#!/usr/bin/env Rscript library(tidyverse) library(ggthemes) library(ggbreak) library(readr) library(patchwork) args <- commandArgs(trailingOnly = TRUE) if (length(args) == 0) { csv_fp <- "data/combined.csv" } else { csv_fp <- args[1] } all_cols <- c( "model", "algo", "config", "attempt", "bpp", "ppl_wikitext", "ppl_c4" ) df_all <- read_csv(csv_fp) |> select(all_of(all_cols)) |> mutate( model = factor( model, levels = c("Llama-2-7b-hf", "Meta-Llama-3-8B", "Llama-2-13b-hf"), labels = c("Llama-2-7B", "Llama-3-8B", "Llama-2-13B") ), algo = factor( algo, levels = c("mxq", "fp16", "awq", "gptq", "bnb", "hqq"), labels = c("MXQ", "FP16", "AWQ", "GPTQ", "BnB", "HQQ"), ), attempt = factor( attempt, levels = c( "mxq1", "head-prioritized_1_05", "head-prioritized_1_10", "tail-prioritized_1_05", "tail-prioritized_1_10", "tail-prioritized_1_15", "tail-prioritized_1_20", "kurt-scaled" ), labels = c( "MXQ1", "HP105", "HP110", "TP105", "TP110", "TP115", "TP120", "KURT-SCALED" ), ) ) df_wikitxt_all <- df_all |> rename(ppl = ppl_wikitext) df_c4_all <- df_all |> rename(ppl = ppl_c4) # Plot Llama-2-7b memory drop vs PPL loss --------------------------------- model_name <- "Llama-2-7B" df_wikitxt <- df_wikitxt_all |> filter( grepl(model_name, model) & bpp >= 2.5 ) min_ppl <- min(df_wikitxt$ppl) min_bpp <- min(df_wikitxt$bpp) plt5 <- ggplot( subset(df_wikitxt, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_wikitxt, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16, name = "% Memery Reduction") ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.20), breaks = seq(5.18, 5.18 * 1.20, 0.20), sec.axis = sec_axis(~ 100 * (. - 5.18) / 5.18, name = "% Degradation") ) + labs(x = "WikiText2", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme(legend.position = "none") + facet_wrap(~model, scales = "free") + scale_color_solarized() df_c4 <- df_c4_all |> filter( grepl(model_name, model) & bpp >= 2.5 ) min_ppl <- min(df_c4$ppl) min_bpp <- min(df_c4$bpp) plt6 <- ggplot( subset(df_c4, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_c4, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16) ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.20), breaks = seq(6.95, 6.95 * 1.20, 0.25), sec.axis = sec_axis(~ 100 * (. - 6.95) / 6.95, name = "% Degradation") ) + labs(x = "C4", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme( legend.position = "bottom", legend.text = element_text(size = 14), legend.title = element_text(size = 14) ) + facet_wrap(~model, scales = "free") + scale_color_solarized() final_plot1 <- plt5 / plt6 + plot_layout(heights = c(2, 3)) pdf.options(reset = TRUE, onefile = FALSE) ggsave( paste0("pdfs/", "mxq-", model_name, ".pdf"), plot = final_plot1, width = 10, height = 7 ) # Plot Llama-2-13b memory drop vs PPL loss --------------------------------- model_name <- "Llama-2-13B" df_wikitxt <- df_wikitxt_all |> filter( model == model_name & bpp >= 2.5 ) min_ppl <- min(df_wikitxt$ppl) min_bpp <- min(df_wikitxt$bpp) plt1 <- ggplot( subset(df_wikitxt, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_wikitxt, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16, name = "% Memery Reduction") ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.20), breaks = seq(4.63, 4.63 * 1.20, 0.20), sec.axis = sec_axis(~ 100 * (. - 4.63) / 4.63, name = "% Degradation") ) + labs(x = "WikiText2", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme( legend.position = "none" ) + facet_wrap(~model, scales = "free") + scale_color_solarized() df_c4 <- df_c4_all |> filter( grepl(model_name, model) & bpp >= 2.5 ) min_ppl <- min(df_c4$ppl) min_bpp <- min(df_c4$bpp) plt2 <- ggplot( subset(df_c4, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_c4, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16) ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.20), breaks = seq(6.45, 6.45 * 1.20, 0.20), sec.axis = sec_axis(~ 100 * (. - 6.45) / 6.45, name = "% Degradation") ) + labs(x = "C4", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme( legend.position = "bottom", legend.text = element_text(size = 14), legend.title = element_text(size = 14) ) + facet_wrap(~model, scales = "free") + scale_color_solarized() final_plot2 <- plt1 / plt2 + plot_layout(heights = c(2, 3)) pdf.options(reset = TRUE, onefile = FALSE) ggsave( paste0("pdfs/", "mxq-", model_name, ".pdf"), plot = final_plot2, width = 10, height = 7 ) # Plot Llama-3-8B memory drop vs PPL loss --------------------------------- model_name <- "Llama-3-8B" df_wikitxt <- df_wikitxt_all |> filter( grepl(model_name, model) & bpp >= 2.5 ) min_ppl <- round(min(df_wikitxt$ppl), digits = 2) min_bpp <- round(min(df_wikitxt$bpp), digits = 2) plt3 <- ggplot( subset(df_wikitxt, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_wikitxt, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16, name = "% Memery Reduction") ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.35), breaks = seq(min_ppl, min_ppl * 1.35, 0.20), sec.axis = sec_axis(~ 100 * (. - 5.81) / 5.81, name = "% Degradation") ) + labs(x = "WikiText2", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme( legend.position = "none" ) + facet_wrap(~model, scales = "free") + scale_color_solarized() df_c4 <- df_c4_all |> filter( grepl(model_name, model) & bpp >= 2.5 ) min_ppl <- round(min(df_c4$ppl), digits = 2) min_bpp <- round(min(df_c4$bpp), digits = 2) plt4 <- ggplot( subset(df_c4, algo != "MXQ"), aes(x = bpp, y = ppl), ) + geom_point( data = subset(df_c4, algo == "MXQ"), size = 0.5, aes(shape = algo, color = attempt, y = ppl) ) + geom_point(size = 1.5, aes(shape = algo, y = ppl)) + geom_hline( yintercept = min_ppl * 1.02, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl * 1.01, linetype = "dashed", size = 0.1, color = "blue" ) + geom_hline( yintercept = min_ppl, linetype = "dashed", size = 0.1, color = "blue" ) + annotate("text", x = 15.8, y = min_ppl * 1.00, label = "FP16") + scale_x_break(c(5.5, 15.6)) + scale_x_continuous( limits = c(2.8, 16.2), breaks = seq(2.8, 5.5, 0.20), sec.axis = sec_axis(~ 100 * (16 - .) / 16) ) + scale_y_continuous( limits = c(min_ppl * 0.99, min_ppl * 1.35), breaks = seq(min_ppl, min_ppl * 1.35, 0.20), sec.axis = sec_axis(~ 100 * (. - 8.98) / 8.98, name = "% Degradation") ) + labs(x = "C4", y = "Perplexity") + theme_gray(base_size = 14) + guides( shape = guide_legend(title = "Method:"), color = guide_legend(title = "Attempt:") ) + theme( legend.position = "bottom", legend.text = element_text(size = 14), legend.title = element_text(size = 14) ) + facet_wrap(~model, scales = "free") + scale_color_solarized() final_plot3 <- plt3 / plt4 + plot_layout(heights = c(4, 5)) pdf.options(reset = TRUE, onefile = FALSE) ggsave( paste0("pdfs/", "mxq-", model_name, ".pdf"), plot = final_plot3, width = 12, height = 9 )