ctx, llama_sampling * smpl, const std::string & prompt, bool stream) {
     std::string result;
 
     const llama_model * model = llama_get_model(ctx);
-    llama_sampling * smpl = llama_get_sampling(ctx);
     llama_token eos_token = llama_token_eos(model);
 
     llama_kv_cache_clear(ctx);
@@ -117,7 +116,7 @@ static std::string generate(llama_context * ctx, const std::string & prompt, boo
         inputs.clear();
 
         llama_decode(ctx, bat);
-        auto logits = llama_get_logits_ith(ctx, bat.n_tokens - 1);
+        auto * logits = llama_get_logits_ith(ctx, bat.n_tokens - 1);
 
         auto candidates = std::vector<llama_token_data>(llama_n_vocab(model));
         auto n_candidates = (int32_t)candidates.size();
@@ -173,6 +172,8 @@ int main(int argc, char * argv[]) {
     // create generation context
     llama_context * ctx = llama_new_context_with_model(model, cparams);
 
+    llama_sampling * smpl = llama_sampling_init(model, nullptr, nullptr);
+
     // ### Embedding/Representation ###
     // samples taken from: https://github.com/ContextualAI/gritlm#basic
     {
@@ -209,9 +210,10 @@ int main(int argc, char * argv[]) {
     // GritLM models are not finetuned with system prompts, as you can just include system-like instructions together with your user instruction
     {
         const std::string prompt = "<|user|>\nPlease write me a poem about my recent hike of Mt. Fuji at midnight in the style of Shakespeare.\n<|assistant|>\n";
-        std::string response = generate(ctx, prompt, true);
+        std::string response = generate(ctx, smpl, prompt, true);
     }
 
+    llama_sampling_free(smpl);
     llama_free(ctx);
     llama_free_model(model);
     llama_backend_free();

examples/infill/infill.cpp

@@ -156,16 +156,9 @@ int main(int argc, char ** argv) {
         LOG_TEE("%s: warning: scaling RoPE frequency by %g.\n", __func__, params.rope_freq_scale);
     }
 
-    LOG_TEE("%s: build = %d (%s)\n",      __func__, LLAMA_BUILD_NUMBER, LLAMA_COMMIT);
-    LOG_TEE("%s: built with %s for %s\n", __func__, LLAMA_COMPILER, LLAMA_BUILD_TARGET);
+    print_build_info();
 
-    if (params.seed == LLAMA_DEFAULT_SEED) {
-        params.seed = time(NULL);
-    }
-
-    LOG_TEE("%s: seed  = %u\n", __func__, params.seed);
-
-    std::mt19937 rng(params.seed);
+    LOG_TEE("%s: seed = %u\n", __func__, params.sparams.seed);
 
     LOG("%s: llama backend init\n", __func__);
     llama_backend_init();
@@ -348,7 +341,7 @@ int main(int argc, char ** argv) {
 
     std::vector<llama_token> embd;
 
-    ctx_sampling = llama_sampling_init(sparams, llama_get_sampling(ctx));
+    ctx_sampling = llama_sampling_init(sparams, model);
 
     while (n_remain != 0 || params.interactive) {
         // predict

examples/llava/llava-cli.cpp

@@ -191,7 +191,7 @@ static void process_prompt(struct llava_context * ctx_llava, struct llava_image_
 
     LOG_TEE("\n");
 
-    struct llama_sampling_context * ctx_sampling = llama_sampling_init(params->sparams, llama_get_sampling(ctx_llava->ctx_llama));
+    struct llama_sampling_context * ctx_sampling = llama_sampling_init(params->sparams, ctx_llava->model);
     if (!ctx_sampling) {
         fprintf(stderr, "%s: failed to initialize sampling subsystem\n", __func__);
         exit(1);